locked
Multiple Users Getting dropped at the same time RRS feed

  • Question

  • I have a issue where several users are being dropped all within the same second and they are all stated as "Reason for Removal: Administrators Request"  All the users are behind a Firewall that is proxying their addresses so they all show as having the same IP.  So I basically havew 2 questions:

    1) When a session times out, does UAG/TMG disconnect the IP address?
    2) What can cause the error message: The user XXX with source IP address 0.0.0.1 was removed from session xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx on trunk xxxx (secure=1). Reason for removal: Administrator Request.

    Tuesday, November 15, 2011 1:37 PM

Answers

  • We changed the default timout on this trunk from 60min to 24 hours and the users stopped complaining.  We also added access to this app from the UK, so latency was cut in half.  Both of these measures seemed to have resolved the issues.  I do not know why all the users in a single site where being dropped at the same time, I suspect it was the vendors internet circuit.  The application has been identified as one being sensitive to latency longer than 500ms. 

    • Marked as answer by JGBoake Monday, February 13, 2012 3:28 PM
    Monday, February 13, 2012 3:28 PM

All replies

  • Hi John,

    I think this happen when the administrator terminate the session from the web monitor using the "Terminate" button:

    Could it be someone from the UAG side played with this?

     

    EDIT: I just tested on my lab, and when pressing this "terminate" the reason is "User requested". So maybe it have to do with stopping the node via the "array managemnet" interface in the web monitor. I could not test right now, but if you are using array and stopped it via the web monitor, this may explain it.

    Ophir.


    Tuesday, November 15, 2011 2:56 PM
  • Hi Ophir,

    I am not using an array, yet.  The trunk they are using is a single application trunk.  The trunk has 2 applications, one is a Generic Client Application (Multiple Servers) the other is a web application.  The WebApp has a dependancy on the Generic Client Application and it all works just fine, except for this one small thing.  It makes sence that the IP is cleared after a session is terminated, I just need confirmation that this is indeed the problem.  If it is, I can simply extend the session timeout on this trunk, if it's not, I need to find out why it's happening.

     

    Tuesday, November 15, 2011 3:41 PM
  • Hi John,

    I tested it, and no matter what I do, the session is being terminate with "Reason: User request".

    I could not make it terminate with "Administrator request". When the session timeout, only the specific client should get disconnect (after a pop-up warn them about it) but the UAG should not terminate all users comming from same source-IP as it is very common for users to arrive via proxy server, etc..

    Do you see this often, or was this a one-time event?

     

    Ophir.

    Tuesday, November 15, 2011 5:18 PM
  • Seems to happen a couple times a day, but not at specific times.

     

    Tuesday, November 15, 2011 7:58 PM
  • Maybe you activate the configuration at the same time, or do other maintanence on the system at the same time? Do you have any events in the system/application that may hint what happen on that specific time ?
    Wednesday, November 16, 2011 11:00 AM
  • We are not doing anything on the UAG(occurs at odd hours at inconsistant times), we only do maintenance once a month on the UAG's, so it's not anyone messing around.  I don't see other events on the UAG, so it's not everyone on every portal(12 or 13/uag) getting dropped.  Just a select group of users in a clients office, where we have no jurisdiction on the network equipment at all.

     

    So a needle in a needle stack.

    Monday, November 21, 2011 4:35 PM
  • We changed the default timout on this trunk from 60min to 24 hours and the users stopped complaining.  We also added access to this app from the UK, so latency was cut in half.  Both of these measures seemed to have resolved the issues.  I do not know why all the users in a single site where being dropped at the same time, I suspect it was the vendors internet circuit.  The application has been identified as one being sensitive to latency longer than 500ms. 

    • Marked as answer by JGBoake Monday, February 13, 2012 3:28 PM
    Monday, February 13, 2012 3:28 PM