MID server thread count is exceeding

[chowdhary]

Hi,

After upgrading to helsinki patch3 I see that our mid server thread count is exceeding and also discovery takes more time to complete the schedules when compared to Fuji version.

I see the thread count is around 140 whereas in the config file we selected the max thread count as 100. I see the same threads multiple times. Could any one let me know why the thread count exceeds and what should be done not to exceed the mid server thread count as some times our schedules are getting cancelled.

Regards

Chowdhary

[Sharon Hobart]

I would lower the max.threads on your mid servers to around 75 and test.   This can be done under Configuration Parameters on your mid server in the GUI.   For the new number to go into affect, you would need to restart the mid server service.       Also, did you validate that your mid servers upgraded to the new version?   Also, depending on your VM configuration hosting your mid server, you should also consider going to the conf/wrapper-override.conf file and remove the "#" off the setting:   wrapper.java.maxmemory=1024.   restart the mid server service after making this change.     Assumption is that your VM is 64-bit.

[chowdhary]

Hi Sharon,

Thanks much for your response. Yes our mid servers are upgraded to new version and our VM is 64 bit. I will do the above changes in the parameters and see how it works.

Best Regards

Chowdhary

[chowdhary]

Hello Sharon,

After following the steps suggested by you the thread count came down but not sure after upgrading to Helsinki patch2 all our discovery schedules are getting cancelled. We have 5 Mid servers and we use only 2 for discovery and from all the mid servers I see the system command (cancel discovery) as in the screenshot(ECC queue) even though we use only 2 mid server for discovery. Finally the discovery gets cancelled with less than 50% success rate. In Fuji version we didn't face this issue and we were able to discover all the devices(around 90%).

Best Regards

Chowdhary