Hello:
I have an odd problem with an IBM EServer XSeries 305 (aka x305). Sometimes it seems to just sit there, completely unresponsive to the command line (ssh and console), for several seconds. Other times it works normally. top will sometimes die with a floating point error while waiting for it to start. When top does work, it often shows a system load of 0.5, with high interrupt load.
At first I though that the tg3 driver was not playing nicely, so I disabled the onboard NICs and installed a 3Com card. This seemed to clear up the problem, but it's now returned two days later.
System is running CentOS 4.4, no X, Apache and MySQL.
CPU is a P4 1.8GHz 512MB RAM 40GB ATA-100 disk
Any suggestions on where to look next are appreciated.
--Chris
I have an odd problem with an IBM EServer XSeries 305 (aka x305). Sometimes it seems to just sit there, completely unresponsive to the command line (ssh and console), for several seconds. Other times it works normally. top will sometimes die with a floating point error while waiting for it to start. When top does work, it often shows a system load of 0.5, with high interrupt load.
At first I though that the tg3 driver was not playing nicely, so I disabled the onboard NICs and installed a 3Com card. This seemed to clear up the problem, but it's now returned two days later.
System is running CentOS 4.4, no X, Apache and MySQL.
CPU is a P4 1.8GHz 512MB RAM 40GB ATA-100 disk
Any suggestions on where to look next are appreciated.
--Chris
Have you run or considered running memtest86+ from the centos install cd ??
If physical memory problems isn't it, have you done this
chkconfig --list | grep :on | more
And shut off any of the items that are not necessary on your unit?
I can provide a list if you need it of things that are typically not needed in our situations.
You more than likely know this yet...
ps ax | more
will show many of them as well.
- rh
-- Robert - Abba Communications Computer & Internet Services (509) 624-7159 - www.abbacomm.net
On Oct 23, 2006, at 3:43 PM, R Lists06 wrote:
Have you run or considered running memtest86+ from the centos install cd ??
Have not tried that. The amount of memory appears to be correct, and no processes are dying in other core dumps (though top may be touching a memory structure that's in bad hardware).
Most of the other services are off, but I'll go back over the list.
Thanks, --Chris
Chris Boyd wrote:
Have not tried that. The amount of memory appears to be correct, and no processes are dying in other core dumps (though top may be touching a memory structure that's in bad hardware).
Do you have the asmi management stuff on that server? You should install the driver for it or try to disable it in the BIOS. With no driver and an active card, strange things can happen when the card borrow the CPU.