[OpenAFS] Performance problems seem to be coming back

Dale Pontius pontius@btv.ibm.com
Fri, 09 Sep 2011 09:41:10 -0400

Earlier this year I posted about AFS performance problems on our 
network, which were related to underlying network performance problems.  
Remedial actions were taken and longer-term strategies put in place, and 
things got better.

At least for a while.  Over the past few weeks, it seems that 
occasionally things are getting bad again, though nowhere nearly as bad 
as they were earlier this year.  I guess I should say here that I don't 
have access to the servers,  in order to do any sort of metrics there.  
I'm gearing up for the battle with the support people, to convince them 
that there is a problem, and I'm not just a "picky user."  That job is 
complicated by the fact that we have many complacent, accepting users 
around here, so this often turns into a matter of, "It's only you guys 

Part of the problem has been metrics.  We (or at least I) don't really 
seem to have to tools and techniques to see what's going on here.  
During the previous problem timeframe, "smokeping" turned out to be a 
critical diagnostic tool, crude as it was.  At that time I got the 
impression that what I'd really like to be doing is monitoring the round 
trip time for afs communications.  I took a brief look at wireshark, 
played with filtering, and managed to watch the afs packets go and 
return, and am pretty sure that there was enough information there to do 
such monitoring.  I also got the impression that wireshark supported 
some sort of "filter macros" to make this process automatic, and that 
some stuff was available for afs.  I started looking into it, but it 
looks to me as if all of this stuff was meant for someone more expert 
than I, so I didn't get far.  Plus about that time our network started 
getting better, and because of the network problems we were behind on 
our "real work", etc.  It got dropped.

Does anyone have advice on how to, as simply and automatically as 
possible, monitor and log afs packet round trip times?


Dale Pontius
Senior Engineer
IBM Corporation
Phone: (802) 769-6850
Tie-Line: 446-6850
email: pontius@us.ibm.com

This e-mail and its attachments, if any, may contain confidential and privileged material for the sole use of the intended recipient. Any review, use, distribution or disclosure by others is strictly prohibited. If you are not the intended recipient (or authorized to receive for the recipient), please contact the sender by reply e-mail and delete all copies of this message from your system without copying it and notify sender of the misdirection by reply e-mail.