From john.hearns at mclaren.com Thu Sep 1 04:27:26 2011 From: john.hearns at mclaren.com (Hearns, John) Date: Thu, 1 Sep 2011 09:27:26 +0100 Subject: [Beowulf] materials for air shroud? References: Message-ID: <207BB2F60743C34496BE41039233A809075A6747@MRL-PWEXCHMB02.mil.tagmclarengroup.com> > > Plastic tape covering the aluminum.. 20 mil "pipe wrap" is useful > stuff. 3M VHB double stick foam tape to hold it in place. > > But, enough of this feeble lash-up idea: I think the real solution is > to have a second cluster doing a complete finite element model of the > instantaneous temperature distribution within the processor in > question, You called? John Hearns | CFD Hardware Specialist | McLaren Racing Limited McLaren Technology Centre, Chertsey Road, Woking, Surrey GU21 4YH, UK > driving a set of actuators to form a dynamically optimized > shroud. Sadly, moveable aerodynamic surfaces, with the exception of the DRS, are forbidden by FIA regulations. Regarding the material http://f1-dictionary.110mb.com/heat_shielding.html The contents of this email are confidential and for the exclusive use of the intended recipient. If you receive this email in error you should not copy it, retransmit it, use it or disclose its contents but should return it to the sender immediately and delete your copy. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Thu Sep 1 09:47:14 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Thu, 1 Sep 2011 06:47:14 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <207BB2F60743C34496BE41039233A809075A6747@MRL-PWEXCHMB02.mil.tagmclarengroup.com> Message-ID: On 9/1/11 1:27 AM, "Hearns, John" wrote: >> >> Plastic tape covering the aluminum.. 20 mil "pipe wrap" is useful >> stuff. 3M VHB double stick foam tape to hold it in place. >> >> But, enough of this feeble lash-up idea: I think the real solution is >> to have a second cluster doing a complete finite element model of the >> instantaneous temperature distribution within the processor in >> question, > >You called? > >John Hearns | CFD Hardware Specialist | McLaren Racing Limited >McLaren Technology Centre, Chertsey Road, Woking, Surrey GU21 4YH, UK > > >> driving a set of actuators to form a dynamically optimized >> shroud. > >Sadly, moveable aerodynamic surfaces, with the exception of the DRS, are >forbidden by FIA regulations. > > Oh ho.. You are more bold than I, so we have real-time adaptive cooling shrouds on a cluster, modeled by a cluster, implemented by a cluster installed on a F1 car using the first cluster to optimize performance. Movable aerodynamic surfaces or forbidden, even if they're just inside a component? (yeah, I can see that.. Sure, it's just a cooling fan for the onboard electronics.. All we happened to have is a 1000 HP blower) (I suppose one could use a high rate data link, but where's the thrill in that.. Clusters everywhere..) > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From ntmoore at gmail.com Thu Sep 1 11:22:42 2011 From: ntmoore at gmail.com (Nathan Moore) Date: Thu, 1 Sep 2011 10:22:42 -0500 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: Steel has a lousy thermal conductivity compared to Aluminum. They sell the "dippable" electrical tape at most home improvement stores. Foam double-stick tape is indeed amazing stuff... On Wed, Aug 31, 2011 at 4:05 PM, Robert G. Brown wrote: > On Wed, 31 Aug 2011, Lux, Jim (337C) wrote: > > Also thin aluminum. You can get aluminum sheeting that you can cut with > scissors and that is easy to bend into shapes if you have a bending jig > (or can make one with two pieces of board stock and a vise). Cheap, > fireproof, meltproof at any temperatures you're likely to reach, no > toxic fumes in a fire, can be glued or screwed. The one drawback is > that it is a PITA to weld or solder if that's important to you, but for > an air shroud you can probably make compression joints (interlocking U > rims, squeezed down) that are adequate. > > Most hardware stores (roof flashing), some auto parts or hobby stores. > Copper too, but more expensive. Don't know about thin "enough" sheet > steel, but probably -- copper or steel would both weld or solder easily. > > rgb > > > Cardboard? Card stock? Masking tape? White glue? (that's what I usually > use for cooling ducts.. easy to cut, glue, tape..) It's no more flammable > than plastic, and it doesn't melt and get soft. Papier Mache, works too. > > > > On the other hand, if you want to mold a smooth curve, then plastic is > the way to go. Vacuforming can make a very nice thing, and the form is made > out of wood (usually), but you don't need to go to that extreme.. you get > some nice thermoplastic, put it in hot water to get it soft, and mold as > needed. (yes, you could use those old LPs you've got stashed away.. ) > > > > Thin, cuttable plastic could be polyethylene (not necessarily High > density) or similar. Polystyrene and acrylic tend to be more brittle. ABS > is very nice to work with. PVC is also easy to work with. Nylon is another > possibility. > > > > Do you want to be able to glue it? > > > > What I would do is call up profesionalplastics.com formerly Cadillac > Plastics (many outlets nationwide) and see what they have. It might be more > useful to find a retail outlet and go look through their scrap bin.. Before > Gem-O-Lite in Woodland Hills went out of business, that's where I used to > go. Plastic Depot in Burbank has a huge selection. > > > > Drive over there, and ask the counter folks what would work for you. > $10-20 will get you more plastic than you know what to do with. > > > > Art supply places (e.g. Blick on Raymond.. any of the countless Michaels > or Aaron Bros) also carry sheet plastic, but I find the plastic places tend > to have more variety, and more practical information about use for > "engineering" applications. > > > > > > Jim Lux > > +1(818)354-2075 > > > >> -----Original Message----- > >> From: beowulf-bounces at beowulf.org [mailto:beowulf-bounces at beowulf.org] > On Behalf Of David Mathog > >> Sent: Wednesday, August 31, 2011 10:29 AM > >> To: beowulf at beowulf.org > >> Subject: [Beowulf] materials for air shroud? > >> > >> Anybody know of a nice cheap, high melting point, easy to work with > >> sheet material, for making a custom air shroud? > >> > >> We have one box with stuff in it that looks similar to HDPE, the > >> material the white flexible cutting boards are made of, but it is a bit > >> thinner and more rigid that that. Unfortunately there are no markings > >> on it, so HDPE is just a guess. Whatever it is, it cut easily with > >> scissors (I had to trim it slightly at one point.) > >> > >> Background. We have an older Supermicro SC-823 server with dual > >> processors. The air shroud it came with only covers the first > >> processor. That didn't matter much when it had two low power processors > >> in it, but after upgrading it to dual Opteron 280s, the uncovered second > >> one runs considerably hotter than the covered front one. (Swapping the > >> processors around didn't help - the heat stayed where it was, so a > >> ventilation issue, not a processor issue.) Supermicro does make a newer > >> shroud which extends to the back of the case, but the manual (google for > >> "SC-823 air shroud user's guide") indicates that it is designed for > >> Intel CPUs. So it may or may not fit around the Opterons. > >> > >> The redesigned air shroud will probably work, but I'm about 90% > >> confident that taping a sheet of plastic onto the back of the existing > >> shroud would work as well - if I can find a plastic that won't flap > >> around or melt. > >> > >> Thanks, > >> > >> David Mathog > >> mathog at caltech.edu > >> Manager, Sequence Analysis Facility, Biology Division, Caltech > >> _______________________________________________ > >> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin > Computing > >> To change your subscription (digest mode or unsubscribe) visit > >> http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > > > Robert G. Brown http://www.phy.duke.edu/~rgb/ > Duke University Dept. of Physics, Box 90305 > Durham, N.C. 27708-0305 > Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu > > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- - - - - - - - - - - - - - - - - - - - - - Nathan Moore Associate Professor, Physics Winona State University - - - - - - - - - - - - - - - - - - - - - -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From mathog at caltech.edu Wed Sep 7 19:05:51 2011 From: mathog at caltech.edu (mathog) Date: Wed, 07 Sep 2011 16:05:51 -0700 Subject: [Beowulf] =?utf-8?q?SGE_qconf=2C_renaming_hosts/queues=3F?= Message-ID: Just finished migrating most everything on the cluster from the old headnode to a new one, but the old headnode has to remain functional for a while. Even the node name moved, and therein lies the rub. In testing I had node queue_names mendel mendel_something newsaf newsaf_something and that worked fine. Now I want to rename node mendel to oldsaf, and all the queue names to oldsaf_something. Currently the mendel_* queues are all jammed, because by IP number mendel is the same machine as newsaf, and it isn't even the same architecture. I must be missing something, because while I see qconf/qmon options to modify nodes and queues, I don't see how to rename these. Failing that, is there an easy way to dump the host/queue description to a text file, edit that, and then re-create from the modified descriptions? There are a bunch of text configuration files in the SGE directory, but I'm guessing that editing/renaming those directly is a big no no. Thanks, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From kilian.cavalotti.work at gmail.com Thu Sep 8 03:04:14 2011 From: kilian.cavalotti.work at gmail.com (Kilian CAVALOTTI) Date: Thu, 8 Sep 2011 09:04:14 +0200 Subject: [Beowulf] SGE qconf, renaming hosts/queues? In-Reply-To: References: Message-ID: Hi David, On Thu, Sep 8, 2011 at 1:05 AM, mathog wrote: > Failing that, is there an easy way to dump > the host/queue description to a text file, edit > that, and then re-create from the modified descriptions? ?There are a > bunch of text configuration files in the > SGE directory, but I'm ?guessing that editing/renaming those directly > is a big no no. Well, it'd probably work, but the risk is to forget some. Instead, you can use qconf to dump queue definitions. Something along the lines of: # qconf -sql | while read QUEUE ; do echo "****************** ${QUEUE} **********************" qconf -sq ${QUEUE} done For host renaming, though, the safest route is probably to: 0. remove the old queues (qconf -dq ) 1. remove old nodes from SGE configuration (qconf -de for execution hosts) 2. change hostnames on the old nodes (depends on your OS) 3. re-add the new old nodes (qconf -ae for execution hosts) 4. recreate the queues from the previous dump (qconf -aq , and paste/edit the dumps) Cheers, -- Kilian _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From reuti at staff.uni-marburg.de Thu Sep 8 05:40:29 2011 From: reuti at staff.uni-marburg.de (Reuti) Date: Thu, 8 Sep 2011 11:40:29 +0200 Subject: [Beowulf] SGE qconf, renaming hosts/queues? In-Reply-To: References: Message-ID: Hi, Am 08.09.2011 um 09:04 schrieb Kilian CAVALOTTI: > Hi David, > > On Thu, Sep 8, 2011 at 1:05 AM, mathog wrote: >> Failing that, is there an easy way to dump >> the host/queue description to a text file, edit >> that, and then re-create from the modified descriptions? There are a >> bunch of text configuration files in the >> SGE directory, but I'm guessing that editing/renaming those directly >> is a big no no. > > Well, it'd probably work, but the risk is to forget some. > > Instead, you can use qconf to dump queue definitions. Something along > the lines of: yep, there are also prepared scripts in $SGE_ROOT/util/upgrade_modules/ called "save_sge_config.sh" and "load_sge_config.sh" . While the "save_sge_config.sh" will do the correct thing, the "load_sge_config.sh" is meant to be used in an empty SGE installation. It might work in some cases even by overriding the previous setting, but if you have e.g. mutal settings in two queues for subordination of the other queue, you first have to remove this subordination in one queue before you can remove the other queue at all. And the "load_sge_config.sh" isn't so sophisticated to cover these cases. -- Reuti > # qconf -sql | while read QUEUE ; do > echo "****************** ${QUEUE} **********************" > qconf -sq ${QUEUE} > done > > For host renaming, though, the safest route is probably to: > > 0. remove the old queues (qconf -dq ) > 1. remove old nodes from SGE configuration (qconf -de > for execution hosts) > 2. change hostnames on the old nodes (depends on your OS) > 3. re-add the new old nodes (qconf -ae for execution hosts) > 4. recreate the queues from the previous dump (qconf -aq , and > paste/edit the dumps) > > Cheers, > -- > Kilian > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From reuti at staff.uni-marburg.de Thu Sep 8 05:44:54 2011 From: reuti at staff.uni-marburg.de (Reuti) Date: Thu, 8 Sep 2011 11:44:54 +0200 Subject: [Beowulf] SGE qconf, renaming hosts/queues? In-Reply-To: References: Message-ID: Am 08.09.2011 um 01:05 schrieb mathog: > Just finished migrating most everything on the cluster from the old > headnode to a new one, but > the old headnode has to remain functional for a while. Even the node > name moved, and > therein lies the rub. In testing I had > > node queue_names > mendel mendel_something > newsaf newsaf_something > > and that worked fine. Now I want to rename node mendel to oldsaf, and > all the queue names This is not forseen. You can only copy/paste the setting to a new queue and then delete the original one afterwards. Do you have one queue per node? For changing hostnames it's best to remove the nodes first from SGE and then add them again with the new name. -- Reuti > to oldsaf_something. Currently the mendel_* queues are all jammed, > because by IP number mendel > is the same machine as newsaf, and it isn't even the same architecture. > > I must be missing something, because while I see qconf/qmon options to > modify nodes and queues, I don't > see how to rename these. Failing that, is there an easy way to dump > the host/queue description to a text file, edit > that, and then re-create from the modified descriptions? There are a > bunch of text configuration files in the > SGE directory, but I'm guessing that editing/renaming those directly > is a big no no. > > Thanks, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Thu Sep 8 15:03:02 2011 From: mathog at caltech.edu (mathog) Date: Thu, 08 Sep 2011 12:03:02 -0700 Subject: [Beowulf] =?utf-8?q?SGE_qconf=2C_renaming_hosts/queues=3F?= In-Reply-To: References: Message-ID: <7a37f12f8806672162ccb4f3adaff114@saf.bio.caltech.edu> On Thu, 8 Sep 2011 09:04:14 +0200, Kilian CAVALOTTI wrote: > Hi David, > Kilian CAVALOTTI wrote: > On Thu, Sep 8, 2011 at 1:05 AM, mathog wrote: >> Failing that, is there an easy way to dump >> the host/queue description to a text file, edit >> that, and then re-create from the modified descriptions? ?There are >> a >> bunch of text configuration files in the >> SGE directory, but I'm ?guessing that editing/renaming those >> directly >> is a big no no. > > Well, it'd probably work, but the risk is to forget some. > > Instead, you can use qconf to dump queue definitions. Something along > the lines of: > > # qconf -sql | while read QUEUE ; do > echo "****************** ${QUEUE} **********************" > qconf -sq ${QUEUE} > done That was a big help. After dumping both the queues and execution hosts this way it was easy to delete the bogus queues and relatively easy to delete the one bogus host entry (after running down everything else that referenced it). (Bogus host because the address now points to a machine of a different type.) The only major complication was that this doesn't quite work: qconf -se hostname >foo qconf -Ae foo there are things -se writes that -Ae won't consume. After editing out the "processors" and "load_values" lines in foo -Ae was happy. Thanks, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Tue Sep 13 16:05:30 2011 From: mathog at caltech.edu (mathog) Date: Tue, 13 Sep 2011 13:05:30 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: References: Message-ID: FYI, Some sheets of 1/16" polypropylene were obtained from smallparts.com. (This is very similar or identical to the case plastic used on DLT and other larger tape cartridges.) So far this material is relatively easy to work with. Plus: 1. Can be cut with a (big) pair of scissors. 2. Can be bent by hand 3. Melting point >130C 4. Can be welded with high temperature hot glue (not verified) 5. Rigid enough for a duct (but not for anything load bearing) Minus: 1. Holds a static charge. (Probably not more than any other plastic though.) 2. Few adhesives stick to it. Generic brown masking tape holds pretty well. Some details I tried bending a sample two ways. First, just putting it on the edge of a desk and then folding it. That worked but it turned white and thinned somewhat at the bend, and it slowly opened back up again from 90 to about 100 degrees. That edge was then melted by gently rubbing it with the barrel of a soldering iron. (Find the point just below where it smokes). After remelting the edge was once again clear and the angle stayed firmly at 90. Another 90 degree bend was made by first heating the flat plastic on both sides with the soldering iron barrel and then bending. That edge turned out a little better, the thickness of the plastic through the turn was very close to that of the flat parts. Hard to get it to just the right temperature though, so there was some smoke. Either method made a good enough 90 degree bend for an air shroud. Finally, I tried gluing two pieces at right angles using a high melt hot glue. The hot glue gun claims to run at 395F, and the glue stick was nothing special, just generic high temperature hot melt. Mixed results. After it cooled and was allowed to set overnight I tried to tear the two pieces apart by hand, pulling in opposite directions, and they held together. However, I was able to snap the pieces apart by folding it at the junction. (Applying quite a lot of torque to the junction.) The glue completely let go of the top of the "T", all of it stayed on the vertical part. At this point it was easy to peel the rest of the glue off. Seems like the bonding was good perpendicular to the surface, but pretty weak parallel to it. If the piece wasn't physically abused it would likely hold together in an air shroud. I had read somewhere that the hot glue melts the polypropylene so that it was effectively a weld, but that is not how it turned out with this glue. Neither piece of plastic was distorted where the glue had been, so clearly not melted. There are specialty hot melts made of polyethylene or polypropylene, and those may actually weld this material. Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Tue Sep 13 19:51:23 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Tue, 13 Sep 2011 16:51:23 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: > -----Original Message----- > From: beowulf-bounces at beowulf.org [mailto:beowulf-bounces at beowulf.org] On Behalf Of mathog > Sent: Tuesday, September 13, 2011 1:06 PM > To: beowulf at beowulf.org > Subject: Re: [Beowulf] materials for air shroud? > > FYI, > > Some sheets of 1/16" polypropylene were obtained from smallparts.com. > (This is very similar or identical > to the case plastic used on DLT and other larger tape cartridges.) So > far this material is > relatively easy to work with. > > I tried bending a sample two ways. First, just putting it on the edge of a desk and then > folding it. That worked but it turned white and thinned somewhat at the bend, and it > slowly opened back up again from 90 to about 100 degrees. That edge was then melted by > gently rubbing it with the barrel of a soldering iron. (Find the point ust below where > it smokes). A hot air gun works nicely for this. A couple pieces of wood laid on TOP of the plastic where you are blowing the hot air forms a nice channel so that the width of the heated strip is uniform. After you get it soft, you pick it up and put it on your form (i.e. the desk edge). A 1 cm gap works nicely. > Neither piece of plastic > was distorted where the glue had been, so clearly not melted. There > are specialty hot melts made > of polyethylene or polypropylene, and those may actually weld this > material. Plastic welding uses what looks like a big soldering iron with an air hose and a hollow tip. The tip usually has a sort of flat spoon sticking out, maybe 1 cm long and 0.3-0.5 cm wide that you can use to push the melted/soft plastic around. You hold the welder in one hand and the rod in the other and it's much like regular welding (except easier). You get the parent material really soft, and then feed the rod in to make a fillet, just like in regular welding. The trick is welding, not brazing (i.e. the parent material has to melt enough to mix with the melted rod.. otherwise you're basically hot gluing two pieces of plastic together) Jim Lux +1(818)354-2075 _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Wed Sep 14 09:02:21 2011 From: prentice at ias.edu (Prentice Bisbal) Date: Wed, 14 Sep 2011 09:02:21 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E70A5DD.6070701@ias.edu> Instead of a soldering iron, have you tried using just an old clothing iron? That should have a more adjustable temperature setting. And once it's heated up, you can fold it like origami paper. (I challenge this list to produce an origami crane this way!) I would keep a large bowl of cold water nearby, so that after you make your folds, you can dunk your finished product in the cold water to thermoset the plastic. For cutting the plastic, I highly recommend these Wiss MultiMaster multi-purpose shears: http://www.cooperhandtools.com/brands/CF_Files/model_detail.cfm?upc=037103126481 I bought them to cut up a rug, thinking I'd never use them again. They are now probably one of the most used tools in my toolbox. They cut through everything: sheetmetal, hardware cloth, zip-ties, obnoxious plastic packaging - you name it! Since they have a compound action, you don't need to use a lot of force, which makes it easier to control your cuts. I think I paid $12-$15 at Lowe's for mine, and they're worth every penny. On 09/13/2011 04:05 PM, mathog wrote: > FYI, > > Some sheets of 1/16" polypropylene were obtained from smallparts.com. > (This is very similar or identical > to the case plastic used on DLT and other larger tape cartridges.) So > far this material is > relatively easy to work with. > > Plus: > 1. Can be cut with a (big) pair of scissors. > 2. Can be bent by hand > 3. Melting point >130C > 4. Can be welded with high temperature hot glue (not verified) > 5. Rigid enough for a duct (but not for anything load bearing) > > Minus: > 1. Holds a static charge. (Probably not more than any other plastic > though.) > 2. Few adhesives stick to it. Generic brown masking tape holds pretty > well. > > Some details > > I tried bending a sample two ways. First, just putting it on the edge > of a desk and then > folding it. That worked but it turned white and thinned somewhat at > the bend, and it > slowly opened back up again from 90 to about 100 degrees. That edge was > then melted by > gently rubbing it with the barrel of a soldering iron. (Find the point > just below where > it smokes). After remelting the edge was once again clear and the > angle stayed firmly at 90. > Another 90 degree bend was made by first heating the flat plastic on > both sides with the > soldering iron barrel and then bending. That edge turned out a little > better, the thickness > of the plastic through the turn was very close to that of the flat > parts. Hard to get it to > just the right temperature though, so there was some smoke. Either > method made a good enough > 90 degree bend for an air shroud. > > Finally, I tried gluing two pieces at right angles using a high melt > hot glue. The hot glue gun > claims to run at 395F, and the glue stick was nothing special, just > generic high temperature > hot melt. Mixed results. After it cooled and was allowed to set > overnight I tried to > tear the two pieces apart by hand, pulling in opposite directions, and > they held together. > However, I was able to snap the pieces apart by folding it at the > junction. (Applying quite > a lot of torque to the junction.) The glue completely let go of the > top of the "T", all of it > stayed on the vertical part. At this point it was easy to peel the > rest of the glue off. > Seems like the bonding was good perpendicular to the surface, but > pretty weak > parallel to it. If the piece wasn't physically abused it would likely > hold together in an > air shroud. I had read somewhere that the hot glue melts the > polypropylene so that it was > effectively a weld, but that is not how it turned out with this glue. > Neither piece of plastic > was distorted where the glue had been, so clearly not melted. There > are specialty hot melts made > of polyethylene or polypropylene, and those may actually weld this > material. > > Regards, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From dnlombar at ichips.intel.com Wed Sep 14 11:29:30 2011 From: dnlombar at ichips.intel.com (David N. Lombard) Date: Wed, 14 Sep 2011 08:29:30 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <20110914152930.GA27207@nlxcldnl2.cl.intel.com> On Tue, Sep 13, 2011 at 02:05:30PM -0600, mathog wrote: > FYI, > > Finally, I tried gluing two pieces at right angles using a high melt > hot glue. The hot glue gun > claims to run at 395F, and the glue stick was nothing special, just > generic high temperature > hot melt. Mixed results. After it cooled and was allowed to set > overnight I tried to > tear the two pieces apart by hand, pulling in opposite directions, and > they held together. > However, I was able to snap the pieces apart by folding it at the > junction. (Applying quite > a lot of torque to the junction.) The glue completely let go of the > top of the "T", all of it > stayed on the vertical part. At this point it was easy to peel the > rest of the glue off. > Seems like the bonding was good perpendicular to the surface, but > pretty weak > parallel to it. If the piece wasn't physically abused it would likely > hold together in an > air shroud. I had read somewhere that the hot glue melts the > polypropylene so that it was > effectively a weld, but that is not how it turned out with this glue. > Neither piece of plastic > was distorted where the glue had been, so clearly not melted. There > are specialty hot melts made > of polyethylene or polypropylene, and those may actually weld this > material. This is my expectation with /consumer/ hot glues--they're good for temporary work, but are not a long-term adhesive. Did you try some of the hobby plastic adhesives? Plastruct has a "Plastic Weld" solvent cement. They list ABS, Butyrate, Styrene, Acrylic, and "more". Perhaps "more" includes your propylne. Another brand is "Tenax 7R". The former is MEK and the latter MC; both need careful handling. Finally, you could try the PVC Priming Fluid or even PVC Pipe Adhesives. These are all solvent glues and will weld the materials. -- David N. Lombard, Intel, Irvine, CA I do not speak for Intel Corporation; all comments are strictly my own. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From john.hearns at mclaren.com Wed Sep 14 12:35:58 2011 From: john.hearns at mclaren.com (Hearns, John) Date: Wed, 14 Sep 2011 17:35:58 +0100 Subject: [Beowulf] materials for air shroud? References: <20110914152930.GA27207@nlxcldnl2.cl.intel.com> Message-ID: <207BB2F60743C34496BE41039233A80907CD9C33@MRL-PWEXCHMB02.mil.tagmclarengroup.com> Thinking about this, would vacuforming not be a good option? http://astore.amazon.com/google022-20/detail/B001N0BTVU The contents of this email are confidential and for the exclusive use of the intended recipient. If you receive this email in error you should not copy it, retransmit it, use it or disclose its contents but should return it to the sender immediately and delete your copy. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From dnlombar at ichips.intel.com Wed Sep 14 18:56:24 2011 From: dnlombar at ichips.intel.com (David N. Lombard) Date: Wed, 14 Sep 2011 15:56:24 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <207BB2F60743C34496BE41039233A80907CD9C33@MRL-PWEXCHMB02.mil.tagmclarengroup.com> References: <20110914152930.GA27207@nlxcldnl2.cl.intel.com> <207BB2F60743C34496BE41039233A80907CD9C33@MRL-PWEXCHMB02.mil.tagmclarengroup.com> Message-ID: <20110914225624.GA31941@nlxcldnl2.cl.intel.com> On Wed, Sep 14, 2011 at 10:35:58AM -0600, Hearns, John wrote: > Thinking about this, > would vacuforming not be a good option? Alternately, try "smash molding". Short story: heat the plastic and pull it over a form. Make a clamp fixture that holds the edges of the plastic on all sides. Arrange the mold in a another fixture so it sits above (stands proud of) the fixture. Heat the plastic until plastic deformation starts. then pull it down over the mold, well past the bottom of the mold. -- David N. Lombard, Intel, Irvine, CA I do not speak for Intel Corporation; all comments are strictly my own. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Thu Sep 15 18:33:44 2011 From: mathog at caltech.edu (mathog) Date: Thu, 15 Sep 2011 15:33:44 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: References: Message-ID: A not so short story about air flow... Yesterday I did some experimenting with different baffles and ducts, each built temporarily out of the cardboard backs from yellow notepads and held together with masking tape. (Not worried about a fire, since it only ran for 10 minutes at a time like that, and I was right there to yank the plug and rip out the cardboard if something went wrong.) The system has a Supermicro H8DC8 motherboard in a Supermicro case. This one: http://www.supermicro.com/products/chassis/2U/823/SC823S-550LP.cfm This is what that motherboard looks like without heat sinks: http://www.supermicro.com/a_images/products/Aplus/MB/H8DC8_spec.jpg and here is a very similar motherboard with heat sinks in place (but not my motherboard, which uses conventional flat passive heat sinks, not the big curved orange monsters in the picture). http://i19.photobucket.com/albums/b165/TeamScream/Wide-2.jpg Notice the 50% overlap in the heat sinks in the direction of the airflow? That is the overlap with more conventional heat sinks too. Yes, it really does feed hot air from CPU1 into CPU2. Putting a little wall in, redirecting the hot air from CPU1 around CPU2 dropped CPU2's temperature by 4C. Nothing else I tried made a bit of difference - including lowering the "ceiling" over CPU2. CPU2 is still hotter than CPU1 even with that fix. The reason it will not get any better is that while there are 4 fans in the system, they are not placed very well for this motherboard. The first one sends all of its air into the PS and so doesn't cool the CPUs at all. Totally a waste since the PS has a fan already. The next fan is partially blocked by CPU1, so maybe 3/4 of its air is available for CPU2. CPU1 then gets the remaining 1/4 of that fan, and most of the next fan, so around 1 whole fan's worth. The last fan blows over the chipset and PCI slots, again, with no contribution to cooling the CPUS. For comparison, here is a Rio-works HDAMA motherboard which we have. For this design airflow was taken into account. Note that the CPU sockets are spaced farther apart perpendicular to the air flow. It is very similar hardware otherwise: http://www.opteronics.com/images/16a_MBLarge.jpg there are some pictures of these with heat sinks in place which may be found by google image search for "HDAMA motherboard" - I didn't want to link to them as they all seem to be on ebay and those links could disappear at any time. Note how the heatsinks do not overlap in the direction of the air flow? We have one of these, with passive heatsinks of approximately the same shape, but a bit taller, stuffed into an old 2U case scavenged from an old machine. In that machine the two CPUs run at very close to the same temperature. The component layout in the case is very similar to the Supermicro except that the heat sinks are not overlapping, so here there is a fan lined up directly on center with each CPU, plus one to cool the chipset/PCI slots. The PS gets by on its internal fan. The old case has been "optimized" for air flow by the simple expedient of placing the 3 fans as just described (originally there was just one fan in it), plus removing the front panel and as much of the back panel as possible, including the shield that normally goes around the jacks on the motherboard. The HDAMA machine is pretty darn ugly, but it definitely "breathes" better than the Supermicro. I found a product with the perfect properties for sticking polypropylene sheets together. This is 3M "Jet-melt" 3731 hot melt adhesive. (Also called "Scotch-Weld"). Unfortunately I need about 2cc of it, but nobody sells it in sizes less than 11 pounds! The only place that sells anything in this whole 3M hot melt line as single sticks is Digikey, and the one they sell http://parts.digikey.com/1/parts/440266-hot-melt-adhesive-vo-5-8-x2-3748-vo-tc.html is not as heat resistant as the 3731. Probably have to use 3748 though, since at least it can be purchased easily. Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From gus at ldeo.columbia.edu Fri Sep 16 10:54:32 2011 From: gus at ldeo.columbia.edu (Gus Correa) Date: Fri, 16 Sep 2011 10:54:32 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E736328.4040608@ldeo.columbia.edu> mathog wrote: > A not so short story about air flow... > > Yesterday I did some experimenting with different baffles and ducts, > each built temporarily > out of the cardboard backs from yellow notepads and held together with > masking tape. (Not worried > about a fire, since it only ran for 10 minutes at a time like that, and > I was right there to > yank the plug and rip out the cardboard if something went wrong.) The > system has a Supermicro H8DC8 > motherboard in a Supermicro case. This one: > > http://www.supermicro.com/products/chassis/2U/823/SC823S-550LP.cfm > > This is what that motherboard looks like without heat sinks: > > http://www.supermicro.com/a_images/products/Aplus/MB/H8DC8_spec.jpg > > and here is a very similar motherboard with heat sinks in place (but > not my > motherboard, which uses conventional flat passive heat sinks, not the > big curved > orange monsters in the picture). > > http://i19.photobucket.com/albums/b165/TeamScream/Wide-2.jpg > > Notice the 50% overlap in the heat sinks in the direction of the > airflow? > That is the overlap with more conventional heat sinks too. Yes, it > really > does feed hot air from CPU1 into CPU2. Putting a little wall in, > redirecting the hot air from CPU1 around CPU2 dropped CPU2's > temperature > by 4C. Nothing else I tried made a bit of difference - including > lowering > the "ceiling" over CPU2. CPU2 is still hotter than CPU1 even with that > fix. The reason it will not get any better is that while there are 4 > fans > in the system, they are not placed very well for this motherboard. The > first one sends all of its air into the PS and so doesn't cool the CPUs > at all. Totally a waste since the PS has a fan already. The next fan > is partially blocked by CPU1, so maybe 3/4 of its air is available for > CPU2. > CPU1 then gets the remaining 1/4 of that fan, and most of the next fan, > so around 1 whole fan's worth. The last fan blows over the chipset and > PCI slots, again, with no contribution to cooling the CPUS. > > For comparison, here is a Rio-works HDAMA motherboard which we have. > For > this design airflow was taken into account. Note that the CPU sockets > are > spaced farther apart perpendicular to the air flow. It is very similar > hardware otherwise: > > http://www.opteronics.com/images/16a_MBLarge.jpg > > there are some pictures of these with heat sinks in place which may be > found > by google image search for "HDAMA motherboard" - I didn't want to link > to them as they > all seem to be on ebay and those links could disappear at any time. > Note how the > heatsinks do not overlap in the direction of the air flow? We have one > of these, > with passive heatsinks of approximately the same shape, but a bit > taller, > stuffed into an old 2U case scavenged from an old machine. In that > machine > the two CPUs run at very close to the same temperature. The component > layout > in the case is very similar to the Supermicro except that the heat > sinks are not > overlapping, so here there is a fan lined up directly on center with > each CPU, > plus one to cool the chipset/PCI slots. The PS gets by on its internal > fan. > The old case has been "optimized" for air flow by the simple expedient > of placing > the 3 fans as just described (originally there was just one fan in it), > plus > removing the front panel and as much of the back panel as possible, > including > the shield that normally goes around the jacks on the motherboard. > > The HDAMA machine is pretty darn ugly, but it definitely "breathes" > better than > the Supermicro. > > I found a product with the perfect properties for sticking > polypropylene sheets > together. This is 3M "Jet-melt" 3731 hot melt adhesive. (Also called > "Scotch-Weld"). > Unfortunately I need about 2cc of it, but nobody sells it in sizes less > than 11 pounds! > The only place that sells anything in this whole 3M hot melt line as > single sticks is Digikey, > and the one they sell > > http://parts.digikey.com/1/parts/440266-hot-melt-adhesive-vo-5-8-x2-3748-vo-tc.html > > is not as heat resistant as the 3731. Probably have to use 3748 > though, since at least > it can be purchased easily. > > Regards, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf Hi David It may complicate matters too much, but to avoid blowing warm air over CPU2 on the SuperMicro board, is it possible to make two ducts, slightly S-shaped ( __/``` ), one blowing air on the (cpu-1-RAM)+CPU2, the other on CPU1+(cpu-2-RAM)? Or perhaps a single duct, but include an S-shaped wall/divider between the two? Gus Correa _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Fri Sep 16 11:35:09 2011 From: prentice at ias.edu (Prentice Bisbal) Date: Fri, 16 Sep 2011 11:35:09 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E736CAD.6010501@ias.edu> Predicting airflow, especially in the turbulent regime, is not trivial, and not always intuitive. Just ask John Hearns about that. He works for McLaren's F1 team, and I bet they spend A LOT of computational power studying the aerodynamics of their cars (I'm sure the exact amount is a closely guarded secret). Luxi, I'm sure, could pipe in, too. >From my experience, that's one of the factors that sets the large server companies, like Dell, HP, Sun, etc. apart from the smaller companies. These larger companies have the larger resources and the engineering resources to model the airflow through their servers and it's effect on heat transfer. Here's an anectdote about this issue: A certain national lab had incentives to use woman-owned and minority-owned businesses. So when they needed to buy ~200 cluster nodes, they went to a local small business that fit these criteria. I never saw the actual servers myself but sounded like the company bought generic 1U cases, and then put generic/commodity server components in them. I'm sure they didn't to any airflow or thermal analysis on the finished product. After the cluster was brought online, they started having significant hardware reliability problems, and overheating was one of those problems, which could have been the cause of the other reliability problems, since hardware failures usually increase with increased temps. Moral of this story and all this noise? Airflow can be a tricky thing, especially in a challenging conduit, like a 1U or 2U server. -- Prentice On 09/15/2011 06:33 PM, mathog wrote: > A not so short story about air flow... > > Yesterday I did some experimenting with different baffles and ducts, > each built temporarily > out of the cardboard backs from yellow notepads and held together with > masking tape. (Not worried > about a fire, since it only ran for 10 minutes at a time like that, and > I was right there to > yank the plug and rip out the cardboard if something went wrong.) The > system has a Supermicro H8DC8 > motherboard in a Supermicro case. This one: > > http://www.supermicro.com/products/chassis/2U/823/SC823S-550LP.cfm > > This is what that motherboard looks like without heat sinks: > > http://www.supermicro.com/a_images/products/Aplus/MB/H8DC8_spec.jpg > > and here is a very similar motherboard with heat sinks in place (but > not my > motherboard, which uses conventional flat passive heat sinks, not the > big curved > orange monsters in the picture). > > http://i19.photobucket.com/albums/b165/TeamScream/Wide-2.jpg > > Notice the 50% overlap in the heat sinks in the direction of the > airflow? > That is the overlap with more conventional heat sinks too. Yes, it > really > does feed hot air from CPU1 into CPU2. Putting a little wall in, > redirecting the hot air from CPU1 around CPU2 dropped CPU2's > temperature > by 4C. Nothing else I tried made a bit of difference - including > lowering > the "ceiling" over CPU2. CPU2 is still hotter than CPU1 even with that > fix. The reason it will not get any better is that while there are 4 > fans > in the system, they are not placed very well for this motherboard. The > first one sends all of its air into the PS and so doesn't cool the CPUs > at all. Totally a waste since the PS has a fan already. The next fan > is partially blocked by CPU1, so maybe 3/4 of its air is available for > CPU2. > CPU1 then gets the remaining 1/4 of that fan, and most of the next fan, > so around 1 whole fan's worth. The last fan blows over the chipset and > PCI slots, again, with no contribution to cooling the CPUS. > > For comparison, here is a Rio-works HDAMA motherboard which we have. > For > this design airflow was taken into account. Note that the CPU sockets > are > spaced farther apart perpendicular to the air flow. It is very similar > hardware otherwise: > > http://www.opteronics.com/images/16a_MBLarge.jpg > > there are some pictures of these with heat sinks in place which may be > found > by google image search for "HDAMA motherboard" - I didn't want to link > to them as they > all seem to be on ebay and those links could disappear at any time. > Note how the > heatsinks do not overlap in the direction of the air flow? We have one > of these, > with passive heatsinks of approximately the same shape, but a bit > taller, > stuffed into an old 2U case scavenged from an old machine. In that > machine > the two CPUs run at very close to the same temperature. The component > layout > in the case is very similar to the Supermicro except that the heat > sinks are not > overlapping, so here there is a fan lined up directly on center with > each CPU, > plus one to cool the chipset/PCI slots. The PS gets by on its internal > fan. > The old case has been "optimized" for air flow by the simple expedient > of placing > the 3 fans as just described (originally there was just one fan in it), > plus > removing the front panel and as much of the back panel as possible, > including > the shield that normally goes around the jacks on the motherboard. > > The HDAMA machine is pretty darn ugly, but it definitely "breathes" > better than > the Supermicro. > > I found a product with the perfect properties for sticking > polypropylene sheets > together. This is 3M "Jet-melt" 3731 hot melt adhesive. (Also called > "Scotch-Weld"). > Unfortunately I need about 2cc of it, but nobody sells it in sizes less > than 11 pounds! > The only place that sells anything in this whole 3M hot melt line as > single sticks is Digikey, > and the one they sell > > http://parts.digikey.com/1/parts/440266-hot-melt-adhesive-vo-5-8-x2-3748-vo-tc.html > > is not as heat resistant as the 3731. Probably have to use 3748 > though, since at least > it can be purchased easily. > > Regards, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Fri Sep 16 12:39:40 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Fri, 16 Sep 2011 09:39:40 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <4E736CAD.6010501@ias.edu> Message-ID: On 9/16/11 8:35 AM, "Prentice Bisbal" wrote: >Predicting airflow, especially in the turbulent regime, is not trivial, >and not always intuitive. Just ask John Hearns about that. He works for >McLaren's F1 team, and I bet they spend A LOT of computational power >studying the aerodynamics of their cars (I'm sure the exact amount is a >closely guarded secret). > >Luxi, I'm sure, could pipe in, too. Unless you're like John with a F1 budget, or an America's Cup team, or someone designing the next tactical fighter, airflow design in equipment is done by skilled engineers using empirical experience, and a whole lot o' prototyping. Modern IR thermal cameras help, too. I would guess that the vast majority of PC case design is done by old fashion cut and try, and what modern technology has done is make the cut and try faster: nifty ways to do the cutting, better tools to do the trying. Interestingly, since it's so "craft-like", I would think that there is no reason a small integrator with a connection to a decent sheet metal fabricator could crank out customized cases for a particular need: and if the job were big enough to cover the cost of the iterations, you could get a decent product. All it takes is happening to have the right guy or gal who has a good intuitive feel for it and some experience. Think of all those guys and gals doing intake manifold design, valve porting on cylinder heads, etc. in the racing or performance car biz. You find out about them by word of mouth and some people get good reputations for having the "touch". Sort of like when I got my wife a custom 7mm wet suit, I got a couple recommendations for "oh, you want to have Jeannie do it", and subsequently (the suit turned out great) I find out that *everyone* knows that Jeannie is who you go to, because she started aeons ago, working with some other guy, who was the "goto" guy back in the 70s. There are lots of places to get a custom wetsuit in southern California, but relatively few of them have the "touch" to make a suit that is comfortable and usable (and there's some application specificity.. I heard there's a guy up in Santa Barbara who is the Urchin Diving Wet Suit god.. He's probably retired since the urchin business is essentially gone) So the question is "how do you find the gifted case designer who intuitively understands airflow".. Maybe that will be David? (when he gives up the comfortable life at CalTech, and sets up shop next to the surfboard shaper and fashion designers, and hangs out his "Mathog's custom high performance cooling PC cases" shingle) The same thing applies in microwave and RF design. There's all kinds of tools out there to do the analysis, but a lot of good design is just experience and knowing how to put the parts together. So there are these little companies that specialize in some fairly narrow specialty (e.g. Krytar for diode detectors, Marki for mixers, Spacek for amplifiers, Wenzel for oscillators). The troubling thing is that when the one or two people at the company who really know how to do things decide to retire or die, the company will usually be bought up, but will be restricted to making copies of things they've done before. And it takes some years before the word gets out.. "Oh, Bob's no longer there, so don't bother going to them for a custom") However, as Prentice points out in his anecdote, there's an awful lot of "rack n stack" vendors out there who are just plain ignorant. And ignorance is fine, as LONG as the buyer knows they're thermal design ignorant. > >>From my experience, that's one of the factors that sets the large server >companies, like Dell, HP, Sun, etc. apart from the smaller companies. >These larger companies have the larger resources and the engineering >resources to model the airflow through their servers and it's effect on >heat transfer. I don't think they actually model, for the most part. Sure, they'll have some nice animations and some simple models they've run to put on the sales brochures, but I'll bet most of it is empiricism. And, in any case, what you really need is the designer.. Whether they actually build the prototype and test or build it virtually in the modeling code, it's not like there's an automatic "thermal management system designer" program. For instance, when we design radios for spaceflight, we start with some notional design in terms of chassis wall thickness, number and thickness of copper layers in the PWB, etc. You run the model and see, Oh, that's a 40C rise from baseplate to component, and since we have a component temp limit of 85C and the baseplate needs to go to 55C, that's not going to work. Let's try making the copper layers thicker.. Run the model again. Not quite.. Well, what about adding some support bosses to carry the heat out from the middle of the board? What about rearranging the parts on the board to move the hotter parts closer to the edge? Oh no, the PWB designer won't be able to fit the traces in., etc. So it doesn't lend itself to a sort of "parametric" optimization. You might do local parametric optimization: Given this design, what if we push 100cfm of air? What about 200cfm? Etc. > >Here's an anectdote about this issue: > >A certain national lab had incentives to use woman-owned and >minority-owned businesses. So when they needed to buy ~200 cluster >nodes, they went to a local small business that fit these criteria. I >never saw the actual servers myself but sounded like the company bought >generic 1U cases, and then put generic/commodity server components in >them. I'm sure they didn't to any airflow or thermal analysis on the >finished product. I'd assume the same.. However, that doesn't mean that the small shop couldn't have done it, or at least TESTED it. Thermocouples are cheap. That's what I find objectionable... It's really SO easy to do a simple validation and test, but it's like A) the procurement person doesn't know to put some sort of test requirement into the requisition/RFP B) the vendor doesn't volunteer to do it, or know how. How hard is it to go through the mfr data sheet and find the recommended maximum temperatures for the processor, then impose a requirement on the vendor to demonstrate they meet it. (well, actually, it *is* hard for the casual purchaser.. That's why the successful cluster integrators earn their money.. They've either done the work or they've at least done the tests and used that in their selection process) > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Fri Sep 16 13:52:04 2011 From: mathog at caltech.edu (mathog) Date: Fri, 16 Sep 2011 10:52:04 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: References: Message-ID: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> Now that I have seen the problem for two sockets, it made me wonder what the manufacturers did for 4 socket motherboard designs. Easy enough to find out. A Google Image search for: "quad socket" motherboard shows: some took air flow into account, and some didn't. There are quad socket motherboards with the sockets in rectangular arrangements, so that air cooling all of those equally in a 1U or 2U case is going to be really hard, and probably it just isn't done. Water cooling would work, or in a 3U or 4U air cooling by snaking in ducts, but either way it would be a PITA. Then there are other motherboards where air flow was clearly taken into account, and the sockets are lined up so that they are not in each other's air flow. Or at least not much. This one there might be a tiny bit of overlap right at the edge of the heat sinks: http://www.tyan.com/product_SKU_spec.aspx?ProductType=MB&pid=670&SKU=600000180 That would be easy enough to handle by redirecting the hot air a few millimeters to the side with a small plastic "wall". In other designs they line the sockets up straight across the board, so that there is no way for one CPU to "breathe" the hot air from another: http://www.zdnet.co.uk/reviews/sme-servers/2010/06/25/dell-poweredge-r810-40089356/ That Poweredge case looked pretty long, so I checked the specs and it was 72 cm, which was 7 cm longer than the longest 2U case in any of my racks. (And only 2 cm longer than some old dead dual motherboard 1U cases from Racksaver that have not yet gone to recycling. Yes, those really had two motherboards in them, one behind the other.) Being Dell they may have gone with a nonstandard motherboard form factor. Standards are for the little guys! At the other end of the spectrum, in the "what were they thinking?" category, there are 2 socket designs like this one (where the manufacturer only recommends a 4U case, with active heatsinks): http://www.supermicro.com/Aplus/motherboard/Opteron4100/SR56x0/H8DCL-6.cfm This definitely points up one advantage of more cores/socket - it is easier to cool all the cores equally if there are fewer sockets. There is also the blade approach, but that is a whole different ballgame in terms of price versus the more generic rack servers. Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Fri Sep 16 14:19:18 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Fri, 16 Sep 2011 11:19:18 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> Message-ID: On 9/16/11 10:52 AM, "mathog" wrote: >Now that I have seen the problem for two sockets, it made me wonder >what >the manufacturers did for 4 socket motherboard designs. Easy enough to >find out. >A Google Image search for: > > "quad socket" motherboard > >shows: some took air flow into account, and some didn't. > >There are quad socket motherboards with the sockets in rectangular >arrangements, so that >air cooling all of those equally in a 1U or 2U case is going to be >really hard, and probably >it just isn't done. Water cooling would work, or in a 3U or 4U air >cooling by snaking in >ducts, but either way it would be a PITA. > >Then there are other motherboards where air flow was clearly taken into >account, and the sockets >are lined up so that they are not in each other's air flow. Or at >least not much. This one >there might be a tiny bit of overlap right at the edge of the heat >sinks: Having done PWB layout in the past, I can shed some light on this.. First off, doing a design for something this complex is really hard. Not only is there the whole thing of just getting all the wires to fit, but there's a whole tradeoff of how many layers vs manufacturability (e.g. you could make a 20 layer board, but that means you'd better have all 20 of those layers aligned really well so that those tiny vias going from layer 10 to layer 17 through layers 11,12,13,14,15,and 16 actually work. On top of this is all the problems with what's called signal integrity, that is, do all the signals get to the correct place at the right time without interference. So there's all these constraints on relative trace lengths, which layers they can be on, and so forth. And don't forget thermal management. Quite a lot of heat is carried out through the pins of the parts (after all, they have a direct physical connection to the die), and that heat spreads through the board, carried by (usually) the power planes, which are fairly solid sheets of copper. (you can put dedicated heat spreader layers in, but that runs up the cost) So what most people do is start with some previous design that is known to work (e.g. Maybe you have a single socket or dual socket board you've been making for a while) and you sort of cut and paste it. You drag the components around, letting the tool try to reroute, and you fix up the 5% of the traces that just don't autoroute by hand. So you can see why some boards truly are horrible and others aren't. Sometimes, the original 1 or 2 CPU layout lends itself to replication and other times it doesn't. I would think that if you look at the layout of previous boards from the same mfr, you'll see the commonality. And since the original design probably wasn't done contemplating tiling it, it's sort of the luck of the draw whether it's suitable or not. The CPU & bridge mfrs, by the way, generally provide a "reference design" for their parts (in a form suitable for ingest into the CAD tool of your choice), so that people can get up and manufacturing quickly, but that reference design isn't necessarily optimized. And it almost always needs to be tweaked for your particular board house's processes. Every manufacturer of boards has slightly different design rules and manufacturing tolerances. It would be the rare reference design where you could just generate Gerber files with no changes and go get your board fabricated and have it actually work. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Fri Sep 16 15:26:09 2011 From: prentice at ias.edu (Prentice Bisbal) Date: Fri, 16 Sep 2011 15:26:09 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E73A2D1.5070807@ias.edu> On 09/16/2011 12:39 PM, Lux, Jim (337C) wrote: > The same thing applies in microwave and RF design. There's all kinds of > tools out there to do the analysis, but a lot of good design is just > experience and knowing how to put the parts together. So there are these > little companies that specialize in some fairly narrow specialty (e.g. > Krytar for diode detectors, Marki for mixers, Spacek for amplifiers, > Wenzel for oscillators). The troubling thing is that when the one or two > people at the company who really know how to do things decide to retire or > die, the company will usually be bought up, but will be restricted to > making copies of things they've done before. And it takes some years > before the word gets out.. "Oh, Bob's no longer there, so don't bother > going to them for a custom" That process many not be modelling, but it's still engineering, in my book. It's trial and error, but engineers are making the decisions about the next trials based on previous results, and their engineering training/experience, right? I'm sure all the top vendors do this sort of thing with their servers to test the limits of them. > > However, as Prentice points out in his anecdote, there's an awful lot of > "rack n stack" vendors out there who are just plain ignorant. And > ignorance is fine, as LONG as the buyer knows they're thermal design > ignorant. > > And few things can detect bad thermal design as quickly a few racks of cluster nodes running pinned for days! I always enjoyed my classes in heat and momentum transport in college. Too bad I do sysadmin and not science these days. I hope Dave shares his final design and experimental data with us when he's done. -- Prentice _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From atp at piskorski.com Fri Sep 16 20:35:40 2011 From: atp at piskorski.com (Andrew Piskorski) Date: Fri, 16 Sep 2011 20:35:40 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> References: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> Message-ID: <20110917003536.GB86335@piskorski.com> On Fri, Sep 16, 2011 at 10:52:04AM -0700, mathog wrote: > http://www.tyan.com/product_SKU_spec.aspx?ProductType=MB&pid=670&SKU=600000180 > > That would be easy enough to handle by redirecting the hot air > a few millimeters to the side with a small plastic "wall". That makes me wonder why they didn't put all 4 sockets in a row. Then you could have just put one giant heat sink across all 4; I don't see any capacitors or such sticking up in the way. -- Andrew Piskorski http://www.piskorski.com/ _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From lindahl at pbm.com Fri Sep 16 22:22:14 2011 From: lindahl at pbm.com (Greg Lindahl) Date: Fri, 16 Sep 2011 19:22:14 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <20110917003536.GB86335@piskorski.com> References: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> <20110917003536.GB86335@piskorski.com> Message-ID: <20110917022214.GB1215@bx9.net> On Fri, Sep 16, 2011 at 08:35:40PM -0400, Andrew Piskorski wrote: > That makes me wonder why they didn't put all 4 sockets in a row. Then > you could have just put one giant heat sink across all 4; I don't see > any capacitors or such sticking up in the way. It's harder than it looks to get short trace lengths for HT and short trace lengths for the dimm slots while keeping the number of layers to a minimum: 4 used to be the magic number for AMD mobos. Oh, and there's power distribution, and a host of other things to worry about. -- greg _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Sat Sep 17 00:44:52 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Fri, 16 Sep 2011 21:44:52 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <20110917003536.GB86335@piskorski.com> Message-ID: On 9/16/11 5:35 PM, "Andrew Piskorski" wrote: >On Fri, Sep 16, 2011 at 10:52:04AM -0700, mathog wrote: > >> >>http://www.tyan.com/product_SKU_spec.aspx?ProductType=MB&pid=670&SKU=6000 >>00180 >> >> That would be easy enough to handle by redirecting the hot air >> a few millimeters to the side with a small plastic "wall". > >That makes me wonder why they didn't put all 4 sockets in a row. Then >you could have just put one giant heat sink across all 4; I don't see >any capacitors or such sticking up in the way. It's probably a routing of traces thing, not a components in the way thing. > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From trainor at presciencetrust.org Sat Sep 17 21:53:29 2011 From: trainor at presciencetrust.org (Douglas J. Trainor) Date: Sat, 17 Sep 2011 21:53:29 -0400 Subject: [Beowulf] OT: three rotor Enigma machine for sale Message-ID: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> "A version of the three rotor Enigma machine -- used by the German military to encrypt messages, the code of which was subsequently cracked by a team at the legendary Bletchley Park complex -- will be auctioned at Christie's on September 29." -- http://www.cnn.com/2011/WORLD/europe/09/16/enigma.machine.auction/index.html _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rgb at phy.duke.edu Sun Sep 18 08:17:09 2011 From: rgb at phy.duke.edu (Robert G. Brown) Date: Sun, 18 Sep 2011 08:17:09 -0400 (EDT) Subject: [Beowulf] OT: three rotor Enigma machine for sale In-Reply-To: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> References: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> Message-ID: On Sat, 17 Sep 2011, Douglas J. Trainor wrote: > "A version of the three rotor Enigma machine -- used by the German military to encrypt messages, the code of which was subsequently cracked by a team at the legendary Bletchley Park complex -- will be auctioned at Christie's on September 29." -- > > http://www.cnn.com/2011/WORLD/europe/09/16/enigma.machine.auction/index.html Yeah, I saw this and lusted after it in my heart for a bit, but then I though -- I can sell the house, the second house, the boat, the cars, and put in the starting bid in on it and hope all of the actual people with money in the world are home with bad colds on that day, but then where would I keep it if I actually won? It's a bit irregular to make a very comfortable pillow sleeping naked under an overpass... So instead it will go to some wealthy high tech entrepreneur who cannot possibly appreciate it the way I would, or worse, some Eurotrash snob who is "investing" in it as an inflation hedge, and thinks that all of the little keys and wires are very nifty but who is more interested in the provenance of this particular machine -- did it come out of a U-boat, or a diplomatic office, or... I'll just have to settle for modern multirotor pseudorngs. Not as pretty, but a lot faster... rgb > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From samuel at unimelb.edu.au Sun Sep 18 08:37:20 2011 From: samuel at unimelb.edu.au (Chris Samuel) Date: Sun, 18 Sep 2011 22:37:20 +1000 Subject: [Beowulf] OT: three rotor Enigma machine for sale In-Reply-To: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> References: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> Message-ID: <201109182237.20717.samuel@unimelb.edu.au> On Sun, 18 Sep 2011 11:53:29 AM Douglas J. Trainor wrote: [quoting CNN] > subsequently cracked by a team at the legendary Bletchley Park Whilst BP did this during the war, CNN miss completely the fact that if it hadn't been for the successful Polish efforts against Engima in the 1930's (they first read German Enigma traffic in late 1933 according to Peter Clvocoressi, and could decrypt 75% of their intercepts at one point), and their gifts of their reverse-engineered devices (with ABCDE keyboards rather than QWERTY) to the British and French in July 1939, BP might not have been as successful as it was. cheers! Chris -- Christopher Samuel - Senior Systems Administrator VLSCI - Victorian Life Sciences Computation Initiative Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 http://www.vlsci.unimelb.edu.au/ _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Mon Sep 19 11:40:03 2011 From: mathog at caltech.edu (mathog) Date: Mon, 19 Sep 2011 08:40:03 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: <90ED90A6DB6748509A8A0A40B17005B5@ibm> References: <90ED90A6DB6748509A8A0A40B17005B5@ibm> Message-ID: On Sat, 17 Sep 2011 18:53:15 +1000, Richard Windsor wrote: > I'm sorry but I have to have a quiet chuckle here. The ignition > temperature of cardboard at 420 C to 430 C is a significantly higher > that the melting point of solders commonly used in the electronics > industry that run from 364 C to 418 C. Your poor little processor > will > be in a hot bath up to its neck long before there is any smoke. > By the way, my cardboard shrouded processor has been running, > virtually non-stop for over 3 years at about 60 C Spontaneous ignition from heat isn't going to ignite the cardboard. However, a spark is well above the ignition temperature, and sparks are a real possibility on short circuit or component failure in electronic equipment. A small number of sparks are not likely to ignite most plastics, but they may ignite paper products. The difference being that the plastic adjacent to a spark will usually melt, spreading the energy around and keeping the peak temperature below the ignition point. Cardboard/paper have no such mechanism, and so a cellulose fiber in contact with a spark can easily rise above its ignition point. David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rigved.sharma123 at gmail.com Tue Sep 20 09:01:48 2011 From: rigved.sharma123 at gmail.com (rigved sharma) Date: Tue, 20 Sep 2011 18:31:48 +0530 Subject: [Beowulf] nfs error Message-ID: Hi, i have rhel 5.5 x64 machine. I always get following error when nfs clients try to access filesystem from this nfs server. nfsd: find_fh_dentry returned a DISCONNECTED directory: /// Is there any solution to this issue? -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From eugen at leitl.org Wed Sep 21 07:02:39 2011 From: eugen at leitl.org (Eugen Leitl) Date: Wed, 21 Sep 2011 13:02:39 +0200 Subject: [Beowulf] $1, 279-per-hour, 30, 000-core cluster built on Amazon EC2 cloud Message-ID: <20110921110239.GR25711@leitl.org> http://arstechnica.com/business/news/2011/09/30000-core-cluster-built-on-amazon-ec2-cloud.ars $1,279-per-hour, 30,000-core cluster built on Amazon EC2 cloud By Jon Brodkin | Published September 20, 2011 10:49 AM Amazon EC2 and other cloud services are expanding the market for high-performance computing. Without access to a national lab or a supercomputer in your own data center, cloud computing lets businesses spin up temporary clusters at will and stop paying for them as soon as the computing needs are met. A vendor called Cycle Computing is on a mission to demonstrate the potential of Amazon?s cloud by building increasingly large clusters on the Elastic Compute Cloud. Even with Amazon, building a cluster takes some work, but Cycle combines several technologies to ease the process and recently used them to create a 30,000-core cluster running CentOS Linux. The cluster, announced publicly this week, was created for an unnamed ?Top 5 Pharma? customer, and ran for about seven hours at the end of July at a peak cost of $1,279 per hour, including the fees to Amazon and Cycle Computing. The details are impressive: 3,809 compute instances, each with eight cores and 7GB of RAM, for a total of 30,472 cores, 26.7TB of RAM and 2PB (petabytes) of disk space. Security was ensured with HTTPS, SSH and 256-bit AES encryption, and the cluster ran across data centers in three Amazon regions in the United States and Europe. The cluster was dubbed ?Nekomata.? Spreading the cluster across multiple continents was done partly for disaster recovery purposes, and also to guarantee that 30,000 cores could be provisioned. ?We thought it would improve our probability of success if we spread it out,? Cycle Computing?s Dave Powers, manager of product engineering, told Ars. ?Nobody really knows how many instances you can get at any one time from any one [Amazon] region.? Amazon offers its own special cluster compute instances, at a higher cost than regular-sized virtual machines. These cluster instances provide 10 Gigabit Ethernet networking along with greater CPU and memory, but they weren?t necessary to build the Cycle Computing cluster. The pharmaceutical company?s job, related to molecular modeling, was ?embarrassingly parallel? so a fast interconnect wasn?t crucial. To further reduce costs, Cycle took advantage of Amazon?s low-price ?spot instances.? To manage the cluster, Cycle Computing used its own management software as well as the Condor High-Throughput Computing software and Chef, an open source systems integration framework. Cycle demonstrated the power of the Amazon cloud earlier this year with a 10,000-core cluster built for a smaller pharma firm called Genentech. Now, 10,000 cores is a relatively easy task, says Powers. ?We think we?ve mastered the small-scale environments,? he said. 30,000 cores isn?t the end game, either. Going forward, Cycle plans bigger, more complicated clusters, perhaps ones that will require Amazon?s special cluster compute instances. The 30,000-core cluster may or may not be the biggest one run on EC2. Amazon isn?t saying. ?I can?t share specific customer details, but can tell you that we do have businesses of all sizes running large-scale, high-performance computing workloads on AWS [Amazon Web Services], including distributed clusters like the Cycle Computing 30,000 core cluster to tightly-coupled clusters often used for science and engineering applications such as computational fluid dynamics and molecular dynamics simulation,? an Amazon spokesperson told Ars. Amazon itself actually built a supercomputer on its own cloud that made it onto the list of the world?s Top 500 supercomputers. With 7,000 cores, the Amazon cluster ranked number 232 in the world last November with speeds of 41.82 teraflops, falling to number 451 in June of this year. So far, Cycle Computing hasn?t run the Linpack benchmark to determine the speed of its clusters relative to Top 500 sites. But Cycle?s work is impressive no matter how you measure it. The job performed for the unnamed pharma company ?would take well over a week for them to run internally,? Powers says. In the end, the cluster performed the equivalent of 10.9 ?compute years of work.? The task of managing such large cloud-based clusters forced Cycle to step up its own game, with a new plug-in for Chef the company calls Grill. ?There is no way that any mere human could keep track of all of the moving parts on a cluster of this scale,? Cycle wrote in a blog post. ?At Cycle, we?ve always been fans of extreme IT automation, but we needed to take this to the next level in order to monitor and manage every instance, volume, daemon, job, and so on in order for Nekomata to be an efficient 30,000 core tool instead of a big shiny on-demand paperweight.? But problems did arise during the 30,000-core run. ?You can be sure that when you run at massive scale, you are bound to run into some unexpected gotchas,? Cycle notes. ?In our case, one of the gotchas included such things as running out of file descriptors on the license server. In hindsight, we should have anticipated this would be an issue, but we didn?t find that in our prelaunch testing, because we didn?t test at full scale. We were able to quickly recover from this bump and keep moving along with the workload with minimal impact. The license server was able to keep up very nicely with this workload once we increased the number of file descriptors.? Cycle also hit a speed bump related to volume and byte limits on Amazon?s Elastic Block Store volumes. But the company is already planning bigger and better things. ?We already have our next use-case identified and will be turning up the scale a bit more with the next run,? the company says. But ultimately, ?it?s not about core counts or terabytes of RAM or petabytes of data. Rather, it?s about how we are helping to transform how science is done.? _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Wed Sep 21 15:35:15 2011 From: mathog at caltech.edu (mathog) Date: Wed, 21 Sep 2011 12:35:15 -0700 Subject: [Beowulf] SGE forgetting its queues at restart following node rename In-Reply-To: References: Message-ID: <3a0078bd5778e78ec4b45f4fc0310a2b@saf.bio.caltech.edu> FYI (Just to have it posted, in case anybody else ever runs into this.) A little while back I moved same names around in the cluster. To do so, in SGE a bunch of queues and some hosts were removed and then added back. There was much trial and error in doing so - I make no claim that the right commands were issued in the proper order. However, in the end all the queues were as desired and they all stayed up and running. Until the node was rebooted, at which point SGE came back up with only two queues present. After much poking around the problem was finally locate: some of the old host names and old queues were still present in files under: $SGEROOT/default/spool/qmaster/qinstances and as soon as SGE hit one of those during startup, it would stop creating all further queues. The error message that resulted when that happened was of this form: 09/21/2011 12:22:56|qmaster|safserver|E|cannot recreate queue all.q from disk because of unknown host mendel and appeared in: $SGEROOT/default/spool/qmaster/messages Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From john.hearns at mclaren.com Thu Sep 1 04:27:26 2011 From: john.hearns at mclaren.com (Hearns, John) Date: Thu, 1 Sep 2011 09:27:26 +0100 Subject: [Beowulf] materials for air shroud? References: Message-ID: <207BB2F60743C34496BE41039233A809075A6747@MRL-PWEXCHMB02.mil.tagmclarengroup.com> > > Plastic tape covering the aluminum.. 20 mil "pipe wrap" is useful > stuff. 3M VHB double stick foam tape to hold it in place. > > But, enough of this feeble lash-up idea: I think the real solution is > to have a second cluster doing a complete finite element model of the > instantaneous temperature distribution within the processor in > question, You called? John Hearns | CFD Hardware Specialist | McLaren Racing Limited McLaren Technology Centre, Chertsey Road, Woking, Surrey GU21 4YH, UK > driving a set of actuators to form a dynamically optimized > shroud. Sadly, moveable aerodynamic surfaces, with the exception of the DRS, are forbidden by FIA regulations. Regarding the material http://f1-dictionary.110mb.com/heat_shielding.html The contents of this email are confidential and for the exclusive use of the intended recipient. If you receive this email in error you should not copy it, retransmit it, use it or disclose its contents but should return it to the sender immediately and delete your copy. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Thu Sep 1 09:47:14 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Thu, 1 Sep 2011 06:47:14 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <207BB2F60743C34496BE41039233A809075A6747@MRL-PWEXCHMB02.mil.tagmclarengroup.com> Message-ID: On 9/1/11 1:27 AM, "Hearns, John" wrote: >> >> Plastic tape covering the aluminum.. 20 mil "pipe wrap" is useful >> stuff. 3M VHB double stick foam tape to hold it in place. >> >> But, enough of this feeble lash-up idea: I think the real solution is >> to have a second cluster doing a complete finite element model of the >> instantaneous temperature distribution within the processor in >> question, > >You called? > >John Hearns | CFD Hardware Specialist | McLaren Racing Limited >McLaren Technology Centre, Chertsey Road, Woking, Surrey GU21 4YH, UK > > >> driving a set of actuators to form a dynamically optimized >> shroud. > >Sadly, moveable aerodynamic surfaces, with the exception of the DRS, are >forbidden by FIA regulations. > > Oh ho.. You are more bold than I, so we have real-time adaptive cooling shrouds on a cluster, modeled by a cluster, implemented by a cluster installed on a F1 car using the first cluster to optimize performance. Movable aerodynamic surfaces or forbidden, even if they're just inside a component? (yeah, I can see that.. Sure, it's just a cooling fan for the onboard electronics.. All we happened to have is a 1000 HP blower) (I suppose one could use a high rate data link, but where's the thrill in that.. Clusters everywhere..) > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From ntmoore at gmail.com Thu Sep 1 11:22:42 2011 From: ntmoore at gmail.com (Nathan Moore) Date: Thu, 1 Sep 2011 10:22:42 -0500 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: Steel has a lousy thermal conductivity compared to Aluminum. They sell the "dippable" electrical tape at most home improvement stores. Foam double-stick tape is indeed amazing stuff... On Wed, Aug 31, 2011 at 4:05 PM, Robert G. Brown wrote: > On Wed, 31 Aug 2011, Lux, Jim (337C) wrote: > > Also thin aluminum. You can get aluminum sheeting that you can cut with > scissors and that is easy to bend into shapes if you have a bending jig > (or can make one with two pieces of board stock and a vise). Cheap, > fireproof, meltproof at any temperatures you're likely to reach, no > toxic fumes in a fire, can be glued or screwed. The one drawback is > that it is a PITA to weld or solder if that's important to you, but for > an air shroud you can probably make compression joints (interlocking U > rims, squeezed down) that are adequate. > > Most hardware stores (roof flashing), some auto parts or hobby stores. > Copper too, but more expensive. Don't know about thin "enough" sheet > steel, but probably -- copper or steel would both weld or solder easily. > > rgb > > > Cardboard? Card stock? Masking tape? White glue? (that's what I usually > use for cooling ducts.. easy to cut, glue, tape..) It's no more flammable > than plastic, and it doesn't melt and get soft. Papier Mache, works too. > > > > On the other hand, if you want to mold a smooth curve, then plastic is > the way to go. Vacuforming can make a very nice thing, and the form is made > out of wood (usually), but you don't need to go to that extreme.. you get > some nice thermoplastic, put it in hot water to get it soft, and mold as > needed. (yes, you could use those old LPs you've got stashed away.. ) > > > > Thin, cuttable plastic could be polyethylene (not necessarily High > density) or similar. Polystyrene and acrylic tend to be more brittle. ABS > is very nice to work with. PVC is also easy to work with. Nylon is another > possibility. > > > > Do you want to be able to glue it? > > > > What I would do is call up profesionalplastics.com formerly Cadillac > Plastics (many outlets nationwide) and see what they have. It might be more > useful to find a retail outlet and go look through their scrap bin.. Before > Gem-O-Lite in Woodland Hills went out of business, that's where I used to > go. Plastic Depot in Burbank has a huge selection. > > > > Drive over there, and ask the counter folks what would work for you. > $10-20 will get you more plastic than you know what to do with. > > > > Art supply places (e.g. Blick on Raymond.. any of the countless Michaels > or Aaron Bros) also carry sheet plastic, but I find the plastic places tend > to have more variety, and more practical information about use for > "engineering" applications. > > > > > > Jim Lux > > +1(818)354-2075 > > > >> -----Original Message----- > >> From: beowulf-bounces at beowulf.org [mailto:beowulf-bounces at beowulf.org] > On Behalf Of David Mathog > >> Sent: Wednesday, August 31, 2011 10:29 AM > >> To: beowulf at beowulf.org > >> Subject: [Beowulf] materials for air shroud? > >> > >> Anybody know of a nice cheap, high melting point, easy to work with > >> sheet material, for making a custom air shroud? > >> > >> We have one box with stuff in it that looks similar to HDPE, the > >> material the white flexible cutting boards are made of, but it is a bit > >> thinner and more rigid that that. Unfortunately there are no markings > >> on it, so HDPE is just a guess. Whatever it is, it cut easily with > >> scissors (I had to trim it slightly at one point.) > >> > >> Background. We have an older Supermicro SC-823 server with dual > >> processors. The air shroud it came with only covers the first > >> processor. That didn't matter much when it had two low power processors > >> in it, but after upgrading it to dual Opteron 280s, the uncovered second > >> one runs considerably hotter than the covered front one. (Swapping the > >> processors around didn't help - the heat stayed where it was, so a > >> ventilation issue, not a processor issue.) Supermicro does make a newer > >> shroud which extends to the back of the case, but the manual (google for > >> "SC-823 air shroud user's guide") indicates that it is designed for > >> Intel CPUs. So it may or may not fit around the Opterons. > >> > >> The redesigned air shroud will probably work, but I'm about 90% > >> confident that taping a sheet of plastic onto the back of the existing > >> shroud would work as well - if I can find a plastic that won't flap > >> around or melt. > >> > >> Thanks, > >> > >> David Mathog > >> mathog at caltech.edu > >> Manager, Sequence Analysis Facility, Biology Division, Caltech > >> _______________________________________________ > >> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin > Computing > >> To change your subscription (digest mode or unsubscribe) visit > >> http://www.beowulf.org/mailman/listinfo/beowulf > > _______________________________________________ > > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > > > Robert G. Brown http://www.phy.duke.edu/~rgb/ > Duke University Dept. of Physics, Box 90305 > Durham, N.C. 27708-0305 > Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu > > > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > -- - - - - - - - - - - - - - - - - - - - - - Nathan Moore Associate Professor, Physics Winona State University - - - - - - - - - - - - - - - - - - - - - -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From mathog at caltech.edu Wed Sep 7 19:05:51 2011 From: mathog at caltech.edu (mathog) Date: Wed, 07 Sep 2011 16:05:51 -0700 Subject: [Beowulf] =?utf-8?q?SGE_qconf=2C_renaming_hosts/queues=3F?= Message-ID: Just finished migrating most everything on the cluster from the old headnode to a new one, but the old headnode has to remain functional for a while. Even the node name moved, and therein lies the rub. In testing I had node queue_names mendel mendel_something newsaf newsaf_something and that worked fine. Now I want to rename node mendel to oldsaf, and all the queue names to oldsaf_something. Currently the mendel_* queues are all jammed, because by IP number mendel is the same machine as newsaf, and it isn't even the same architecture. I must be missing something, because while I see qconf/qmon options to modify nodes and queues, I don't see how to rename these. Failing that, is there an easy way to dump the host/queue description to a text file, edit that, and then re-create from the modified descriptions? There are a bunch of text configuration files in the SGE directory, but I'm guessing that editing/renaming those directly is a big no no. Thanks, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From kilian.cavalotti.work at gmail.com Thu Sep 8 03:04:14 2011 From: kilian.cavalotti.work at gmail.com (Kilian CAVALOTTI) Date: Thu, 8 Sep 2011 09:04:14 +0200 Subject: [Beowulf] SGE qconf, renaming hosts/queues? In-Reply-To: References: Message-ID: Hi David, On Thu, Sep 8, 2011 at 1:05 AM, mathog wrote: > Failing that, is there an easy way to dump > the host/queue description to a text file, edit > that, and then re-create from the modified descriptions? ?There are a > bunch of text configuration files in the > SGE directory, but I'm ?guessing that editing/renaming those directly > is a big no no. Well, it'd probably work, but the risk is to forget some. Instead, you can use qconf to dump queue definitions. Something along the lines of: # qconf -sql | while read QUEUE ; do echo "****************** ${QUEUE} **********************" qconf -sq ${QUEUE} done For host renaming, though, the safest route is probably to: 0. remove the old queues (qconf -dq ) 1. remove old nodes from SGE configuration (qconf -de for execution hosts) 2. change hostnames on the old nodes (depends on your OS) 3. re-add the new old nodes (qconf -ae for execution hosts) 4. recreate the queues from the previous dump (qconf -aq , and paste/edit the dumps) Cheers, -- Kilian _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From reuti at staff.uni-marburg.de Thu Sep 8 05:40:29 2011 From: reuti at staff.uni-marburg.de (Reuti) Date: Thu, 8 Sep 2011 11:40:29 +0200 Subject: [Beowulf] SGE qconf, renaming hosts/queues? In-Reply-To: References: Message-ID: Hi, Am 08.09.2011 um 09:04 schrieb Kilian CAVALOTTI: > Hi David, > > On Thu, Sep 8, 2011 at 1:05 AM, mathog wrote: >> Failing that, is there an easy way to dump >> the host/queue description to a text file, edit >> that, and then re-create from the modified descriptions? There are a >> bunch of text configuration files in the >> SGE directory, but I'm guessing that editing/renaming those directly >> is a big no no. > > Well, it'd probably work, but the risk is to forget some. > > Instead, you can use qconf to dump queue definitions. Something along > the lines of: yep, there are also prepared scripts in $SGE_ROOT/util/upgrade_modules/ called "save_sge_config.sh" and "load_sge_config.sh" . While the "save_sge_config.sh" will do the correct thing, the "load_sge_config.sh" is meant to be used in an empty SGE installation. It might work in some cases even by overriding the previous setting, but if you have e.g. mutal settings in two queues for subordination of the other queue, you first have to remove this subordination in one queue before you can remove the other queue at all. And the "load_sge_config.sh" isn't so sophisticated to cover these cases. -- Reuti > # qconf -sql | while read QUEUE ; do > echo "****************** ${QUEUE} **********************" > qconf -sq ${QUEUE} > done > > For host renaming, though, the safest route is probably to: > > 0. remove the old queues (qconf -dq ) > 1. remove old nodes from SGE configuration (qconf -de > for execution hosts) > 2. change hostnames on the old nodes (depends on your OS) > 3. re-add the new old nodes (qconf -ae for execution hosts) > 4. recreate the queues from the previous dump (qconf -aq , and > paste/edit the dumps) > > Cheers, > -- > Kilian > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From reuti at staff.uni-marburg.de Thu Sep 8 05:44:54 2011 From: reuti at staff.uni-marburg.de (Reuti) Date: Thu, 8 Sep 2011 11:44:54 +0200 Subject: [Beowulf] SGE qconf, renaming hosts/queues? In-Reply-To: References: Message-ID: Am 08.09.2011 um 01:05 schrieb mathog: > Just finished migrating most everything on the cluster from the old > headnode to a new one, but > the old headnode has to remain functional for a while. Even the node > name moved, and > therein lies the rub. In testing I had > > node queue_names > mendel mendel_something > newsaf newsaf_something > > and that worked fine. Now I want to rename node mendel to oldsaf, and > all the queue names This is not forseen. You can only copy/paste the setting to a new queue and then delete the original one afterwards. Do you have one queue per node? For changing hostnames it's best to remove the nodes first from SGE and then add them again with the new name. -- Reuti > to oldsaf_something. Currently the mendel_* queues are all jammed, > because by IP number mendel > is the same machine as newsaf, and it isn't even the same architecture. > > I must be missing something, because while I see qconf/qmon options to > modify nodes and queues, I don't > see how to rename these. Failing that, is there an easy way to dump > the host/queue description to a text file, edit > that, and then re-create from the modified descriptions? There are a > bunch of text configuration files in the > SGE directory, but I'm guessing that editing/renaming those directly > is a big no no. > > Thanks, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Thu Sep 8 15:03:02 2011 From: mathog at caltech.edu (mathog) Date: Thu, 08 Sep 2011 12:03:02 -0700 Subject: [Beowulf] =?utf-8?q?SGE_qconf=2C_renaming_hosts/queues=3F?= In-Reply-To: References: Message-ID: <7a37f12f8806672162ccb4f3adaff114@saf.bio.caltech.edu> On Thu, 8 Sep 2011 09:04:14 +0200, Kilian CAVALOTTI wrote: > Hi David, > Kilian CAVALOTTI wrote: > On Thu, Sep 8, 2011 at 1:05 AM, mathog wrote: >> Failing that, is there an easy way to dump >> the host/queue description to a text file, edit >> that, and then re-create from the modified descriptions? ?There are >> a >> bunch of text configuration files in the >> SGE directory, but I'm ?guessing that editing/renaming those >> directly >> is a big no no. > > Well, it'd probably work, but the risk is to forget some. > > Instead, you can use qconf to dump queue definitions. Something along > the lines of: > > # qconf -sql | while read QUEUE ; do > echo "****************** ${QUEUE} **********************" > qconf -sq ${QUEUE} > done That was a big help. After dumping both the queues and execution hosts this way it was easy to delete the bogus queues and relatively easy to delete the one bogus host entry (after running down everything else that referenced it). (Bogus host because the address now points to a machine of a different type.) The only major complication was that this doesn't quite work: qconf -se hostname >foo qconf -Ae foo there are things -se writes that -Ae won't consume. After editing out the "processors" and "load_values" lines in foo -Ae was happy. Thanks, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Tue Sep 13 16:05:30 2011 From: mathog at caltech.edu (mathog) Date: Tue, 13 Sep 2011 13:05:30 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: References: Message-ID: FYI, Some sheets of 1/16" polypropylene were obtained from smallparts.com. (This is very similar or identical to the case plastic used on DLT and other larger tape cartridges.) So far this material is relatively easy to work with. Plus: 1. Can be cut with a (big) pair of scissors. 2. Can be bent by hand 3. Melting point >130C 4. Can be welded with high temperature hot glue (not verified) 5. Rigid enough for a duct (but not for anything load bearing) Minus: 1. Holds a static charge. (Probably not more than any other plastic though.) 2. Few adhesives stick to it. Generic brown masking tape holds pretty well. Some details I tried bending a sample two ways. First, just putting it on the edge of a desk and then folding it. That worked but it turned white and thinned somewhat at the bend, and it slowly opened back up again from 90 to about 100 degrees. That edge was then melted by gently rubbing it with the barrel of a soldering iron. (Find the point just below where it smokes). After remelting the edge was once again clear and the angle stayed firmly at 90. Another 90 degree bend was made by first heating the flat plastic on both sides with the soldering iron barrel and then bending. That edge turned out a little better, the thickness of the plastic through the turn was very close to that of the flat parts. Hard to get it to just the right temperature though, so there was some smoke. Either method made a good enough 90 degree bend for an air shroud. Finally, I tried gluing two pieces at right angles using a high melt hot glue. The hot glue gun claims to run at 395F, and the glue stick was nothing special, just generic high temperature hot melt. Mixed results. After it cooled and was allowed to set overnight I tried to tear the two pieces apart by hand, pulling in opposite directions, and they held together. However, I was able to snap the pieces apart by folding it at the junction. (Applying quite a lot of torque to the junction.) The glue completely let go of the top of the "T", all of it stayed on the vertical part. At this point it was easy to peel the rest of the glue off. Seems like the bonding was good perpendicular to the surface, but pretty weak parallel to it. If the piece wasn't physically abused it would likely hold together in an air shroud. I had read somewhere that the hot glue melts the polypropylene so that it was effectively a weld, but that is not how it turned out with this glue. Neither piece of plastic was distorted where the glue had been, so clearly not melted. There are specialty hot melts made of polyethylene or polypropylene, and those may actually weld this material. Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Tue Sep 13 19:51:23 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Tue, 13 Sep 2011 16:51:23 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: > -----Original Message----- > From: beowulf-bounces at beowulf.org [mailto:beowulf-bounces at beowulf.org] On Behalf Of mathog > Sent: Tuesday, September 13, 2011 1:06 PM > To: beowulf at beowulf.org > Subject: Re: [Beowulf] materials for air shroud? > > FYI, > > Some sheets of 1/16" polypropylene were obtained from smallparts.com. > (This is very similar or identical > to the case plastic used on DLT and other larger tape cartridges.) So > far this material is > relatively easy to work with. > > I tried bending a sample two ways. First, just putting it on the edge of a desk and then > folding it. That worked but it turned white and thinned somewhat at the bend, and it > slowly opened back up again from 90 to about 100 degrees. That edge was then melted by > gently rubbing it with the barrel of a soldering iron. (Find the point ust below where > it smokes). A hot air gun works nicely for this. A couple pieces of wood laid on TOP of the plastic where you are blowing the hot air forms a nice channel so that the width of the heated strip is uniform. After you get it soft, you pick it up and put it on your form (i.e. the desk edge). A 1 cm gap works nicely. > Neither piece of plastic > was distorted where the glue had been, so clearly not melted. There > are specialty hot melts made > of polyethylene or polypropylene, and those may actually weld this > material. Plastic welding uses what looks like a big soldering iron with an air hose and a hollow tip. The tip usually has a sort of flat spoon sticking out, maybe 1 cm long and 0.3-0.5 cm wide that you can use to push the melted/soft plastic around. You hold the welder in one hand and the rod in the other and it's much like regular welding (except easier). You get the parent material really soft, and then feed the rod in to make a fillet, just like in regular welding. The trick is welding, not brazing (i.e. the parent material has to melt enough to mix with the melted rod.. otherwise you're basically hot gluing two pieces of plastic together) Jim Lux +1(818)354-2075 _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Wed Sep 14 09:02:21 2011 From: prentice at ias.edu (Prentice Bisbal) Date: Wed, 14 Sep 2011 09:02:21 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E70A5DD.6070701@ias.edu> Instead of a soldering iron, have you tried using just an old clothing iron? That should have a more adjustable temperature setting. And once it's heated up, you can fold it like origami paper. (I challenge this list to produce an origami crane this way!) I would keep a large bowl of cold water nearby, so that after you make your folds, you can dunk your finished product in the cold water to thermoset the plastic. For cutting the plastic, I highly recommend these Wiss MultiMaster multi-purpose shears: http://www.cooperhandtools.com/brands/CF_Files/model_detail.cfm?upc=037103126481 I bought them to cut up a rug, thinking I'd never use them again. They are now probably one of the most used tools in my toolbox. They cut through everything: sheetmetal, hardware cloth, zip-ties, obnoxious plastic packaging - you name it! Since they have a compound action, you don't need to use a lot of force, which makes it easier to control your cuts. I think I paid $12-$15 at Lowe's for mine, and they're worth every penny. On 09/13/2011 04:05 PM, mathog wrote: > FYI, > > Some sheets of 1/16" polypropylene were obtained from smallparts.com. > (This is very similar or identical > to the case plastic used on DLT and other larger tape cartridges.) So > far this material is > relatively easy to work with. > > Plus: > 1. Can be cut with a (big) pair of scissors. > 2. Can be bent by hand > 3. Melting point >130C > 4. Can be welded with high temperature hot glue (not verified) > 5. Rigid enough for a duct (but not for anything load bearing) > > Minus: > 1. Holds a static charge. (Probably not more than any other plastic > though.) > 2. Few adhesives stick to it. Generic brown masking tape holds pretty > well. > > Some details > > I tried bending a sample two ways. First, just putting it on the edge > of a desk and then > folding it. That worked but it turned white and thinned somewhat at > the bend, and it > slowly opened back up again from 90 to about 100 degrees. That edge was > then melted by > gently rubbing it with the barrel of a soldering iron. (Find the point > just below where > it smokes). After remelting the edge was once again clear and the > angle stayed firmly at 90. > Another 90 degree bend was made by first heating the flat plastic on > both sides with the > soldering iron barrel and then bending. That edge turned out a little > better, the thickness > of the plastic through the turn was very close to that of the flat > parts. Hard to get it to > just the right temperature though, so there was some smoke. Either > method made a good enough > 90 degree bend for an air shroud. > > Finally, I tried gluing two pieces at right angles using a high melt > hot glue. The hot glue gun > claims to run at 395F, and the glue stick was nothing special, just > generic high temperature > hot melt. Mixed results. After it cooled and was allowed to set > overnight I tried to > tear the two pieces apart by hand, pulling in opposite directions, and > they held together. > However, I was able to snap the pieces apart by folding it at the > junction. (Applying quite > a lot of torque to the junction.) The glue completely let go of the > top of the "T", all of it > stayed on the vertical part. At this point it was easy to peel the > rest of the glue off. > Seems like the bonding was good perpendicular to the surface, but > pretty weak > parallel to it. If the piece wasn't physically abused it would likely > hold together in an > air shroud. I had read somewhere that the hot glue melts the > polypropylene so that it was > effectively a weld, but that is not how it turned out with this glue. > Neither piece of plastic > was distorted where the glue had been, so clearly not melted. There > are specialty hot melts made > of polyethylene or polypropylene, and those may actually weld this > material. > > Regards, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From dnlombar at ichips.intel.com Wed Sep 14 11:29:30 2011 From: dnlombar at ichips.intel.com (David N. Lombard) Date: Wed, 14 Sep 2011 08:29:30 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <20110914152930.GA27207@nlxcldnl2.cl.intel.com> On Tue, Sep 13, 2011 at 02:05:30PM -0600, mathog wrote: > FYI, > > Finally, I tried gluing two pieces at right angles using a high melt > hot glue. The hot glue gun > claims to run at 395F, and the glue stick was nothing special, just > generic high temperature > hot melt. Mixed results. After it cooled and was allowed to set > overnight I tried to > tear the two pieces apart by hand, pulling in opposite directions, and > they held together. > However, I was able to snap the pieces apart by folding it at the > junction. (Applying quite > a lot of torque to the junction.) The glue completely let go of the > top of the "T", all of it > stayed on the vertical part. At this point it was easy to peel the > rest of the glue off. > Seems like the bonding was good perpendicular to the surface, but > pretty weak > parallel to it. If the piece wasn't physically abused it would likely > hold together in an > air shroud. I had read somewhere that the hot glue melts the > polypropylene so that it was > effectively a weld, but that is not how it turned out with this glue. > Neither piece of plastic > was distorted where the glue had been, so clearly not melted. There > are specialty hot melts made > of polyethylene or polypropylene, and those may actually weld this > material. This is my expectation with /consumer/ hot glues--they're good for temporary work, but are not a long-term adhesive. Did you try some of the hobby plastic adhesives? Plastruct has a "Plastic Weld" solvent cement. They list ABS, Butyrate, Styrene, Acrylic, and "more". Perhaps "more" includes your propylne. Another brand is "Tenax 7R". The former is MEK and the latter MC; both need careful handling. Finally, you could try the PVC Priming Fluid or even PVC Pipe Adhesives. These are all solvent glues and will weld the materials. -- David N. Lombard, Intel, Irvine, CA I do not speak for Intel Corporation; all comments are strictly my own. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From john.hearns at mclaren.com Wed Sep 14 12:35:58 2011 From: john.hearns at mclaren.com (Hearns, John) Date: Wed, 14 Sep 2011 17:35:58 +0100 Subject: [Beowulf] materials for air shroud? References: <20110914152930.GA27207@nlxcldnl2.cl.intel.com> Message-ID: <207BB2F60743C34496BE41039233A80907CD9C33@MRL-PWEXCHMB02.mil.tagmclarengroup.com> Thinking about this, would vacuforming not be a good option? http://astore.amazon.com/google022-20/detail/B001N0BTVU The contents of this email are confidential and for the exclusive use of the intended recipient. If you receive this email in error you should not copy it, retransmit it, use it or disclose its contents but should return it to the sender immediately and delete your copy. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From dnlombar at ichips.intel.com Wed Sep 14 18:56:24 2011 From: dnlombar at ichips.intel.com (David N. Lombard) Date: Wed, 14 Sep 2011 15:56:24 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <207BB2F60743C34496BE41039233A80907CD9C33@MRL-PWEXCHMB02.mil.tagmclarengroup.com> References: <20110914152930.GA27207@nlxcldnl2.cl.intel.com> <207BB2F60743C34496BE41039233A80907CD9C33@MRL-PWEXCHMB02.mil.tagmclarengroup.com> Message-ID: <20110914225624.GA31941@nlxcldnl2.cl.intel.com> On Wed, Sep 14, 2011 at 10:35:58AM -0600, Hearns, John wrote: > Thinking about this, > would vacuforming not be a good option? Alternately, try "smash molding". Short story: heat the plastic and pull it over a form. Make a clamp fixture that holds the edges of the plastic on all sides. Arrange the mold in a another fixture so it sits above (stands proud of) the fixture. Heat the plastic until plastic deformation starts. then pull it down over the mold, well past the bottom of the mold. -- David N. Lombard, Intel, Irvine, CA I do not speak for Intel Corporation; all comments are strictly my own. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Thu Sep 15 18:33:44 2011 From: mathog at caltech.edu (mathog) Date: Thu, 15 Sep 2011 15:33:44 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: References: Message-ID: A not so short story about air flow... Yesterday I did some experimenting with different baffles and ducts, each built temporarily out of the cardboard backs from yellow notepads and held together with masking tape. (Not worried about a fire, since it only ran for 10 minutes at a time like that, and I was right there to yank the plug and rip out the cardboard if something went wrong.) The system has a Supermicro H8DC8 motherboard in a Supermicro case. This one: http://www.supermicro.com/products/chassis/2U/823/SC823S-550LP.cfm This is what that motherboard looks like without heat sinks: http://www.supermicro.com/a_images/products/Aplus/MB/H8DC8_spec.jpg and here is a very similar motherboard with heat sinks in place (but not my motherboard, which uses conventional flat passive heat sinks, not the big curved orange monsters in the picture). http://i19.photobucket.com/albums/b165/TeamScream/Wide-2.jpg Notice the 50% overlap in the heat sinks in the direction of the airflow? That is the overlap with more conventional heat sinks too. Yes, it really does feed hot air from CPU1 into CPU2. Putting a little wall in, redirecting the hot air from CPU1 around CPU2 dropped CPU2's temperature by 4C. Nothing else I tried made a bit of difference - including lowering the "ceiling" over CPU2. CPU2 is still hotter than CPU1 even with that fix. The reason it will not get any better is that while there are 4 fans in the system, they are not placed very well for this motherboard. The first one sends all of its air into the PS and so doesn't cool the CPUs at all. Totally a waste since the PS has a fan already. The next fan is partially blocked by CPU1, so maybe 3/4 of its air is available for CPU2. CPU1 then gets the remaining 1/4 of that fan, and most of the next fan, so around 1 whole fan's worth. The last fan blows over the chipset and PCI slots, again, with no contribution to cooling the CPUS. For comparison, here is a Rio-works HDAMA motherboard which we have. For this design airflow was taken into account. Note that the CPU sockets are spaced farther apart perpendicular to the air flow. It is very similar hardware otherwise: http://www.opteronics.com/images/16a_MBLarge.jpg there are some pictures of these with heat sinks in place which may be found by google image search for "HDAMA motherboard" - I didn't want to link to them as they all seem to be on ebay and those links could disappear at any time. Note how the heatsinks do not overlap in the direction of the air flow? We have one of these, with passive heatsinks of approximately the same shape, but a bit taller, stuffed into an old 2U case scavenged from an old machine. In that machine the two CPUs run at very close to the same temperature. The component layout in the case is very similar to the Supermicro except that the heat sinks are not overlapping, so here there is a fan lined up directly on center with each CPU, plus one to cool the chipset/PCI slots. The PS gets by on its internal fan. The old case has been "optimized" for air flow by the simple expedient of placing the 3 fans as just described (originally there was just one fan in it), plus removing the front panel and as much of the back panel as possible, including the shield that normally goes around the jacks on the motherboard. The HDAMA machine is pretty darn ugly, but it definitely "breathes" better than the Supermicro. I found a product with the perfect properties for sticking polypropylene sheets together. This is 3M "Jet-melt" 3731 hot melt adhesive. (Also called "Scotch-Weld"). Unfortunately I need about 2cc of it, but nobody sells it in sizes less than 11 pounds! The only place that sells anything in this whole 3M hot melt line as single sticks is Digikey, and the one they sell http://parts.digikey.com/1/parts/440266-hot-melt-adhesive-vo-5-8-x2-3748-vo-tc.html is not as heat resistant as the 3731. Probably have to use 3748 though, since at least it can be purchased easily. Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From gus at ldeo.columbia.edu Fri Sep 16 10:54:32 2011 From: gus at ldeo.columbia.edu (Gus Correa) Date: Fri, 16 Sep 2011 10:54:32 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E736328.4040608@ldeo.columbia.edu> mathog wrote: > A not so short story about air flow... > > Yesterday I did some experimenting with different baffles and ducts, > each built temporarily > out of the cardboard backs from yellow notepads and held together with > masking tape. (Not worried > about a fire, since it only ran for 10 minutes at a time like that, and > I was right there to > yank the plug and rip out the cardboard if something went wrong.) The > system has a Supermicro H8DC8 > motherboard in a Supermicro case. This one: > > http://www.supermicro.com/products/chassis/2U/823/SC823S-550LP.cfm > > This is what that motherboard looks like without heat sinks: > > http://www.supermicro.com/a_images/products/Aplus/MB/H8DC8_spec.jpg > > and here is a very similar motherboard with heat sinks in place (but > not my > motherboard, which uses conventional flat passive heat sinks, not the > big curved > orange monsters in the picture). > > http://i19.photobucket.com/albums/b165/TeamScream/Wide-2.jpg > > Notice the 50% overlap in the heat sinks in the direction of the > airflow? > That is the overlap with more conventional heat sinks too. Yes, it > really > does feed hot air from CPU1 into CPU2. Putting a little wall in, > redirecting the hot air from CPU1 around CPU2 dropped CPU2's > temperature > by 4C. Nothing else I tried made a bit of difference - including > lowering > the "ceiling" over CPU2. CPU2 is still hotter than CPU1 even with that > fix. The reason it will not get any better is that while there are 4 > fans > in the system, they are not placed very well for this motherboard. The > first one sends all of its air into the PS and so doesn't cool the CPUs > at all. Totally a waste since the PS has a fan already. The next fan > is partially blocked by CPU1, so maybe 3/4 of its air is available for > CPU2. > CPU1 then gets the remaining 1/4 of that fan, and most of the next fan, > so around 1 whole fan's worth. The last fan blows over the chipset and > PCI slots, again, with no contribution to cooling the CPUS. > > For comparison, here is a Rio-works HDAMA motherboard which we have. > For > this design airflow was taken into account. Note that the CPU sockets > are > spaced farther apart perpendicular to the air flow. It is very similar > hardware otherwise: > > http://www.opteronics.com/images/16a_MBLarge.jpg > > there are some pictures of these with heat sinks in place which may be > found > by google image search for "HDAMA motherboard" - I didn't want to link > to them as they > all seem to be on ebay and those links could disappear at any time. > Note how the > heatsinks do not overlap in the direction of the air flow? We have one > of these, > with passive heatsinks of approximately the same shape, but a bit > taller, > stuffed into an old 2U case scavenged from an old machine. In that > machine > the two CPUs run at very close to the same temperature. The component > layout > in the case is very similar to the Supermicro except that the heat > sinks are not > overlapping, so here there is a fan lined up directly on center with > each CPU, > plus one to cool the chipset/PCI slots. The PS gets by on its internal > fan. > The old case has been "optimized" for air flow by the simple expedient > of placing > the 3 fans as just described (originally there was just one fan in it), > plus > removing the front panel and as much of the back panel as possible, > including > the shield that normally goes around the jacks on the motherboard. > > The HDAMA machine is pretty darn ugly, but it definitely "breathes" > better than > the Supermicro. > > I found a product with the perfect properties for sticking > polypropylene sheets > together. This is 3M "Jet-melt" 3731 hot melt adhesive. (Also called > "Scotch-Weld"). > Unfortunately I need about 2cc of it, but nobody sells it in sizes less > than 11 pounds! > The only place that sells anything in this whole 3M hot melt line as > single sticks is Digikey, > and the one they sell > > http://parts.digikey.com/1/parts/440266-hot-melt-adhesive-vo-5-8-x2-3748-vo-tc.html > > is not as heat resistant as the 3731. Probably have to use 3748 > though, since at least > it can be purchased easily. > > Regards, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf Hi David It may complicate matters too much, but to avoid blowing warm air over CPU2 on the SuperMicro board, is it possible to make two ducts, slightly S-shaped ( __/``` ), one blowing air on the (cpu-1-RAM)+CPU2, the other on CPU1+(cpu-2-RAM)? Or perhaps a single duct, but include an S-shaped wall/divider between the two? Gus Correa _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Fri Sep 16 11:35:09 2011 From: prentice at ias.edu (Prentice Bisbal) Date: Fri, 16 Sep 2011 11:35:09 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E736CAD.6010501@ias.edu> Predicting airflow, especially in the turbulent regime, is not trivial, and not always intuitive. Just ask John Hearns about that. He works for McLaren's F1 team, and I bet they spend A LOT of computational power studying the aerodynamics of their cars (I'm sure the exact amount is a closely guarded secret). Luxi, I'm sure, could pipe in, too. >From my experience, that's one of the factors that sets the large server companies, like Dell, HP, Sun, etc. apart from the smaller companies. These larger companies have the larger resources and the engineering resources to model the airflow through their servers and it's effect on heat transfer. Here's an anectdote about this issue: A certain national lab had incentives to use woman-owned and minority-owned businesses. So when they needed to buy ~200 cluster nodes, they went to a local small business that fit these criteria. I never saw the actual servers myself but sounded like the company bought generic 1U cases, and then put generic/commodity server components in them. I'm sure they didn't to any airflow or thermal analysis on the finished product. After the cluster was brought online, they started having significant hardware reliability problems, and overheating was one of those problems, which could have been the cause of the other reliability problems, since hardware failures usually increase with increased temps. Moral of this story and all this noise? Airflow can be a tricky thing, especially in a challenging conduit, like a 1U or 2U server. -- Prentice On 09/15/2011 06:33 PM, mathog wrote: > A not so short story about air flow... > > Yesterday I did some experimenting with different baffles and ducts, > each built temporarily > out of the cardboard backs from yellow notepads and held together with > masking tape. (Not worried > about a fire, since it only ran for 10 minutes at a time like that, and > I was right there to > yank the plug and rip out the cardboard if something went wrong.) The > system has a Supermicro H8DC8 > motherboard in a Supermicro case. This one: > > http://www.supermicro.com/products/chassis/2U/823/SC823S-550LP.cfm > > This is what that motherboard looks like without heat sinks: > > http://www.supermicro.com/a_images/products/Aplus/MB/H8DC8_spec.jpg > > and here is a very similar motherboard with heat sinks in place (but > not my > motherboard, which uses conventional flat passive heat sinks, not the > big curved > orange monsters in the picture). > > http://i19.photobucket.com/albums/b165/TeamScream/Wide-2.jpg > > Notice the 50% overlap in the heat sinks in the direction of the > airflow? > That is the overlap with more conventional heat sinks too. Yes, it > really > does feed hot air from CPU1 into CPU2. Putting a little wall in, > redirecting the hot air from CPU1 around CPU2 dropped CPU2's > temperature > by 4C. Nothing else I tried made a bit of difference - including > lowering > the "ceiling" over CPU2. CPU2 is still hotter than CPU1 even with that > fix. The reason it will not get any better is that while there are 4 > fans > in the system, they are not placed very well for this motherboard. The > first one sends all of its air into the PS and so doesn't cool the CPUs > at all. Totally a waste since the PS has a fan already. The next fan > is partially blocked by CPU1, so maybe 3/4 of its air is available for > CPU2. > CPU1 then gets the remaining 1/4 of that fan, and most of the next fan, > so around 1 whole fan's worth. The last fan blows over the chipset and > PCI slots, again, with no contribution to cooling the CPUS. > > For comparison, here is a Rio-works HDAMA motherboard which we have. > For > this design airflow was taken into account. Note that the CPU sockets > are > spaced farther apart perpendicular to the air flow. It is very similar > hardware otherwise: > > http://www.opteronics.com/images/16a_MBLarge.jpg > > there are some pictures of these with heat sinks in place which may be > found > by google image search for "HDAMA motherboard" - I didn't want to link > to them as they > all seem to be on ebay and those links could disappear at any time. > Note how the > heatsinks do not overlap in the direction of the air flow? We have one > of these, > with passive heatsinks of approximately the same shape, but a bit > taller, > stuffed into an old 2U case scavenged from an old machine. In that > machine > the two CPUs run at very close to the same temperature. The component > layout > in the case is very similar to the Supermicro except that the heat > sinks are not > overlapping, so here there is a fan lined up directly on center with > each CPU, > plus one to cool the chipset/PCI slots. The PS gets by on its internal > fan. > The old case has been "optimized" for air flow by the simple expedient > of placing > the 3 fans as just described (originally there was just one fan in it), > plus > removing the front panel and as much of the back panel as possible, > including > the shield that normally goes around the jacks on the motherboard. > > The HDAMA machine is pretty darn ugly, but it definitely "breathes" > better than > the Supermicro. > > I found a product with the perfect properties for sticking > polypropylene sheets > together. This is 3M "Jet-melt" 3731 hot melt adhesive. (Also called > "Scotch-Weld"). > Unfortunately I need about 2cc of it, but nobody sells it in sizes less > than 11 pounds! > The only place that sells anything in this whole 3M hot melt line as > single sticks is Digikey, > and the one they sell > > http://parts.digikey.com/1/parts/440266-hot-melt-adhesive-vo-5-8-x2-3748-vo-tc.html > > is not as heat resistant as the 3731. Probably have to use 3748 > though, since at least > it can be purchased easily. > > Regards, > > David Mathog > mathog at caltech.edu > Manager, Sequence Analysis Facility, Biology Division, Caltech > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Fri Sep 16 12:39:40 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Fri, 16 Sep 2011 09:39:40 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <4E736CAD.6010501@ias.edu> Message-ID: On 9/16/11 8:35 AM, "Prentice Bisbal" wrote: >Predicting airflow, especially in the turbulent regime, is not trivial, >and not always intuitive. Just ask John Hearns about that. He works for >McLaren's F1 team, and I bet they spend A LOT of computational power >studying the aerodynamics of their cars (I'm sure the exact amount is a >closely guarded secret). > >Luxi, I'm sure, could pipe in, too. Unless you're like John with a F1 budget, or an America's Cup team, or someone designing the next tactical fighter, airflow design in equipment is done by skilled engineers using empirical experience, and a whole lot o' prototyping. Modern IR thermal cameras help, too. I would guess that the vast majority of PC case design is done by old fashion cut and try, and what modern technology has done is make the cut and try faster: nifty ways to do the cutting, better tools to do the trying. Interestingly, since it's so "craft-like", I would think that there is no reason a small integrator with a connection to a decent sheet metal fabricator could crank out customized cases for a particular need: and if the job were big enough to cover the cost of the iterations, you could get a decent product. All it takes is happening to have the right guy or gal who has a good intuitive feel for it and some experience. Think of all those guys and gals doing intake manifold design, valve porting on cylinder heads, etc. in the racing or performance car biz. You find out about them by word of mouth and some people get good reputations for having the "touch". Sort of like when I got my wife a custom 7mm wet suit, I got a couple recommendations for "oh, you want to have Jeannie do it", and subsequently (the suit turned out great) I find out that *everyone* knows that Jeannie is who you go to, because she started aeons ago, working with some other guy, who was the "goto" guy back in the 70s. There are lots of places to get a custom wetsuit in southern California, but relatively few of them have the "touch" to make a suit that is comfortable and usable (and there's some application specificity.. I heard there's a guy up in Santa Barbara who is the Urchin Diving Wet Suit god.. He's probably retired since the urchin business is essentially gone) So the question is "how do you find the gifted case designer who intuitively understands airflow".. Maybe that will be David? (when he gives up the comfortable life at CalTech, and sets up shop next to the surfboard shaper and fashion designers, and hangs out his "Mathog's custom high performance cooling PC cases" shingle) The same thing applies in microwave and RF design. There's all kinds of tools out there to do the analysis, but a lot of good design is just experience and knowing how to put the parts together. So there are these little companies that specialize in some fairly narrow specialty (e.g. Krytar for diode detectors, Marki for mixers, Spacek for amplifiers, Wenzel for oscillators). The troubling thing is that when the one or two people at the company who really know how to do things decide to retire or die, the company will usually be bought up, but will be restricted to making copies of things they've done before. And it takes some years before the word gets out.. "Oh, Bob's no longer there, so don't bother going to them for a custom") However, as Prentice points out in his anecdote, there's an awful lot of "rack n stack" vendors out there who are just plain ignorant. And ignorance is fine, as LONG as the buyer knows they're thermal design ignorant. > >>From my experience, that's one of the factors that sets the large server >companies, like Dell, HP, Sun, etc. apart from the smaller companies. >These larger companies have the larger resources and the engineering >resources to model the airflow through their servers and it's effect on >heat transfer. I don't think they actually model, for the most part. Sure, they'll have some nice animations and some simple models they've run to put on the sales brochures, but I'll bet most of it is empiricism. And, in any case, what you really need is the designer.. Whether they actually build the prototype and test or build it virtually in the modeling code, it's not like there's an automatic "thermal management system designer" program. For instance, when we design radios for spaceflight, we start with some notional design in terms of chassis wall thickness, number and thickness of copper layers in the PWB, etc. You run the model and see, Oh, that's a 40C rise from baseplate to component, and since we have a component temp limit of 85C and the baseplate needs to go to 55C, that's not going to work. Let's try making the copper layers thicker.. Run the model again. Not quite.. Well, what about adding some support bosses to carry the heat out from the middle of the board? What about rearranging the parts on the board to move the hotter parts closer to the edge? Oh no, the PWB designer won't be able to fit the traces in., etc. So it doesn't lend itself to a sort of "parametric" optimization. You might do local parametric optimization: Given this design, what if we push 100cfm of air? What about 200cfm? Etc. > >Here's an anectdote about this issue: > >A certain national lab had incentives to use woman-owned and >minority-owned businesses. So when they needed to buy ~200 cluster >nodes, they went to a local small business that fit these criteria. I >never saw the actual servers myself but sounded like the company bought >generic 1U cases, and then put generic/commodity server components in >them. I'm sure they didn't to any airflow or thermal analysis on the >finished product. I'd assume the same.. However, that doesn't mean that the small shop couldn't have done it, or at least TESTED it. Thermocouples are cheap. That's what I find objectionable... It's really SO easy to do a simple validation and test, but it's like A) the procurement person doesn't know to put some sort of test requirement into the requisition/RFP B) the vendor doesn't volunteer to do it, or know how. How hard is it to go through the mfr data sheet and find the recommended maximum temperatures for the processor, then impose a requirement on the vendor to demonstrate they meet it. (well, actually, it *is* hard for the casual purchaser.. That's why the successful cluster integrators earn their money.. They've either done the work or they've at least done the tests and used that in their selection process) > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Fri Sep 16 13:52:04 2011 From: mathog at caltech.edu (mathog) Date: Fri, 16 Sep 2011 10:52:04 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: References: Message-ID: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> Now that I have seen the problem for two sockets, it made me wonder what the manufacturers did for 4 socket motherboard designs. Easy enough to find out. A Google Image search for: "quad socket" motherboard shows: some took air flow into account, and some didn't. There are quad socket motherboards with the sockets in rectangular arrangements, so that air cooling all of those equally in a 1U or 2U case is going to be really hard, and probably it just isn't done. Water cooling would work, or in a 3U or 4U air cooling by snaking in ducts, but either way it would be a PITA. Then there are other motherboards where air flow was clearly taken into account, and the sockets are lined up so that they are not in each other's air flow. Or at least not much. This one there might be a tiny bit of overlap right at the edge of the heat sinks: http://www.tyan.com/product_SKU_spec.aspx?ProductType=MB&pid=670&SKU=600000180 That would be easy enough to handle by redirecting the hot air a few millimeters to the side with a small plastic "wall". In other designs they line the sockets up straight across the board, so that there is no way for one CPU to "breathe" the hot air from another: http://www.zdnet.co.uk/reviews/sme-servers/2010/06/25/dell-poweredge-r810-40089356/ That Poweredge case looked pretty long, so I checked the specs and it was 72 cm, which was 7 cm longer than the longest 2U case in any of my racks. (And only 2 cm longer than some old dead dual motherboard 1U cases from Racksaver that have not yet gone to recycling. Yes, those really had two motherboards in them, one behind the other.) Being Dell they may have gone with a nonstandard motherboard form factor. Standards are for the little guys! At the other end of the spectrum, in the "what were they thinking?" category, there are 2 socket designs like this one (where the manufacturer only recommends a 4U case, with active heatsinks): http://www.supermicro.com/Aplus/motherboard/Opteron4100/SR56x0/H8DCL-6.cfm This definitely points up one advantage of more cores/socket - it is easier to cool all the cores equally if there are fewer sockets. There is also the blade approach, but that is a whole different ballgame in terms of price versus the more generic rack servers. Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Fri Sep 16 14:19:18 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Fri, 16 Sep 2011 11:19:18 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> Message-ID: On 9/16/11 10:52 AM, "mathog" wrote: >Now that I have seen the problem for two sockets, it made me wonder >what >the manufacturers did for 4 socket motherboard designs. Easy enough to >find out. >A Google Image search for: > > "quad socket" motherboard > >shows: some took air flow into account, and some didn't. > >There are quad socket motherboards with the sockets in rectangular >arrangements, so that >air cooling all of those equally in a 1U or 2U case is going to be >really hard, and probably >it just isn't done. Water cooling would work, or in a 3U or 4U air >cooling by snaking in >ducts, but either way it would be a PITA. > >Then there are other motherboards where air flow was clearly taken into >account, and the sockets >are lined up so that they are not in each other's air flow. Or at >least not much. This one >there might be a tiny bit of overlap right at the edge of the heat >sinks: Having done PWB layout in the past, I can shed some light on this.. First off, doing a design for something this complex is really hard. Not only is there the whole thing of just getting all the wires to fit, but there's a whole tradeoff of how many layers vs manufacturability (e.g. you could make a 20 layer board, but that means you'd better have all 20 of those layers aligned really well so that those tiny vias going from layer 10 to layer 17 through layers 11,12,13,14,15,and 16 actually work. On top of this is all the problems with what's called signal integrity, that is, do all the signals get to the correct place at the right time without interference. So there's all these constraints on relative trace lengths, which layers they can be on, and so forth. And don't forget thermal management. Quite a lot of heat is carried out through the pins of the parts (after all, they have a direct physical connection to the die), and that heat spreads through the board, carried by (usually) the power planes, which are fairly solid sheets of copper. (you can put dedicated heat spreader layers in, but that runs up the cost) So what most people do is start with some previous design that is known to work (e.g. Maybe you have a single socket or dual socket board you've been making for a while) and you sort of cut and paste it. You drag the components around, letting the tool try to reroute, and you fix up the 5% of the traces that just don't autoroute by hand. So you can see why some boards truly are horrible and others aren't. Sometimes, the original 1 or 2 CPU layout lends itself to replication and other times it doesn't. I would think that if you look at the layout of previous boards from the same mfr, you'll see the commonality. And since the original design probably wasn't done contemplating tiling it, it's sort of the luck of the draw whether it's suitable or not. The CPU & bridge mfrs, by the way, generally provide a "reference design" for their parts (in a form suitable for ingest into the CAD tool of your choice), so that people can get up and manufacturing quickly, but that reference design isn't necessarily optimized. And it almost always needs to be tweaked for your particular board house's processes. Every manufacturer of boards has slightly different design rules and manufacturing tolerances. It would be the rare reference design where you could just generate Gerber files with no changes and go get your board fabricated and have it actually work. _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From prentice at ias.edu Fri Sep 16 15:26:09 2011 From: prentice at ias.edu (Prentice Bisbal) Date: Fri, 16 Sep 2011 15:26:09 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: References: Message-ID: <4E73A2D1.5070807@ias.edu> On 09/16/2011 12:39 PM, Lux, Jim (337C) wrote: > The same thing applies in microwave and RF design. There's all kinds of > tools out there to do the analysis, but a lot of good design is just > experience and knowing how to put the parts together. So there are these > little companies that specialize in some fairly narrow specialty (e.g. > Krytar for diode detectors, Marki for mixers, Spacek for amplifiers, > Wenzel for oscillators). The troubling thing is that when the one or two > people at the company who really know how to do things decide to retire or > die, the company will usually be bought up, but will be restricted to > making copies of things they've done before. And it takes some years > before the word gets out.. "Oh, Bob's no longer there, so don't bother > going to them for a custom" That process many not be modelling, but it's still engineering, in my book. It's trial and error, but engineers are making the decisions about the next trials based on previous results, and their engineering training/experience, right? I'm sure all the top vendors do this sort of thing with their servers to test the limits of them. > > However, as Prentice points out in his anecdote, there's an awful lot of > "rack n stack" vendors out there who are just plain ignorant. And > ignorance is fine, as LONG as the buyer knows they're thermal design > ignorant. > > And few things can detect bad thermal design as quickly a few racks of cluster nodes running pinned for days! I always enjoyed my classes in heat and momentum transport in college. Too bad I do sysadmin and not science these days. I hope Dave shares his final design and experimental data with us when he's done. -- Prentice _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From atp at piskorski.com Fri Sep 16 20:35:40 2011 From: atp at piskorski.com (Andrew Piskorski) Date: Fri, 16 Sep 2011 20:35:40 -0400 Subject: [Beowulf] materials for air shroud? In-Reply-To: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> References: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> Message-ID: <20110917003536.GB86335@piskorski.com> On Fri, Sep 16, 2011 at 10:52:04AM -0700, mathog wrote: > http://www.tyan.com/product_SKU_spec.aspx?ProductType=MB&pid=670&SKU=600000180 > > That would be easy enough to handle by redirecting the hot air > a few millimeters to the side with a small plastic "wall". That makes me wonder why they didn't put all 4 sockets in a row. Then you could have just put one giant heat sink across all 4; I don't see any capacitors or such sticking up in the way. -- Andrew Piskorski http://www.piskorski.com/ _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From lindahl at pbm.com Fri Sep 16 22:22:14 2011 From: lindahl at pbm.com (Greg Lindahl) Date: Fri, 16 Sep 2011 19:22:14 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <20110917003536.GB86335@piskorski.com> References: <08d6afc30a65cf9b09c30f9c34245629@saf.bio.caltech.edu> <20110917003536.GB86335@piskorski.com> Message-ID: <20110917022214.GB1215@bx9.net> On Fri, Sep 16, 2011 at 08:35:40PM -0400, Andrew Piskorski wrote: > That makes me wonder why they didn't put all 4 sockets in a row. Then > you could have just put one giant heat sink across all 4; I don't see > any capacitors or such sticking up in the way. It's harder than it looks to get short trace lengths for HT and short trace lengths for the dimm slots while keeping the number of layers to a minimum: 4 used to be the magic number for AMD mobos. Oh, and there's power distribution, and a host of other things to worry about. -- greg _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From james.p.lux at jpl.nasa.gov Sat Sep 17 00:44:52 2011 From: james.p.lux at jpl.nasa.gov (Lux, Jim (337C)) Date: Fri, 16 Sep 2011 21:44:52 -0700 Subject: [Beowulf] materials for air shroud? In-Reply-To: <20110917003536.GB86335@piskorski.com> Message-ID: On 9/16/11 5:35 PM, "Andrew Piskorski" wrote: >On Fri, Sep 16, 2011 at 10:52:04AM -0700, mathog wrote: > >> >>http://www.tyan.com/product_SKU_spec.aspx?ProductType=MB&pid=670&SKU=6000 >>00180 >> >> That would be easy enough to handle by redirecting the hot air >> a few millimeters to the side with a small plastic "wall". > >That makes me wonder why they didn't put all 4 sockets in a row. Then >you could have just put one giant heat sink across all 4; I don't see >any capacitors or such sticking up in the way. It's probably a routing of traces thing, not a components in the way thing. > _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From trainor at presciencetrust.org Sat Sep 17 21:53:29 2011 From: trainor at presciencetrust.org (Douglas J. Trainor) Date: Sat, 17 Sep 2011 21:53:29 -0400 Subject: [Beowulf] OT: three rotor Enigma machine for sale Message-ID: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> "A version of the three rotor Enigma machine -- used by the German military to encrypt messages, the code of which was subsequently cracked by a team at the legendary Bletchley Park complex -- will be auctioned at Christie's on September 29." -- http://www.cnn.com/2011/WORLD/europe/09/16/enigma.machine.auction/index.html _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rgb at phy.duke.edu Sun Sep 18 08:17:09 2011 From: rgb at phy.duke.edu (Robert G. Brown) Date: Sun, 18 Sep 2011 08:17:09 -0400 (EDT) Subject: [Beowulf] OT: three rotor Enigma machine for sale In-Reply-To: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> References: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> Message-ID: On Sat, 17 Sep 2011, Douglas J. Trainor wrote: > "A version of the three rotor Enigma machine -- used by the German military to encrypt messages, the code of which was subsequently cracked by a team at the legendary Bletchley Park complex -- will be auctioned at Christie's on September 29." -- > > http://www.cnn.com/2011/WORLD/europe/09/16/enigma.machine.auction/index.html Yeah, I saw this and lusted after it in my heart for a bit, but then I though -- I can sell the house, the second house, the boat, the cars, and put in the starting bid in on it and hope all of the actual people with money in the world are home with bad colds on that day, but then where would I keep it if I actually won? It's a bit irregular to make a very comfortable pillow sleeping naked under an overpass... So instead it will go to some wealthy high tech entrepreneur who cannot possibly appreciate it the way I would, or worse, some Eurotrash snob who is "investing" in it as an inflation hedge, and thinks that all of the little keys and wires are very nifty but who is more interested in the provenance of this particular machine -- did it come out of a U-boat, or a diplomatic office, or... I'll just have to settle for modern multirotor pseudorngs. Not as pretty, but a lot faster... rgb > _______________________________________________ > Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf > Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From samuel at unimelb.edu.au Sun Sep 18 08:37:20 2011 From: samuel at unimelb.edu.au (Chris Samuel) Date: Sun, 18 Sep 2011 22:37:20 +1000 Subject: [Beowulf] OT: three rotor Enigma machine for sale In-Reply-To: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> References: <518F8301-58AF-4DC3-98F3-EE75B055BB4A@presciencetrust.org> Message-ID: <201109182237.20717.samuel@unimelb.edu.au> On Sun, 18 Sep 2011 11:53:29 AM Douglas J. Trainor wrote: [quoting CNN] > subsequently cracked by a team at the legendary Bletchley Park Whilst BP did this during the war, CNN miss completely the fact that if it hadn't been for the successful Polish efforts against Engima in the 1930's (they first read German Enigma traffic in late 1933 according to Peter Clvocoressi, and could decrypt 75% of their intercepts at one point), and their gifts of their reverse-engineered devices (with ABCDE keyboards rather than QWERTY) to the British and French in July 1939, BP might not have been as successful as it was. cheers! Chris -- Christopher Samuel - Senior Systems Administrator VLSCI - Victorian Life Sciences Computation Initiative Email: samuel at unimelb.edu.au Phone: +61 (0)3 903 55545 http://www.vlsci.unimelb.edu.au/ _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Mon Sep 19 11:40:03 2011 From: mathog at caltech.edu (mathog) Date: Mon, 19 Sep 2011 08:40:03 -0700 Subject: [Beowulf] =?utf-8?q?materials_for_air_shroud=3F?= In-Reply-To: <90ED90A6DB6748509A8A0A40B17005B5@ibm> References: <90ED90A6DB6748509A8A0A40B17005B5@ibm> Message-ID: On Sat, 17 Sep 2011 18:53:15 +1000, Richard Windsor wrote: > I'm sorry but I have to have a quiet chuckle here. The ignition > temperature of cardboard at 420 C to 430 C is a significantly higher > that the melting point of solders commonly used in the electronics > industry that run from 364 C to 418 C. Your poor little processor > will > be in a hot bath up to its neck long before there is any smoke. > By the way, my cardboard shrouded processor has been running, > virtually non-stop for over 3 years at about 60 C Spontaneous ignition from heat isn't going to ignite the cardboard. However, a spark is well above the ignition temperature, and sparks are a real possibility on short circuit or component failure in electronic equipment. A small number of sparks are not likely to ignite most plastics, but they may ignite paper products. The difference being that the plastic adjacent to a spark will usually melt, spreading the energy around and keeping the peak temperature below the ignition point. Cardboard/paper have no such mechanism, and so a cellulose fiber in contact with a spark can easily rise above its ignition point. David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From rigved.sharma123 at gmail.com Tue Sep 20 09:01:48 2011 From: rigved.sharma123 at gmail.com (rigved sharma) Date: Tue, 20 Sep 2011 18:31:48 +0530 Subject: [Beowulf] nfs error Message-ID: Hi, i have rhel 5.5 x64 machine. I always get following error when nfs clients try to access filesystem from this nfs server. nfsd: find_fh_dentry returned a DISCONNECTED directory: /// Is there any solution to this issue? -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf From eugen at leitl.org Wed Sep 21 07:02:39 2011 From: eugen at leitl.org (Eugen Leitl) Date: Wed, 21 Sep 2011 13:02:39 +0200 Subject: [Beowulf] $1, 279-per-hour, 30, 000-core cluster built on Amazon EC2 cloud Message-ID: <20110921110239.GR25711@leitl.org> http://arstechnica.com/business/news/2011/09/30000-core-cluster-built-on-amazon-ec2-cloud.ars $1,279-per-hour, 30,000-core cluster built on Amazon EC2 cloud By Jon Brodkin | Published September 20, 2011 10:49 AM Amazon EC2 and other cloud services are expanding the market for high-performance computing. Without access to a national lab or a supercomputer in your own data center, cloud computing lets businesses spin up temporary clusters at will and stop paying for them as soon as the computing needs are met. A vendor called Cycle Computing is on a mission to demonstrate the potential of Amazon?s cloud by building increasingly large clusters on the Elastic Compute Cloud. Even with Amazon, building a cluster takes some work, but Cycle combines several technologies to ease the process and recently used them to create a 30,000-core cluster running CentOS Linux. The cluster, announced publicly this week, was created for an unnamed ?Top 5 Pharma? customer, and ran for about seven hours at the end of July at a peak cost of $1,279 per hour, including the fees to Amazon and Cycle Computing. The details are impressive: 3,809 compute instances, each with eight cores and 7GB of RAM, for a total of 30,472 cores, 26.7TB of RAM and 2PB (petabytes) of disk space. Security was ensured with HTTPS, SSH and 256-bit AES encryption, and the cluster ran across data centers in three Amazon regions in the United States and Europe. The cluster was dubbed ?Nekomata.? Spreading the cluster across multiple continents was done partly for disaster recovery purposes, and also to guarantee that 30,000 cores could be provisioned. ?We thought it would improve our probability of success if we spread it out,? Cycle Computing?s Dave Powers, manager of product engineering, told Ars. ?Nobody really knows how many instances you can get at any one time from any one [Amazon] region.? Amazon offers its own special cluster compute instances, at a higher cost than regular-sized virtual machines. These cluster instances provide 10 Gigabit Ethernet networking along with greater CPU and memory, but they weren?t necessary to build the Cycle Computing cluster. The pharmaceutical company?s job, related to molecular modeling, was ?embarrassingly parallel? so a fast interconnect wasn?t crucial. To further reduce costs, Cycle took advantage of Amazon?s low-price ?spot instances.? To manage the cluster, Cycle Computing used its own management software as well as the Condor High-Throughput Computing software and Chef, an open source systems integration framework. Cycle demonstrated the power of the Amazon cloud earlier this year with a 10,000-core cluster built for a smaller pharma firm called Genentech. Now, 10,000 cores is a relatively easy task, says Powers. ?We think we?ve mastered the small-scale environments,? he said. 30,000 cores isn?t the end game, either. Going forward, Cycle plans bigger, more complicated clusters, perhaps ones that will require Amazon?s special cluster compute instances. The 30,000-core cluster may or may not be the biggest one run on EC2. Amazon isn?t saying. ?I can?t share specific customer details, but can tell you that we do have businesses of all sizes running large-scale, high-performance computing workloads on AWS [Amazon Web Services], including distributed clusters like the Cycle Computing 30,000 core cluster to tightly-coupled clusters often used for science and engineering applications such as computational fluid dynamics and molecular dynamics simulation,? an Amazon spokesperson told Ars. Amazon itself actually built a supercomputer on its own cloud that made it onto the list of the world?s Top 500 supercomputers. With 7,000 cores, the Amazon cluster ranked number 232 in the world last November with speeds of 41.82 teraflops, falling to number 451 in June of this year. So far, Cycle Computing hasn?t run the Linpack benchmark to determine the speed of its clusters relative to Top 500 sites. But Cycle?s work is impressive no matter how you measure it. The job performed for the unnamed pharma company ?would take well over a week for them to run internally,? Powers says. In the end, the cluster performed the equivalent of 10.9 ?compute years of work.? The task of managing such large cloud-based clusters forced Cycle to step up its own game, with a new plug-in for Chef the company calls Grill. ?There is no way that any mere human could keep track of all of the moving parts on a cluster of this scale,? Cycle wrote in a blog post. ?At Cycle, we?ve always been fans of extreme IT automation, but we needed to take this to the next level in order to monitor and manage every instance, volume, daemon, job, and so on in order for Nekomata to be an efficient 30,000 core tool instead of a big shiny on-demand paperweight.? But problems did arise during the 30,000-core run. ?You can be sure that when you run at massive scale, you are bound to run into some unexpected gotchas,? Cycle notes. ?In our case, one of the gotchas included such things as running out of file descriptors on the license server. In hindsight, we should have anticipated this would be an issue, but we didn?t find that in our prelaunch testing, because we didn?t test at full scale. We were able to quickly recover from this bump and keep moving along with the workload with minimal impact. The license server was able to keep up very nicely with this workload once we increased the number of file descriptors.? Cycle also hit a speed bump related to volume and byte limits on Amazon?s Elastic Block Store volumes. But the company is already planning bigger and better things. ?We already have our next use-case identified and will be turning up the scale a bit more with the next run,? the company says. But ultimately, ?it?s not about core counts or terabytes of RAM or petabytes of data. Rather, it?s about how we are helping to transform how science is done.? _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mathog at caltech.edu Wed Sep 21 15:35:15 2011 From: mathog at caltech.edu (mathog) Date: Wed, 21 Sep 2011 12:35:15 -0700 Subject: [Beowulf] SGE forgetting its queues at restart following node rename In-Reply-To: References: Message-ID: <3a0078bd5778e78ec4b45f4fc0310a2b@saf.bio.caltech.edu> FYI (Just to have it posted, in case anybody else ever runs into this.) A little while back I moved same names around in the cluster. To do so, in SGE a bunch of queues and some hosts were removed and then added back. There was much trial and error in doing so - I make no claim that the right commands were issued in the proper order. However, in the end all the queues were as desired and they all stayed up and running. Until the node was rebooted, at which point SGE came back up with only two queues present. After much poking around the problem was finally locate: some of the old host names and old queues were still present in files under: $SGEROOT/default/spool/qmaster/qinstances and as soon as SGE hit one of those during startup, it would stop creating all further queues. The error message that resulted when that happened was of this form: 09/21/2011 12:22:56|qmaster|safserver|E|cannot recreate queue all.q from disk because of unknown host mendel and appeared in: $SGEROOT/default/spool/qmaster/messages Regards, David Mathog mathog at caltech.edu Manager, Sequence Analysis Facility, Biology Division, Caltech _______________________________________________ Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.