[Dxspider-support] Spider stall.
rene_olsen at post3.tele.dk
Mon Oct 9 22:10:07 BST 2006
Thanks for all the replies.
I have done some more investigation. Spider didn't actually stop to respond until Sunday at
09Z or something like that.
But when looking at the mrtg graphs it looks like the problem started already on Friday
evening around 21Z. At that point I can see the CPU load starting to rise from its usual close
to nothing and During Saturday and Sunday until it stopped responding it is just a curve that
rises all the time.
And at the end spider simply stalls and stops responding.
I did not check the netstat status before I killed the cluster.pl proccess, but if it happens
again I will try to check it out.
I have a node on ampr net that I link to, and sometimes that link is very shaky. So could
imagine that sometimes that link could get a rather large amount of data waiting to be sent.
But it seems like this stall was 1 1/2 day under way, as I have never seen such CPU loads
under normal conditions.
Vy 73 de Rene / OZ1LQH
More information about the Dxspider-support