URGENT: LOAD.ASHX CAUSES MAJOR ISSUES WITH LOAD BALANCER

  •  01-09-2012, 7:31 AM

    URGENT: LOAD.ASHX CAUSES MAJOR ISSUES WITH LOAD BALANCER

    We run our dotnet solution using cutesoft in a load balanced configuration employing the use of an f5 load balancer network appliance.

    We recently upgraded to the 11-10-2011 build when we started experiencing problems experienced MAJOR problems with load.ashx not being properly returned to the client.  I upgraded this evening to the 12-20-2011 build of cutesoft editor, and the problem still persists.

    The resulting adverse behavior was that requests for resources through load.ashx were not received by the client, as a result of the load balancer aborting the request.

    Failures resulted from the failure of delivery of

    • Graphics files
    • Javascript files (disastrous implications)
    • css files

     As you can see – this affects pretty much all the content served through load.ashx, and is critical to the operation of cuteeditor on the client.

     Any failure to deliver content results in a completely unusable and unreliable HTML editor.

     We looked at our load balancer logs and noticed specific errors that are detailed here:

    http://support.f5.com/kb/en-us/solutions/public/5000/900/sol5922.html?sr=14345402

     Of note:

    Any response that is not compliant with RFC2616 will cause this error message, such as a web server response which includes a Content-Length header indicating a smaller value than the length of the data in the response. Another common cause of this error message is an extra HTTP 100 Continue message, which may be included in a POST response by Microsoft Internet Information Services (IIS) version 6.0.

     Have modifications been made to the underlying load.ashx code?  I’m finding the most recent builds are returning invalid HTTP headers on the response, which is causing our load balancer to discard and fail the resource requests.

     Can you outline what changes that have been made to writing out load.ashx?  I need to provide specific details to my technical team to discuss the implications so we can come up with the best fix to the problem in my live production system.

     
    We are  day 3 since the reporting and discovery of this problem, and I still have not heard back from you regarding this matter.

     Your response would greatly be appreciated.

View Complete Thread