What are the steps to problem solve when all I can get are 503s?
Jacques
whshub at gmail.com
Fri Jul 9 17:18:20 CEST 2010
I'm not currently using a vcl, just configuring via the command line for
simplicity. Adding the connect_timeout parameter below had no impact on the
503s. Nor did adding all the parameters listed at letsdugg.com.
Communication with the external server works perfectly. (Telnet and via
lynx.) It's actually the varnish-cache.org site just to introduce a minimum
number of variables.
The only thing that I see in the varnish log is the FetchError, no backend
connection.
Here is a startup, request, shutdown cycle of varnish log:
# /opt/extra/bin/varnishlog
0 WorkThread - fffffd7ff9dc0d20 start
0 CLI - Rd vcl.load "boot" ./vcl.ORk8t3RP.so
0 CLI - Wr 200 Loaded "./vcl.ORk8t3RP.so" as "boot"
0 CLI - Rd vcl.use "boot"
0 CLI - Wr 200
0 CLI - Rd start
0 Debug - "Acceptor poll space increased to 512"
0 Debug - "Acceptor is poll"
0 CLI - Wr 200
0 WorkThread - fffffd7ff93c5d20 start
0 WorkThread - fffffd7ff91c6d20 start
0 WorkThread - fffffd7ff8fc7d20 start
0 WorkThread - fffffd7ff8dc8d20 start
0 WorkThread - fffffd7ff8bc9d20 start
0 WorkThread - fffffd7ff89cad20 start
0 WorkThread - fffffd7ff87cbd20 start
0 WorkThread - fffffd7ff85ccd20 start
0 WorkThread - fffffd7ff83cdd20 start
0 CLI - Rd ping
0 CLI - Wr 200 PONG 1278687217 1.0
0 CLI - Rd ping
0 CLI - Wr 200 PONG 1278687220 1.0
9 SessionOpen c 10.100.200.1 64024 :80
9 ReqStart c 10.100.200.1 64024 1455837036
9 RxRequest c GET
9 RxURL c /
9 RxProtocol c HTTP/1.0
9 RxHeader c Host: 10.100.200.2
9 RxHeader c Accept: text/html, text/plain, text/css, text/sgml,
*/*;q=0.01
9 RxHeader c Accept-Encoding: gzip, compress, bzip2
9 RxHeader c Accept-Language: en
9 RxHeader c User-Agent: Lynx/2.8.6rel.5 libwww-FM/2.14 SSL-MM/1.4.1
OpenSSL/0.9.8l
9 VCL_call c recv
9 VCL_return c lookup
9 VCL_call c hash
9 VCL_return c hash
9 VCL_call c miss
9 VCL_return c fetch
9 FetchError c no backend connection
9 VCL_call c error
9 VCL_return c deliver
9 Length c 488
9 VCL_call c deliver
9 VCL_return c deliver
9 TxProtocol c HTTP/1.1
9 TxStatus c 503
9 TxResponse c Service Unavailable
9 TxHeader c Server: Varnish
9 TxHeader c Retry-After: 0
9 TxHeader c Content-Type: text/html; charset=utf-8
9 TxHeader c Content-Length: 488
9 TxHeader c Date: Fri, 09 Jul 2010 14:53:42 GMT
9 TxHeader c X-Varnish: 1455837036
9 TxHeader c Age: 0
9 TxHeader c Via: 1.1 varnish
9 TxHeader c Connection: close
9 ReqEnd c 1455837036 1278687222.201183081 1278687222.201552391
0.000498056 0.000254393 0.000114918
9 SessionClose c error
9 StatSess c 10.100.200.1 64024 0 1 1 0 0 0 235 488
0 CLI - Rd ping
0 CLI - Wr 200 PONG 1278687223 1.0
0 CLI - Rd ping
0 CLI - Wr 200 PONG 1278687226 1.0
0 CLI - Rd ping
0 CLI - Wr 200 PONG 1278687229 1.0
On Fri, Jul 9, 2010 at 12:58 AM, Jorge Díaz <jdzstz at gmail.com> wrote:
> ¿There is communication between Varnish host and 87.238.47.204 machine??
>
> Try first making " telnet 87.238.47.204 80 " in order testing there is no
> firewall or a rule configured in webserver.
>
> If comunication is fine, the problem must be in VCL configuration. Try
> executing varnishlog while you are testing and look for errors in output.
>
> In Solaris is also important to run Varnish with proper parameter that are
> listen in:
>
> - http://letsgetdugg.com/2009/12/04/varnish-on-solaris/
>
>
> I had a lot of comunication problems without parameter:
>
> * -p connect_timeout=0s # Important bug work around for Solaris*
>
>
>
>
>
> 2010/7/9 Bedis 9 <bedis9 at gmail.com>
>
> Hey,
>>
>> Can you share your VCL conf?
>>
>> cheers
>>
>> On Fri, Jul 9, 2010 at 7:08 AM, Jacques <whshub at gmail.com> wrote:
>>
>>> I am trying to use Varnish on OpenSolaris. Due the help of others, I
>>> have versions compiling and not crashing that are based on three different
>>> versions of Varnish.
>>>
>>> I'm using 4516, 2.1.2 with the 649 patch, and trunk. I've also tried
>>> compiling with gcc 4.3.3 in addition to the OpenSolaris standard of 3.4.3.
>>> That changed nothing.
>>>
>>> All compiled copies of varnish seem to start just fine. I can use
>>> varnishlog and see my requests arriving at Varnish.
>>>
>>> However, responses are always 503 with " FetchError: no backend
>>> connection". Using tcpdump, I see no calls to the backend server from
>>> Varnish, ever.
>>>
>>> For simplicity, I'm using the basic start of:
>>>
>>> varnishd -a :80 -b 87.238.47.204:80
>>>
>>> What are right steps to determine what is wrong? I see nothing in
>>> syslogs. Varnish doesn't complain. Using all the same steps on ubuntu, I
>>> get exactly what I expect, a cached version of the backend server.
>>>
>>> Any pointers about how I should figure out what is failing would be
>>> helpful.
>>>
>>> Thanks,
>>> Jacques
>>>
>>> I've also noticed that the cache is starting at 500GB. Changing the
>>> cache size to 512MB didn't have any impact on the 503s.
>>>
>>> _______________________________________________
>>> varnish-misc mailing list
>>> varnish-misc at varnish-cache.org
>>> http://lists.varnish-cache.org/mailman/listinfo/varnish-misc
>>>
>>
>>
>> _______________________________________________
>> varnish-misc mailing list
>> varnish-misc at varnish-cache.org
>> http://lists.varnish-cache.org/mailman/listinfo/varnish-misc
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.varnish-cache.org/pipermail/varnish-misc/attachments/20100709/ebe42089/attachment-0001.html>
More information about the varnish-misc
mailing list