Skip to Content
Former Member
Oct 16, 2009 at 03:14 PM

CRM 2007 IC webclientu2013 u201C400 session timed outu201D issues


<p>We have recently implemented CRM 2007. We are at Service Pack level 5. We are using Interaction Center functionality through the IC Web Client. The user logs into Enterprise Portal to authenticate, but upon clicking the link for CRM the Portal spawns a separate Internet Explorer session for the CRM IC Web Client. We are using Cisco CSS for load balancing across 20 CRM application servers.</p>

<p>We initially deployed the new CRM System in May 2009. Over the last several months we have been working through some startup issues. One error in particular that has been occurring has been an HTTP 400 error. This error would occasionally affect the user by producing the following error on their screen (see screenshot u2013 the message is u201C400 Session timed out u2013 please log in againu201D). The user would need to close out of their session and open a new one to recover from this error. This caused considerable user disruption.</p>

<p>SAP has told us that this issue may be caused by users creating multiple windows of CRM 2007. Every time a user clicks on the relevant tab in Enterprise Portal, CRM 2007 opens in a new window but in the same u201Cbrowser processu201D. SAP Note 1328629 states that 'CRM2007 is designed for 1 session per browser processu2019. We have been training our users not create multiple windows, but users like the flexibility of having multiple windows and we have found no way to enforce this in the system.

Note-although this is u201Ctechnically unsupportedu201D by SAP, we hear that many SAP customers use multiple browser windows successfully, but they donu2019t launch the IE sessions from the same link, therefore, there is 1 session per browser process. This is a complicated workaround for us from a user perspective - and, we need to find a way to generate a new portal window for them to get a new link to CRM. </p>

<p>SAP also pointed out that the Cisco CSS was sending traffic to another application server other than the one the user was on, when these timeouts happened. By working with Cisco and manipulating the CSS settings (increasing the sticky inactivity timer), we believe we have eliminated this particular phenomenon. The number of user complaints reduced drastically with this change. </p>

<p>We have implemented Coradiant TrueSight to monitor our HTTP traffic in this environment. Coradiant tracks server errors (like http 400) in our environment. While we are not certain that every server error logged correlates to a user-experienced timeout, these numbers provide a good u201Cbarometeru201D of the extent of user interruption. We have grown to rely on these numbers, as users do always not report every issue they get and we want to be proactive in addressing issues that impact user adoption. </p>

<p>While the number of server errors has gone down significantly after we made the CSS adjustments, we are still seeing several of them every hour in Coradiant. This type of error seems to be prevalent during user profile determination when the user is logging into the system. </p>

<p>We are looking for input from others who may have also observed this type of issue in their CRM implementation. Can anyone explain what is going on with these u201C400 errorsu201D in network traffic? Did you face such issues in your implementation, and how did you overcome them? Did you have to enforce u201Csingle CRM windowsu201D to avoid the issue?</p>

Edited by: Bijoy George on Oct 16, 2009 5:24 PM