What is the server OS ?

  Can you strace/truss relevant processes (save, nsrindexd) ?

  Can you rule out a hardware fault (SAN switch, HBA, NIC, switch) ?

  Are there any relevant error messages in your syslog ?


On 11/24/2012 06:59 PM, jeronimo wrote:

Hi there,

I have a problem during backup when all sessions almost stall.
In NMC I can see a throughput of like 100KB/sec or less from the 4-5 hosts 
currently backing up to the server.
Up to that point all was fast, then it stalls for a few hours, then it's back 
to full speed again.
Doesn't look like a bottleneck to me. Client's resources are ok. Server's 
resources are ok as well. (cpu,disk,network)
But what's going on?
This is Networker 7.5 and just for the sake of illustration see the io graph 
for one of the hosts here: http://jeronimo.servehttp.com/dataio.jpg

During the stall I was watching savepnpc using lsof, it was slowly accessing 
every file/dir one after another as it should, but that directory is nothing 
special, the filesystem isn't corrupt. Well, I still have to find out if it 
always stalls in there.
But as I said, *all* clients stall, not only that one. Points me to trouble 
with the server...

Any ideas???

Thanks

marki

+----------------------------------------------------------------------
|This was sent by jm+backupcentral AT roth DOT lu via Backup Central.
|Forward SPAM to abuse AT backupcentral DOT com.
+----------------------------------------------------------------------

--

-- Yaron.