Microsoft Platform elements suddenly in timeout after DataMiner upgrade

117 viewsdma upgrade Microsoft Platform Microsoft Windows timeout Timeout issue
3
2 Comments

Hi Dojo,

we recently upgraded our DMS to version 10.6 CU6. Everything went smoothly, but after the upgrade being finished our elements representing the Windows environments of our DMAs, using the Microsoft Platform connector, all went into timeout.

I cannot find any reason for that:

  • all settings on the host machines have not been touched and are correctly established as before (double-checked; including user permissions, WMI/DCOM settings etc.)
  • the correct credentials are provided in the element settings
  • element restart as well as fumbling with the element connection settings ("Edit" menu) has not helped
  • an update of the connector to the latest version also has not improved anything

Surprisingly, all the data actually seems to be available from the Windows environments. In Stream Viewer we can see the data coming in, as well as in the elements' tables etc.: all values for all parameters seem to be live, and we can check trending graphs etc. It's just as if the timeout alarm is somehow laying on top of the element, despite everything appearing to be functional and navigable underneath.

Any thoughts from anyone? Thanks a lot! 🙂

Pedro Debevere [SLC] [DevOps Enabler] Answered question
Miguel Obregon [SLC] [DevOps Catalyst] commented

Hi Nils,
Could you let us know which version of the Microsoft Platform connector are you running?
Nils Hoven [DevOps Advocate] commented

Hi Miguel,
gladly: we've been on version 6.0.0.17, and have updated to 6.0.0.19 after the issue came up.

1 Answer

2
3.42K 1 Comment

Hi Nils,

I could reproduce the issue. The Microsoft Platform 6.0.0.19 connector defines a WMI action (id 50030) that has connection 200 defined. (This is the only action that uses connection 200, the other actions define connection 100.) This query fails and therefore the element goes (and stays) into timeout. The queries on connection 100 however succeed, which explains the behavior you observed: data comes in but the element is in timeout.

We suspect this to be a software issue that is related to a recent change that was done regarding WMI (RN 45851). A task has been created to further investigate and fix this.

Nils Hoven [DevOps Advocate] Posted new comment
Nils Hoven [DevOps Advocate] commented

Hi Pedro,
thanks a lot for this update and for confirming this seems to be a software issue! At least good to hear this is not a problem easily fixable from the place in front of the screen. 😉
So we'll wait for a fix and will ignore/mask the timeout alarms until then.
Just to confirm: the task you created is not visible for me, right (which is alright of course – just checking)?
BR
Nils
You are viewing 1 out of 1 answers, click here to view all answers.