RH8 vs RH9

Robert G. Brown rgb at phy.duke.edu
Wed Oct 1 08:24:13 EDT 2003


On Wed, 1 Oct 2003, Cannon, Andrew wrote:

> Hi All,
> 
> We have a small test cluster running RH8 which seems to work well. We are
> going to expand this cluster and I was wondering what, if any, are the
> advantages of installing the cluster using RH9 instead of RH8? Are there any
> disadvantages?

Many humans wonder about that, given the very short time that RH8 was
around before RH9 came out.  The usual rule is that major number
upgrades are associated with changes in core libraries that break binary
compatibility, so that binaries built for RH 8 are not guaranteed to
work for RH 9.

I think that the easiest way for you to determine precisely what changed
is to look at e.g. 

  ftp://ftp.dulug.duke.edu/pub/redhat/linux/9/en/os/i386/RELEASE-NOTES

and see if anything in there is important to your work.  Beyond that,
there are a few issues to consider:

   a) 8 will, probably fairly soon, be no longer maintained.  9 will be,
at least for a while (possibly for one more year).  Of course the
maintenance issue right now is very cloudy for RH in general with the
Fedora/RHEL situation a work in progress.  However, maintenance alone is
(in my opinion) a good reason to be using 9 and to move from 8 to 9 to
achieve it.  Fedora will likely be strongly derived from 9 and the
current rawhide in any event.  How the "community based" RH release will
end up being maintained is the interesting question.  One possibility is
"as rapidly as RHEL plus a few days", the difference being the time
required to download the GPL-required logo-free source rpm(s) after an
update and rebuild them and insert them into the community version.  Or
of course you can spring for a RHEL license (set) for your cluster,
which may or may not be reasonable in cost or scale well per node by the
time all the University-price dickering is done.

   b) 9 had some fairly significant library upgrades, service upgrades,
and bug fixes.  That doesn't mean 8 is "bad" -- it just means that your
chances of encountering trouble with 9 are in principle smaller than
with 8, and one hopes that the upgrades added a bit to performance as
well.

  c) A lot of the enhancements in 9 were more useful or relevant to
userspace and LAN client operation (CUPS or Open Office, for example)
than they were to cluster nodes.  So in that sense perhaps it doesn't
matter as much.

We're using 9 on a bunch of hosts and nodes with happiness.  We're also
using 7.3 (still) on a bunch of hosts and nodes with happiness.  We
skipped 8 only because they released 9 before we finished creating a
stable/tested 8 repository as RH changed their release cycle and dropped
the .0, .1 and so forth "correction" releases.

I don't know that we'll ever use RHEL with happiness unless RH charges
something like $1 per system as their university price (which isn't
insane, actually, given that an entire university can install and
maintain, as Duke does, off of a single campus-local repository largely
run by and debugged by and maintained by campus administrators, so RH's
costs don't scale at all strongly with the number of internal campus RH
systems).  

Fedora, quite possibly, but as noted we are fearful, uncertain, and
doubtful at the moment, for once because of real issues and not just as
a sort of Microsoft joke...

    rgb

> 
> Thanks
> 
> Andrew
> 
> Andrew Cannon, Nuclear Technology (J2), NNC Ltd, Booths Hall, Knutsford,
> Cheshire, WA16 8QZ.
> 
> Telephone; +44 (0) 1565 843768
> email: mailto:andrew.cannon at nnc.co.uk
> NNC website: http://www.nnc.co.uk
> 
> 
> 
> NNC's UK Operating Companies : NNC Holdings Limited (no. 3725076), NNC Limited (no. 1120437), National Nuclear Corporation Limited (no. 2290928), STATS-NNC Limited (no. 4339062) and Technica-NNC Limited (no. 235856).  The registered office of each company is at Booths Hall, Chelford Road, Knutsford, Cheshire WA16 8QZ except for Technica-NNC Limited whose registered office is at 6 Union Row, Aberdeen AB10 1DQ.
> 
> This email and any files transmitted with it have been sent to you by the relevant UK operating company and are confidential and intended for the use of the individual or entity to whom they are addressed.  If you have received this e-mail in error please notify the NNC system manager by e-mail at eadm at nnc.co.uk.
> 
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> 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
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf



More information about the Beowulf mailing list