On 21/02/19 6:23 pm, Van Order, Drew (US - Hermitage) wrote: > Hello folks, we’re running Squid 3.5.28 on four Windows 2016 VM’s, each > has 4 CPU, 8 GB memory, 10 GB NIC. We implemented Squid to support > forwarding Azure Log Analytics data, it’s all CONNECT. The Squids are > load balanced through a F5. There are less than 1,000 servers sending > data to Log Analytics. > > > > All four Squids are regularly crashing, and I don’t know how to > interpret the errors in cache.log. It crashes if we disable caching too. > ... > > 2019/02/20 09:42:32 kid1| FD 12, 10.5.11.12 [Stopped, reason:Listener > socket closed job1]: (14) Bad address >
Something other than Squid closed the network socket Squid was using to receive new client connections (the http_port socket). The only things which should know that socket even exists are Squid and the operating system. This is not an error I've seen before. Is this Squid maybe running in a container or VM which is being hibernated, or suspended, or anything along those lines which may cause the OS filedescriptors to change unexpectedly? Amos _______________________________________________ squid-users mailing list squid-users@lists.squid-cache.org http://lists.squid-cache.org/listinfo/squid-users