>
> Hi Amos
>
> Answer to your questions
>
>> What exactly is this 504 error saying?
>
> - Timeout issue to the backend servers
> - The servers will be in an overload state because all the requests are being sent to the backend
>
>> And why are the 200 status when they do occur taking so very long to generate?
>
> - The load balancer will be switching from PRD to DR because of PRD overload state
>
> Our Setup
>
> -----> = traffic flow
>
> Apache / Squid ----------> F5 Load balancer ------> Origin servers (CMS System)
>
> This configuration allows us to Load Balance between PRD and DR servers (CMS)
>
> - When the Purge request is sent tp the Backend Origin servers the cms system goes into an overload state
> - The health checks on the Load balancer will fail and divert the traffic from PRD to DR backend (PRD and DR are replicated environments)
>
> Extra Info -- Purge Option
>
> - We Provide the University Staff a PHP page to clear objects of squid cache
> - The tool calls CURL and sends a Purge request
>
> Code
>
> $ch = curl_init();
> curl_setopt($ch, CURLOPT_URL, $value);
> curl_setopt($ch, CURLOPT_PROXY, $proxyserver);
> curl_setopt($ch, CURLOPT_CUSTOMREQUEST , "PURGE");
> curl_setopt($ch, CURLOPT_TIMEOUT, 5);
> curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
> curl_setopt($ch, CURLOPT_FRESH_CONNECT, true);
> curl_setopt($ch, CURLOPT_HEADER, true);
> curl_setopt($ch, CURLOPT_NOBODY, true);
>
> if (!curl_exec($ch)) { return false; }
> curl_close($ch);
>
>
>> What are the headers sent from the origin server to Squid?
>
> Cached Request from Local Workstation to Squid
>
> reidwatson:script rwat090$ curl -I http://www.test.auckland.ac.nz/uoa/
> HTTP/1.1 200 OK
> Date: Fri, 28 Jun 2013 02:21:45 GMT
> Server: Apache-Coyote/1.1
> Cache-Control: no-cache
> Pragma: no-cache
> Expires: Thu, 01 Jan 1970 00:00:00 GMT
> Content-Type: text/html;charset=UTF-8
> Age: 77
> X-Cache: HIT from webroute.test.auckland.ac.nz
> X-Cache-Lookup: HIT from webroute.test.auckland.ac.nz:3128
> Via: 1.0 webroute-test1.its.auckland.ac.nz (squid)
> Set-Cookie: webroute=server.; path=/;
> Vary: Accept-Encoding
> Set-Cookie: BIGipServer~Devtest~webroute-test-80_pool=310641930.20480.0000; path=/
>
> Request after PURGE -- Local Workstation to Squid then Backend server
>
> 1372386101.882 0 10.1.1.1 TCP_MISS/200 295 PURGE http://www.test.auckland.ac.nz/uoa/ - NONE/- -
>
> curl -I http://www.test.auckland.ac.nz/uoa/
> HTTP/1.1 200 OK
> Date: Fri, 28 Jun 2013 02:21:45 GMT
> Server: Apache-Coyote/1.1
> Cache-Control: no-cache
> Pragma: no-cache
> Expires: Thu, 01 Jan 1970 00:00:00 GMT
> Content-Type: text/html;charset=UTF-8
> X-Cache: MISS from webroute.test.auckland.ac.nz
> X-Cache-Lookup: MISS from webroute.test.auckland.ac.nz:3128
> Via: 1.0 webroute-test1.its.auckland.ac.nz (squid)
> Set-Cookie: JSESSIONID=40F3613FE4AC815A87CAC858EE38E112.jvm1; Path=/
> Set-Cookie: Jesi_Template_Username=guest; Path=/
> Set-Cookie: webroute=server.; path=/;
> Vary: Accept-Encoding
> Set-Cookie: BIGipServer~Devtest~webroute-test-80_pool=310641930.20480.0000; path=/
>
>
> I have also include a wire shark dump from the request after the purge -- From Backend server
>
> GET /uoa/ HTTP/1.1
> Host: www.test.auckland.ac.nz
> User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:21.0) Gecko/20100101 Firefox/21.0
> Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8
> Accept-Language: en-gb,en;q=0.5
> Accept-Encoding: gzip, deflate
> Cookie: __utma=18951208.362154167.1367379679.1372030825.1372114069.43; __utmz=18951208.1367379679.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none); __utma=265832586.969691744.1368401175.1372300411.1372385456.16; __utmz=265832586.1372193146.14.3.utmcsr=iam.test.auckland.ac.nz|utmccn=(referral)|utmcmd=referral|utmcct=/profile/SAML2/Redirect/SSO; __utmc=18951208; BIGipServer~Devtest~webroute-test-443_pool=327419146.47873.0000; webroute=server.; __utmc=265832586; _shibsession_63656e7472616c2d74657374687474703a2f2f7777772e746573742e6175636b6c616e642e61632e6e7a=_298760cf8dc4f7d5ae29a97b3febbf5f; JSESSIONID=F81F7E49D3DE061EE0358E77B9D604EC.jvm1; Jesi_Template_Username=guest; BIGipServer~Devtest~webroute-test-80_pool=327419146.20480.0000; __utmb=265832586.4.10.1372385456; __atuvc=2%7C26
> X-Forwarded-Protocol: http
> X-Forwarded-User:
> X-Forwarded-Host: www.test.auckland.ac.nz
> X-Forwarded-Server: www.test.auckland.ac.nz
> Via: 1.1 webroute-test1.its.auckland.ac.nz (squid)
> Surrogate-Capability: unset-id="Surrogate/1.0 ESI/1.0"
> X-Forwarded-For: 130.216.168.2, 127.0.0.1
> Cache-Control: max-age=259200
> Connection: keep-alive
>
> Cheers
>
> Reid
>
>
> On 27/06/2013, at 2:34 AM, Amos Jeffries <squid3@xxxxxxxxxxxxx> wrote:
>
>> On 25/06/2013 5:51 p.m., Reid Watson wrote:
>>> Hi,
>>>
>>> We seem to be having a strange issue with SQUID caching an object once it has been purged - this issue does not occur with all objects within the cache..
>>>
>>> Quick Overview
>>>
>>> - Four Apache Server with Squid Installed
>>> - RedHat 5.7 i386
>>> - squid-3.1.8-1.el5
>>>
>>>
>>> Issue
>>>
>>> - Admin sends a Purge request simultaneously to all four server
>>>
>>> Example
>>> 1372113766.459 0 10.10.10.10 TCP_MISS/200 280 PURGE http://www.auckland.ac.nz/uoa/ - NONE/- -
>>>
>>> - Squid seems to fail on caching the object
>>> - All requests are sent to backend servers
>>>
>>> Server 1
>>> 1372113766.459 0 10.10.10.10 TCP_MISS/200 280 PURGE http://www.auckland.ac.nz/uoa/ - NONE/- -
>>> 372113766.779 23 10.10.10.3 TCP_MISS/504 5063 GET http://www.auckland.ac.nz/uoa/ - NONE/- text/html
>>
>> What exactly is this 504 error saying?
>>
>>> 1372114043.139 106378 127.0.0.1 TCP_MISS/200 9891 GET http://www.auckland.ac.nz/uoa/ - FIRST_UP_PARENT/vip-dr text/html
>>> 1372114043.303 10112 127.0.0.1 TCP_MISS/500 24731 GET http://www.auckland.ac.nz/uoa/ - ANY_PARENT/vip-dr text/html
>>> 1372114105.918 0 10.10.10.3 TCP_MISS/504 4591 GET http://www.auckland.ac.nz/uoa/ - NONE/- text/html
>>> 1372114106.414 57182 127.0.0.1 TCP_MISS/200 69995 GET http://www.auckland.ac.nz/uoa/ - FIRST_UP_PARENT/vip-dr text/html
>>> 1372114106.419 55771 127.0.0.1 TCP_MISS/200 45853 GET http://www.auckland.ac.nz/uoa/ - ANY_PARENT/vip-dr text/html
>>> 1372114106.436 17387 127.0.0.1 TCP_MISS/200 69967 GET http://www.auckland.ac.nz/uoa/ - FIRST_UP_PARENT/vip-dr text/html
>>> 1372114106.490 60516 127.0.0.1 TCP_MISS/200 70115 GET http://www.auckland.ac.nz/uoa/ - FIRST_UP_PARENT/vip-dr text/html
>>> 1372114106.707 58628 127.0.0.1 TCP_MISS/200 70115 GET http://www.auckland.ac.nz/uoa/ - ANY_PARENT/vip-dr text/html
>>
>> And why are the 200 status when they do occur taking so very long to generate?
>>
>>> - if you send a reload command (/sbin/service squid reload) the object is stored within the cache instanlty
>>
>> What are the headers sent from the origin server to Squid?
>>
>> Amos
>>
>
