Understood (of course). I'll assemble the information needed.<div><br></div><div> - Stig<br><br><div class="gmail_quote">On Tue, Dec 21, 2010 at 9:51 AM, Kristian Lyngstol <span dir="ltr"><<a href="mailto:kristian@varnish-software.com">kristian@varnish-software.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im">On Tue, Dec 21, 2010 at 09:36:14AM +0100, Stig Bakken wrote:<br>
> Kristian, please let me know whether I should file a ticket for this.<br>
<br>
</div>So far I haven't seen anything that suggests a bug.<br>
<br>
A single slow response is all the logs show - for better or worse. That<br>
could just as well just be a slow client or packet loss.<br>
<br>
What I need to be able to determine if this is a bug in Varnish,<br>
network-issue or phase-of-moon glitch is:<br>
<br>
- Several occurrences of slow responses and the requests around it in<br>
varnishlog (you can obfuscate the IP - but I need to be able to tell<br>
which requests come from the same IP: ie: if there are two IP's, A.A.A.A<br>
and B.B.B.B are ok, but X.X.X.X for both are not).<br>
- 'varnishstat -1' output<br>
- 'sysctl -a' output<br>
- Network configuration on the machine(s), specially with regards to any<br>
load balancers<br>
- 'dmesg'<br>
- Complete VCL (I do believe I already have that)<br>
- All start-up parameters. Any other running services on the machines.<br>
<br>
If that doesn't reveal the nature issue, I'll need to see tcpdump-output of<br>
the incident to track it further - at that point, obfuscation is not<br>
practically possible without risking the integrity of the data.<br>
<br>
In short: all you can give me.<br>
<br>
The issue doesn't have an obvious explanation and there isn't really any<br>
reason why a cache hit would take any real amount of time when we rule out<br>
busy objects. I should also point out that I can only really spend time on<br>
this to rule out or confirm if it is a varnish bug or not.<br>
<font color="#888888"><br>
--<br>
Kristian Lyngstøl<br>
Product Specialist<br>
Varnish Software<br>
</font><br>-----BEGIN PGP SIGNATURE-----<br>
Version: GnuPG v1.4.10 (GNU/Linux)<br>
<br>
iQEcBAEBAgAGBQJNEGqeAAoJEIC1IG5rXEfqLC4IAJm4o8ernD3XLfHvzHO0ui77<br>
dp8Dy+zXYr9mJ5Mjfyz1CPjHOvuloRBlV5Gt20z872mtcPTedVMuk8w9fWlUo1RU<br>
lwWaMnZ4qHhwhWg5zWiDi/hqno5mpR6BPcVoT+pQyOJp2Z4jB/rRvkV3IkXt1Eb/<br>
TtLhKROxkiGUIa/uayzPGyb2KMjD+KrSn8Jny66vzrvetPpIUd3/A7txuJ4RHNMI<br>
MWx4zed8O9Bm4FR9z109Nr3umr89Lw4AcheAa17m91ZTP1DcFx+DfRTn2IlpFeb0<br>
6+fAB9svoC+K7+PJqnMo9XYp9COzCNWWN9fv34x30rIouH6DSRtOwOyJ9gihOGw=<br>
=Jmur<br>
-----END PGP SIGNATURE-----<br>
<br></blockquote></div><br><br clear="all"><br>-- <br>Stig Bakken<br>CTO, Zedge.net - free your phone!<br>
</div>