tracker issue : CF-4206777

select a category, or use search below
(searches all categories and all time range)
Title:

Performance Monitoring Toolset - Unresponsive alert thresholds incorrect

| View in Tracker

Status/Resolution/Reason: To Fix//BugVerified

Reporter/Name(from Bugbase): Toby W. / ()

Created: 01/21/2020

Components: PMT

Versions: 2018

Failure Type: Incorrectly functioning

Found In Build/Fixed In Build: 2018 /

Priority/Frequency: Normal / All users will encounter

Locale/System: English / Win 2012 Server x64

Vote Count: 0

Problem Description:

If you set a value of threads to be busy before alert, it will notify you even if not all those threads have been busy for the allocated time.

i.e. 15 or more threads are busy for more than 60 seconds.

Only 1 thread was busy for longer then 60 seconds, there was a 15 other threads running for various times less then 60 seconds. This triggers the alert.

The old Coldfusion Monitor did this correctly and alerted only when 15 threads were all running for longer then 60 seconds.

Its painful as we could have 1 long thread but keep getting alerts because a lot of quicker running threads are also running.

The alert will only last for about 3-4 seconds when all the other threads finish.


Steps to Reproduce:

Setup alert for x amount or more threads are busy for more than x seconds. Trigger the alert with running threads

Actual Result:

Alert for unresponsive server.

Expected Result:

No alert until all threads running exceed the limit


Any Workarounds:

No

Attachments:

Comments:

Where do we find or get notified about updates to the PMT??
Comment by Toby W.
33259 | March 09, 2020 04:28:48 PM GMT
Where do we find or get notified about updates to the PMT??
Comment by Toby W.
33260 | March 09, 2020 04:28:53 PM GMT
I have noticed that the time doesn't apply at all to this. We seem to get email alerts for this threshold even if there are not threads running over 60 seconds.
Comment by Toby W.
33261 | March 11, 2020 11:17:51 AM GMT
Also noticed that the alert seems to be more related to the ART of an application then the overall server or running threads. I noticed 1 application which had an ART of over 60 seconds (which we expected) kept spamming the alert email - Another application has loads of threads running (i.e. more then 15) which must trigger the alert.
Comment by Toby W.
33262 | March 11, 2020 02:28:05 PM GMT