[squid-users] Squid RP in front of Atlassian Stash with SSL - 100% CPU and not responding

I've been stuck on this for 2 days now.

After accessing this cache a couple of times, the CPU jumps to 100% and
squid stops forwarding requests.
It remained like this for 24 hours until I killed the process.

Adding the following lines to the config has resolved the issue:

########################
always_direct allow all
sslproxy_flags DONT_VERIFY_PEER
########################

BOTH of those lines must be added otherwise the symptoms return.

But why? What are the security implications of these settings and why would
they resolve the infinite loop problem?

Config snippet
########################
https_port 443 accel cert=/path/to/CertAuth/cert.cert
key=/path/to/CertAuth/key.pem vhost defaultsite=www.domain.com

cache_peer source.domain.com parent 443 0 no-query originserver ssl
sslversion=3 connect-timeout=8 connect-fail-limit=2 sslflags=DONT
_VERIFY_PEER front-end-https=on name=source login=PASSTHRU
acl sites_source dstdomain source.domain.com
cache_peer_access source allow sites_source
acl http proto http
acl https proto https
########################

Replicated problem with:
Ubuntu 12.04 - Squid 3.2 - compiled.
CentOS 6 - Squid 3.3 - compiled.
CentOS 6 - Squid 3.1 - installed from repository.

Atlassian Stash with paid for SSL Certificate - looks fine connecting
directly in firefox.
Atlassian Stash with self signed SSL certificate.

P.S. This appears to be the same problem:
http://www.squid-cache.org/mail-archive/squid-users/201111/0416.html

--
View this message in context: http://squid-web-proxy-cache.1019090.n4.nabble.com/Squid-RP-in-front-of-Atlassian-Stash-with-SSL-100-CPU-and-not-responding-tp4661490.html
Sent from the Squid - Users mailing list archive at Nabble.com.