On Tue, Sep 11, 2012 at 05:46:17PM -0400, Brodie, Kent wrote:
Ian- thanks. (anyone else?)
I am starting to suspect that the Dell 8024 10-gig switches that are at the
heart of my cluster infrastructure are.. crap. CONFIRMING that is of
course, difficult.,.....
Check your counters (as mentioned before). You really should watch
these anway to get sense of your overall network load.
Check for firmware upgrades.
Check for something "portfast" (I think it's called) on the nodes too.
That's been known to cause trouble in the HPC realm.
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
--
Jesse Becker
NHGRI Linux support (Digicon Contractor)
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
