On 24.07.2012 06:14, Dean Weimer wrote:
I have had a Squid server that has been running for some time, and
all of the sudden started having problems, this server runs as both a
forward and reverse proxy on different ports. The reverse proxy part
seems to be responding fine, but the forwarding is all of the sudden
logging errors in cache.log file about tunnelSate The parent server
appears to be running fine, and is serving requests for a few hundred
clients that access it directly without issue.
Of course the full log message is exactly what the subject of this
email message is.
tunnelConnectTimeout(): tunnelState->servers is NULL
The parent server is running 3.1.20 and this server is on 3.1.18,
anyone have any idea what would cause this type of behavior to start
happening?
Something is timing out in CONNECT. Which should not happen normally,
but Squid closes the connections cleanly so its just a warning.
It's not exactly clear whether this is a timeout event arriving after
the tunnel has already closed properly (erasing its "servers" list), or
if it is a real unable-to-connect problem failing after a set of
retries. Are your client connections through the 3.1.18 proxy having
trouble with HTTPS?
Amos
