Help! Is this a bug? Jira Service Desk calculating first response time incorrectly

[Monica Morris]

I've got a custom service desk report that shows the average first support response time for issues submitted to the service desk.

When I view details for individual tickets on the report, there are several Elapsed Times that show up as WAY higher than what they actually are on the ticket.

Here's the most egregious example:

• User created the Issue March 16, 2022, 8:22 AM
• Support added a Comment March 16, 2022, 8:32 AM
• Service desk report shows Elapsed Time to first response as 155:03 instead of 0:10 - basically 6.5 days instead of 10 minutes

This is totally skewing my reporting and reduces the credibility of my entire dashboard!

Things I've already looked into that don't seem to be causing this problem:

• Support comment to User was a customer-facing comment, not an Internal Note
• User is in a user group covered by one of the SLA filters, and has submitted other tickets where the first response time is being calculated correctly in this same report
• Ticket status also changed at 8:32 AM, which should have triggered the SLA to stop counting even if JSD 'missed' the comment since ticket status changing is also one of the SLA rules
• Nothing happened on the ticket 6 days after it was created

Is there something I'm missing? Is this a bug that needs to be reported?

Thank you in advance for any replies!

[Monica Morris · Accepted solution]

ANSWER: the miscalculation I'm seeing is caused by this bug: https://jira.atlassian.com/browse/JSDCLOUD-7546 

Please vote for it if you think the fix would be helpful for your team!

[Joseph Chung Yin]

@Monica Morris -

Hi Monica:

Did you check your SLA setup for the JSM project in question? 

1) There are three conditions for the "Time to first response" SLA which has it's own separate conditions, and it's own start/stop/pause options. 

2) Make sure that the jql filter linked to the SLA is correctly configured so that the SLA is being applied to the correct issues

Hope this helps.

Best, Joseph Chung Yin

Jira/JSM Functional Lead, Global Infrastructure Applications Team

Viasat Inc.

[Monica Morris]

Hi Joseph,

Thanks for your reply!

I did check the SLA setup for this service project. The ticket in question met 2 of the 3 possible conditions for stopping the 'Time to First Response' SLA - a comment was posted, and the status was changed. It seems that neither of these triggered the SLA to stop on this ticket. Based on how the other first response times in this report are being measured (excepting the ones that seem really far off, which prompted this post), Jira should have stopped the SLA and counted the Elapsed Time as 10 minutes.

I've also confirmed that the JQL filter linked to this SLA is correctly configured - as I mentioned, I manually checked and most of the tickets in this report seem to be calculating correctly. The inconsistency is a big part of the reason I think this may be a bug.

Hope this is helpful context.

Monica

[Joseph Chung Yin]

@Monica Morris -

Did the SLA setup changed after the JSM issue in question was created and commented.  Meaning the issue was created prior to SLA was re-configured.  If so, then it explains the possible cause because SLA updates don't automatically applied against issues created prior to the changes.

If not, then I will recommend you contact Atlassian Support for further assistance. 

Best, Joseph

[Monica Morris]

Hi Joseph, no the SLA setup has not changed in quite a while. I've just submitted a bug report to Atlassian (JST-750877). Appreciate your help!

[Joseph Chung Yin]

@Monica Morris -

Thanks for the update.  Please let me know what Atlassian Support has to say about your reported issue and possible solution.

In the meantime, if you feel my responses assisted your ask, please click on Accept Answer when you have time.

Best, Joseph

[Monica Morris]

Hi @Joseph Chung Yin , following up on this - it looks like the miscalculation I'm seeing is caused by this bug: https://jira.atlassian.com/browse/JSDCLOUD-7546 

I've just posted another question around the specifics of the workaround described in the bug ticket, would very much appreciate any info you may have!

[Joseph Chung Yin]

@Monica Morris -

Thanks for the update.  Please feel to click on Accept Answer when you have a chance if my replies to this case have helped you.

Best, Joseph