locked
Consistently receiving HTTP error 408 using CRM 2011 (IFD) RRS feed

  • Question

  • Hi everybody,

    Since UR7 was applied, I get consistent HTTP errors 408 (time-out) while browsing the CRM 2011.

    My client setup is:

    - 2 physical servers : 1 with CRM 2011 (12Gb RAM) / 1 with SQL 2008 R2 + ADFS (16Gb RAM)

    - CRM is setup with IFD & Claims and can be accessed internally & externally without any problems (DNS are functioning OK)

    - 1 TMG server in front of both servers (not managed by us)

    - UR7 applied (CRM + E-mail router)

    The HTTP error 408 appears when browsing CRM 2011 and not everytime on the same screen! It can happen while editing account details, or while editing security roles, or while viewing system settings, etc.

    I have used Fiddler to search for the error but could not find anything.

    I have setup the CRM server to use WCF compression per this site (could this be the problem?) : http://bingsoft.wordpress.com/2012/02/20/optimise-dynamics-crm-2011-using-wcf-compression/ 

    Both servers are dual quad-cores so I'm not sure the WCF compression is the problem.

    Do you think that enabling TRACING would help? Can it be the ADFS server? The SSL certificate (wildcard) is correctly bound to both web sites (CRM & ADFS).

    Hope someone can help me, this error is driving us crazy!

    Olivier


    Currently blogging @ www.furnemont.eu
    Currently twittering @bugsoli
    Currently restoring my Mustang @www.mustangpassion.com

    Sunday, April 22, 2012 12:01 PM

All replies

  • I can't help resolve your issue, but I can say I am seeing this HTTP 408 error in an almost identical environment:

    - CRM 2011 UR7 on one server, SQL 2008 R2 on another server (both virtual as it happens, both server 2008R2)

    - IFD with ADFS (incidentally I'm pretty sure ADFS is not a factor here, as once the user is authenticated and issued a token, ADFS should have nothing to do until that expires and re-authenticates). Internal and external access working great (internal uses SSO based on split DNS).

    - TMG publishing this externally (not internally on the LAN, which seems to work fine)

    (This is a new environment built to do a swing migration from CRM4, and has been UR7 since day 1, so I can't comment on whether this would or would not have been the case in a previous rollup under the same conditions. It did not seem to happen for the first week or so, but now happens way too often for my comfort).

    other forum posts (mainly about OWA) seem to point to an option in the rule so "Request appear to come from the Forefront TMG computer"

    I get this most often with loading forms, occasionally dialogs and even loading views sometimes.

    Hope this helps. Adam Vero, MCT

    Monday, April 23, 2012 9:16 PM
  • Hello Adam,

    Thanks for the input, I'll check the TMG side today with my client and keep you updated. You are right about comfort: the product is barely useable and the thing is: before UR7, browsing was a snap and there were no errors whatsoever!

    Interestingly, while looking to the Event Viewer on the CRM server, I also found a lot of errors coming from a problem with the MUI language pack which is the same as this post: https://community.dynamics.com/product/crm/f/117/p/72523/144403.aspx

    So I might try this to see if it helps: http://social.microsoft.com/Forums/de-DE/crmdeployment/thread/d81133c7-0fba-4a7a-a5e4-0784ab2430ad

    The weird thing is that I don't have any language pack at all on this server, it's only English! Do you have also this error? This could be the problem but I don't see why, I'll try to provision the language again to see if it helps...

    Olivier


    Currently blogging @ www.furnemont.eu
    Currently twittering @bugsoli
    Currently restoring my Mustang @www.mustangpassion.com

    Tuesday, April 24, 2012 6:07 AM
  • Minor update:

    client changed TMG settings to request CRM to compress data, and performance improved significantly using IE9. Fiddler traces show much less content being passed back to the web client, this seems to be mainly because it is able to use cache where previously (for some unknown reason) it would fetch everything every time - scripts (75% of content!) ribbons, images etc.

    However, despite improved speed I still get 408s from time to time. It seems to be slightly less than before, but still too frequent. Overall I now seem to have better "bandwidth" to CRM but still suffer from latency being too high so I get these timeouts.

    I'll update as I continue troubleshooting.


    Hope this helps. Adam Vero, MCT

    Thursday, April 26, 2012 9:10 PM
  • Were you able to track this down? We have a client with the same issue.

    Grant Thompson

    Tuesday, October 16, 2012 8:50 PM
  • Hi Grant,

    As far as I recall, it had something to do with the physical firewall that was set up to cache SSL requests but this feature has a bug in it so the IT team has deactivated it and now everything is good, no more HTTP 408 since a couple of months!

    HTH,

    Olivier


    Currently blogging @ www.furnemont.eu
    Currently twittering @bugsoli
    Currently restoring my Mustang @www.mustangpassion.com

    Wednesday, October 17, 2012 7:20 AM