Now I'm getting messages like this in my log:

2004/04/27 20:08:39| idnsSendQuery: FD 8: sendto: (105) No buffer space
available
2004/04/27 20:08:39| comm_udp_sendto: FD 8, 207.35.37.21, port 53: (105)
No buffer space available

Please are having problems logging into hotmail but they can get to the
home page.  Many other sites are having problems but this is a big example.

I'm running this on a Fedora Core1 system.  This is a dedicated cache
server running wccp.  Is there a better recommendation for an OS to use
squid on that is historically more stable/reliable?  This cache is
supporting our dial-in userbase and we have about 170 dialin lines.  We
get a lot of complaints when squid acts up.

Thanks,
Serge.

[ Post a follow-up to this message ]

s] More Errors
>=20
>=20
> Now I'm getting messages like this in my log:
>=20
> 2004/04/27 20:08:39| idnsSendQuery: FD 8: sendto: (105) No=20
> buffer space=20
> available
> 2004/04/27 20:08:39| comm_udp_sendto: FD 8, 207.35.37.21,=20
> port 53: (105)=20
> No buffer space available

Some kernel resource problem.
Check :
=20
% free

Also check /var/log/messages ; check any kernel messages
related to networking (...NET...)

>=20
> Please are having problems logging into hotmail but they can=20
> get to the=20
> home page.  Many other sites are having problems but this is=20
> a big example.

Does it work when the browser is configured to use SQUID directly; via
proxy config ?

M.
>=20
> I'm running this on a Fedora Core1 system.  This is a dedicated cache=20
> server running wccp.  Is there a better recommendation for an=20
> OS to use=20
> squid on that is historically more stable/reliable?  This cache is=20
> supporting our dial-in userbase and we have about 170 dialin=20
> lines.  We=20
> get a lot of complaints when squid acts up.
>=20
> Thanks,
> Serge.
>=20

[ Post a follow-up to this message ]