Planning Analytics with Watson

Expand all | Collapse all

IdleConnectionTimeout difference in functionality

  • 1.  IdleConnectionTimeout difference in functionality

    Posted Tue June 22, 2021 04:39 AM

    Our TM1S.CFG previously used IdleConnectionTimeout=3600. We noticed that the name of this parameter has changed to IdleConnectionTimeoutSeconds - a perfectily reasonable change so that it is obvious that the units are in seconds. However, the functionality also appears to have changed.

    With the older parameter, we get logged out, but when we go back in, there is a CAM dialog box and if we log back in, then we can carry on with our session.

    With the newer parameter, not only are we logged out, but our session is also destroyed. This is particularly annoying in development, when you go off for a meeting and come back after an hour to find that you need to log back in and re-open all the views and processes that you were working on.

    Ideally there should be two parameters, one to log out but keep session alive and one to terminate the session. The latter is more likely to be useful in Production, and the former more useful in Non-Production.


    Paul SImon

    Paul Simon

  • 2.  RE: IdleConnectionTimeout difference in functionality

    Posted Thu June 24, 2021 12:59 PM
    Hello Paul,

    I personnaly use IdleConnectionTimeOutSeconds=0 to be "never" disconnected (until some unknow error happend every 2-3 days)
    because I agree it is really annoying to be disconnected like that. I am pretty sure this is not recommended but IBM support didn't say it was not when I asked.