Re: [SQU] Squid 2.2-STABLE4 hanging

Hi, I thought of the card but three things prevented me from
changing it:
1. The NIC is on-board
2. There is more than one computer on the network causing the
problem - this is the only one I have identified so far
3. If I telnet to the proxy port from this workstation I do not have
the same fault.

On 25 Oct 00, at 7:34, John Cuson wrote:

> just a thought ... have you thought about changing your network card?
> once in a blue moon i get one that fails in a unique manner, may affect
> one network service but no others, etc. it always surprises me when the
> card is the reason for similar failures, but then that's part of the fun,
> right?
>
> john cuson
> jcuson@labs.isdh.state.in.us, jcuson@alumni.indiana.edu
>
>
> The trouble with our times is that the future is not what it used to be.
> -- Paul Valery
>
>
>
> >>> "Eugene Geldenhuys" <eugeneg@tfx.com.au> 10/24 9:24 PM >>>
>
> Hi - It's been a long while, but I think that I may have narrowed
> down the cause of this problem.
> I was testing the cache today and I found that whenever I envoked
> the web browser from a particular system on the LAN it caused
> squid to hang.
> Investigating further I tailed the access.log and cache.log files and
> found that when I accessed the browser on this particular system squid
> hangs with no log entries, it's as though it attempts to process the
> request and hangs before it can log the activity. I did a tcpdump for the
> particular system and compared it to a dump from another system - I can
> see no differences up to the point that the system hangs. To explain the
> configuration of the Linux system - it is a highly modified Redhat 6.1
> system with kernel 2.2.10. The server has 2 LAN segments attached, both
> which use squid. The workstation in question is a Windows 98 with
> Mickeysoft Explorer 4.0 (I upgraded to 5.0 with the same result) This is a
> really obscure problem, but if we can nail down the cause, we may be able
> to prevent anyone else from suffering the same problems.
>
> > Looks like I may have someone forcing my cache to hang, after
> > months of reliable service my cache has suddenly become prone to
> > hanging.
>
> Not the case, it is a system rather than an individual
> >
> > I have checked the logs and the only clue I have is that just prior to
> > the cache hanging I get the following entries in cache.log:
> >
> > 2000/09/06 14:59:00| sslReadClient: FD 39: read failure: (104)
> > Connection reset by peer
> > or
> > 2000/09/07 11:02:01| urlParse: Illegal character in hostname
> > 'www.lange_devil.homestead.com'
> >
> > Is there some way that the cache can be affected by illegal chars
> > in URL names ?
> >
> > The access.log files holds no clues that I can see.
> >
> > I have gone to the lengths of replacing the box in it's entirety
> > including a fresh load of Linux and I still have the same problem.
>
> I no longer think that this is a configuration issue
> Best Regards
> Eugene Geldenhuys
> MCNE ECNE MCSE MCP
>
> TFX SOLUTIONS -
> PROFESSIONAL NETWORK DESIGN ,IMPLEMENTATION AND SUPPORT
>
> --
> To unsubscribe, see http://www.squid-cache.org/mailing-lists.html
>
>

Best Regards
Eugene Geldenhuys
MCNE ECNE MCSE MCP

TFX SOLUTIONS -
PROFESSIONAL NETWORK DESIGN ,IMPLEMENTATION AND SUPPORT

--
To unsubscribe, see http://www.squid-cache.org/mailing-lists.html