Well the first step was to spread things around time wise so everything is not
happening at once. I'm going to see what that produces as far as change
goes. But I see something called suse.de-check-battery that I'm pretty sure
doesn't need to execute on a daily basis, or at all for that matter. But I
will tend to those things after I'm back on the 11th.
Thanks for the various insights gang. Much appreciated.
Rob van der Heij <[EMAIL PROTECTED]>
Sent by: Linux on 390 Port
<[email protected]>
To
[email protected]
cc
09/01/2006 10:20 AM
Subject
Re:
Strange amounts of paging reported in VM at 4:00 AM.
Please respond to
Linux on 390 Port <[email protected]>
On 9/1/06, James Melin <[EMAIL PROTECTED]> wrote:
> Not 100% sure. But I'll go through those scripts over the weekend and look
> for a list of specific things.
Things like locatedb, logrotate or backup of the RPM database or
backup in general are likely candidates. They read a lot of data into
page cache and with sufficiently large virtual machines it moves the
hot spot through the virtual machine. When you overcommit storage on
z/VM enough, you will find z/VM page the same amount that Linux reads
and writes during that process.
I know of one configuration where we had VM thrash for hours when
cron.daily took off at 00:15. If these are useful things you might
want to spread it over the night, but quite often they're not really
needed at all.
Rob
--
Rob van der Heij
Velocity Software, Inc
http://velocitysoftware.com/
----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
