Re: [cobalt-users] KSWAPD

[flash22@xxxxxxx]

On Fri, 26 Oct 2001, Paul Enlund wrote:

> Hi
> 
> I'am not to familiar with a Cobalt server a colleague runs but have found
> that its current state does not look correct.
> 
> kswapd is zombied "root         5  0.0  0.0     0    0 
> ?        Z    Oct18   0:06 [kswapd <defunct"
> and the cpu info looks bad.
> uptime: 12:16am  up 7 days,  2:17,  1 user,  load average: 2.78, 2.71, 2.68
> 
> The machine is quiter at this time of day but its daytime operation is 
> critical.
> 
> In my colleagues absence is this machine best rebooted  now ?

run badblocks on the swap partition first, hard disk errors on the swap
area will zombie the swapper...other than that, a near kernel panic..
prolly find VM errors in syslog/messages i suspect also...

[That is a very evil error]

gsh