Hi there All!
I use Squid 2.6 STABLE5 in our intranet (about 150-200 permanent users)
with NTLM users authentication via SAMBA ntlm_auth external helper.
All works good.
But I can see some wornings in squid cache.log:
2006/11/28 18:43:53| comm_accept: FD 74: (233) No buffer space available
2006/11/28 18:43:53| httpAccept: FD 74: accept failure: (233) No buffer
space available
2006/11/28 18:44:06| comm_accept: FD 74: (233) No buffer space available
2006/11/28 18:44:06| httpAccept: FD 74: accept failure: (233) No buffer
space available
...
And so on... about 2-3 per minutes.
This indicates a kernel resource problem.
You probably need to increase network buffers.
But I am not a wizzard on HP-UX.
On some UNIX-es :
# netstat -mb
may provide more info.
M.
