On 4/03/2013 11:38 p.m., Wolfgang Grim wrote:
Hi there, I am currently builing a reverse proxy scenario with 3 SQUID Servers where one squid is located in a cloud environment an two are located on one site. The squid in the cloud answers all requests and forwards the packages to one of the other servers without round robin. So when the first squid is available, the request should be forwarded to this server, when the server has issues and is not reachable, all requests should be forwarded to the second server.
This is the config:
squid3 --- squid1
|-squid2 (only if squid1 is NOT rechable)
The problem is, when squid1 is not reachable, there is a weird behaviour of squid3.
the requests are not forwared and every 30 seconds the following information appears in cache.log
2013/03/04 11:27:23| BUG #3329: Orphan Comm::Connection: local=91.250.97.5:49863 remote=188.118.198.34:443 FD 14 flags=1
2013/03/04 11:27:23| NOTE: 9 Orphans since last started.
2013/03/04 11:27:54| BUG #3329: Orphan Comm::Connection: local=91.250.97.5:49921 remote=188.118.198.34:443 FD 14 flags=1
2013/03/04 11:27:54| NOTE: 10 Orphans since last started.
Bug 3329 has been silenced in this months 3.2.8 package.
Sometimes also
2013/03/04 11:26:52| fwdNegotiateSSL: Error negotiating SSL connection on FD 14: error:1408E0F4:SSL routines:SSL3_GET_MESSAGE:unexpected message (1/-1/0)
2013/03/04 11:26:52| TCP connection to 195.202.164.227/443 failed
SSL issues are separate. You may find them gone as well with 3.2.8,
however there is nothing I'm aware of specifically for fixing these
messages.
FYI: they usualy mean plain text or broken SSL response when connecting
to the remote HTTPS service.
Amos
