Hi,
I am using SQUID 2.4STABLE2 on a Linux 2.4.5 machine and I noticed that
every 5-10 sec. the CPU spikes to 50-100% for about 2 sec. then comes back
to normal, only to reach 100% again 10 sec later.
The cache has no load at all, i.e. nobody is sending requests, it is just
sitting there "idle". I don't think these CPU usage spikes are normal.
I tried both ufs and aufs and they both have the same problem. I was not
able to start it with the storage type 'null' even though I compiled it
with '--enable-storeio=aufs,null,ufs'
ICP and WCCP are turned off. The configuration is very basic.
I ran an strace on the process to see what it's doing, but I can't make
any sense of it:
*repeated a lot*
poll([{fd=10, events=POLLIN}], 1, 0) = 0
*shows up once in a while, possibly corresponding to the CPU spikes??*
poll(0xbfffcc10, 0, 0
) = 0
Has anybody noticed this behavior? Any solutions?
Thanks.
Received on Sun Sep 09 2001 - 13:11:25 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:02:06 MST