Re: Bizarre slowdowns

Bill Marcum wrote:
On Wed, 22 Aug 2007 09:09:53 -0700, CptDondo <yan@xxxxxxxxxxxxxxxx> wrote:

I have a newer server that I built using a gigabit 945Pl-S3 mobo, with 2 GB ram, and a Core2Duo CPU.

It has 4 ea. 400 GB SATA drives, Seagate ST3400620AS drives in a RAID-5 array. The SATA controller is an intel 82801 chipset.

Once in a while, the system starts thrashing for no apparent reason. Response time goes to crap - 10 minutes to open a terminal window, that sort of thing - but there is nothing in the log files that would indicate a problem. Once the system starts thrashing, the only way to recover is a hard power-off.

Does this always happen around 2 am? If so, it could be a cron job, maybe something like updatedb or beagle. You could modify /etc/crontab
(or /etc/cron.d/anacron) to run daily|weekly|monthly tasks when they won't interfere with other activities.



No, there doesn't seem to be any reason for it that I can tell.

There are some jobs that hit the disks somewhat hard, but nothing too difficult. (rsync via internet, myth recording to disk, myth commflagging, that sort of thing.)

--Yan
.

Relevant Pages

  • Re: Bizarre slowdowns
    ... Response time goes to crap - 10 minutes to open a terminal window, that sort of thing - but there is nothing in the log files that would indicate a problem. ... the only way to recover is a hard power-off. ... There are some jobs that hit the disks somewhat hard, ... (rsync via internet, myth recording to disk, myth commflagging, that sort of thing.) ...
    (comp.os.linux.misc)
  • Re: Bizarre slowdowns
    ... Response time goes to crap - 10 minutes to open a terminal window, that sort of thing - but there is nothing in the log files that would indicate a problem. ... the only way to recover is a hard power-off. ... There are some jobs that hit the disks somewhat hard, ... (rsync via internet, myth recording to disk, myth commflagging, that sort of thing.) ...
    (comp.os.linux.misc)
  • Re: bios disk numbers and device names
    ... You can wire disks down in SCSI, not sure about ATA. ... That said if you run 5.x devfs makes this sort of thing MUCH less painful.. ...
    (freebsd-stable)
  • Re: Corrupted Drive
    ... > Is there something I should be doing to prevent this sort of thing? ... > system that doesn't get rebooted very often, should fsck be run manually ... Disks that run continuously for months/years are most likely to fail ... fsck'ing a filesystem periodically probably isn't necessary; ...
    (Fedora)
  • Re: New volume group not imported on reboot
    ... > Auto-On is set to 'y' ... What sort of disks are involved? ... For example, EMC Symmetrix BCV volumes ...
    (comp.unix.aix)