Hi folks, We have a slink system that has been running quite nicely for some time, which acts as a pretty important server (for us, at least). One of the functions that it serves is as a proxy server, running squid (the current slink version, whatever that is). Yesterday I noticed that it had stopped accepting connections. The proxy server to which it was chained was still working perfectly, but our little squid was quite dead. Trying to fetch a URL using squidclient, for example, just froze (or rather, did nothing). While not responding, it wasn't logging any connections in any of it's logs. Having a look via top, it seemed that the parent squid process (is `parent' the right term - it was the one that the /var/run/squid.pid defined, anyway) was using 90%+ CPU time. After starting and stoping it, it climbed to this high CPU usage very rapidly. The only thing that has changed lately is the proxy we are chaining through (windoze box - don't ask), but squid worked fine with that proxy for a while at least, and all the other browsers, etc., can use that proxy fine by themselves. Nothing has changed in the squid.conf file during this time. If anyone has any suggestions, I'd be really interested in hearing them. At the moment I have a hacked-up netcat session running out of inetd to redirect to the other proxy, but that's hardly an acceptable situation (and really defeats the purpose of a proxy anyway). If conf/log files would help diagnose this problem, let me know and I'll post them. cheers, damon -- Damon Muller (dm-sig6@empire.net.au) / It's not a sense of humor. * Criminologist / It's a sense of irony * Webmeister / disguised as one. * Linux Geek / - Bruce Sterling - Running Debian GNU/Linux: Doing my bit for World Domination (tm) -
Attachment:
pgpAkD1_OrAZW.pgp
Description: PGP signature