Hits: 4064

Recently, Bill Broadley of UC Davis presented the Beowulf mailing list with a survey on NFS and HPC. Bill has collected the results and given permission to post a summary on Cluster Monkey. Many thanks for all the 27 respondents (who are seasoned HPC administrators) . Note that not all respondents answered all questions..

1) cluster OS:

2) Appliance/NAS or linux server

3) Appliances used (one each, free form answers):

4) Which kernel do you use:

5) what kernel changes do you make

6) Do you often see problems like nfs: server not responding, timed out:

7) If you see NFS time outs what do you do (free form answers)

8) which NFS options do you use (free form):

9) Any explanations:

10) what parts of the file system do you use NFS for (free form):

11) How many nodes can mount a single NFS server at once:

12) How many NFSd daemons do you run per NFS server

13) Do you use NFSd or user space

14) What interconnect do you use with NFS?

15) If IB what transport (10 responses)

16) If IB, do you use connected mode (8 responses)

17) Do you use UDP or TCP (25 responses)

18) Which other network file systems do you use? (24 responses)

19) Are the other network file systems more or less reliable than NFS?

20) Do you support MPI-IO (not just MPI)

21) Any tips for making NFS perform better or more reliably?

22) Any tips for making NFS clients perform better or more reliably?

23) Anything you would like to add: