Re: [Networker] mminfo slow to respond
2009-10-12 16:23:18
Rudy Munguia wrote:
The nsr and mm directories are on the same ufs filesystem, The index is on a
separate san attached 1.2TB volume. There are 28 1.2 TB AFTD's supporting DBO
operations. All SAN disk are provided by a Clariion cx700. Average utilization
of the system is 78% and the dev i/o waits are nothing to worry about.
i tried attaching the output requested, but there were over 10,000 lines which
was bounced by the listserver.
Actually I was able to look at it. I cannot explain what I see. Can
you run truss on the mminfo process and see if you can understand what
is it waiting for ? If you don't understand the output send a hundred or
so lines.
The system time (in sar -u output) does not make much sense. It
should indicate lots of I/O but in your case the disks are not busy at all.
Are you sure your SAN drivers are OK ?
Do you observe any other slowness of the system ?
Can you explain the high system time, or correlate it to another
activity currently happening on the system ?
What do you see in top ?
Do you see any relevant messages in /var/adm/messages or dmesg ?
I have in /etc/syslog.conf:
kern.debug /var/adm/kernel
Are you sure the Clariion is not very busy ? Do you share it with
anyone ? Do you have your own RAID group ?
________________________________
From: Yaron Zabary <yaron AT ARISTO.TAU.AC DOT IL>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Sent: Sunday, October 11, 2009 3:57:49 AM
Subject: Re: [Networker] mminfo slow to respond
I have seen similar problems if your /nsr is on the same file system with an
aftd device.
What is the output of
df -h /nsr/mm
df -h /nsr/index
df -h /nsr/res
And while the mminfo is running:
sar -d 1 10
sar -u 1 10
Rudy Munguia wrote:
Hello,
I am running a very simple mminfo query 'mminfo -avc xyz -r"volume"' in order
to find all tapes that contain savesets for the particular client. The first query took
about 20hours, the second took about 30, the third query has been running for 4 days now.
This is running on 7.4.4 on Solaris 9 on a Sun V440 with 4 Processors and 8GB
of RAM. It handles only about 300 clients. We have a total library of only
14,000 tapes.
I have about 50 more clients that I need to run this same query for.
Any suggestions?
To sign off this list, send email to listserv AT listserv.temple DOT edu and type
"signoff networker" in the body of the email. Please write to networker-request
AT listserv.temple DOT edu if you have any problems with this list. You can access the
archives at http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
--
-- Yaron.
To sign off this list, send email to listserv AT listserv.temple DOT edu and type
"signoff networker" in the body of the email. Please write to networker-request
AT listserv.temple DOT edu if you have any problems with this list. You can access the
archives at http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
|
|
|