[OpenAFS] Monitoring performance of fileservers using cacti or munin

Jose Calhariz jose.calhariz@tagus.ist.utl.pt
Thu, 13 Oct 2011 14:39:05 +0100


I am investigating possible performance issues in my openafs cell.  

But the Linux measurements are all normal.  The load average is usually
below 1.  Using "iostat -k 5" to monitor the vicepa partition I see at
most 140 tps, transactions per second.  The problems could be on the
clients, not the fileservers.  I don't have admin access to the
client machines to do an investigation there so I am just guessing
where the problem is.

My Google searches let me find nagios-plugins-afs software.  Now I am
using it.  No issues were found on the fileservers. 

So I would like to get performance values specific to openafs.  I
would like to find what are the peak hours or the fileservers that are
under most stress.  I know only about scout command. 

What do you use?  Are someone using cacti or munin for performance

I am missing something from the manual pages or openafs documentation?   

My cell is running Debian 5.0, using openafs 1.4.11.  Debian oldstable
now.  In the near future will change to Debian 6.0 using openafs

     José Calhariz


As amizades reatadas requerem maiores cuidados do que as que nunca foram rompidas

--La Rochefoucauld