> No. Swap usage is very low and much lower than memory being used by the
> process. One thing I forgot to mention in my original post is the error
> messages that keep appearing in the cache.log.
>
> My original configuration had the two internal caches as siblings of
> each other but I continually received the following error messages:
>
> 2002/05/26 14:34:09| Detected DEAD Sibling: 132.147.199.1/3128/3130
> 2002/05/26 14:34:14| Detected REVIVED Sibling: 132.147.199.1/3128/3130
>
> The servers where always in contact with each other but the squid
> process seemed to have problems. I disabled the siblings and just left
> it with the parent cache but still get errors of the sort:
>
> 2002/05/29 15:40:29| TCP connection to 132.147.193.1/3128 failed
> 2002/05/29 15:43:59| TCP connection to 132.147.193.1/3128 failed
> 2002/05/29 15:43:59| TCP connection to 132.147.193.1/3128 failed
>
> Once again network connectivity is still their but it's as if the squid
> process has fallen off the server and as such has told the originating
> cache to forget about it.
I think you have to have a look at the NICs or the TCP stack. Something is
definetly wrong with your network connection, maybe temporary, maybe all the
time.
Hermann
Received on Wed May 29 2002 - 03:25:28 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:08:15 MST