Resolve Time Measures

[Drew Davies]

Hi all I'm hoping someone will be able to provide guidance on this one. I'm looking to create a Widget displaying the Average Resolution time of our Incidents. Using the demo data I copied the measure that the widget would be using as it's data source and believe it is working.

The issue however is that the average time that it has calculated for November is drastically higher than expected. Due to this, I am guessing that the timer counts the seconds between 'h_datelogged' and 'h_fixtime' without taking our working hours or 'on hold' statuses into account? 

Have I got the wrong impression with this or is there a way to make it take the above into consideration?

 

Summary: I'd like to know the Average amount of time (using the SLA counter that is paused when out of office hours or on hold) an Incident was either 'New' or 'Open' before it was resolved.

 

Here is the Measure I've got:

 

Thanks in advance for any help!

 

Drew

[Victor]

41 minutes ago, Drew Davies said:

I am guessing that the timer counts the seconds between 'h_datelogged' and 'h_fixtime' without taking our working hours or 'on hold' statuses into account? 

@Drew Davies no, not really...  'h_fixtime' actually represents the number of seconds it took from the moment the fix/resolve timer was created (which could be the same as h_datelogged but not necessarily) and the moment the fix/resolve timer was stopped (which, again could be the same as h_dateresolved but not necessarily). This "fix time" is calculated based on service desk calendar (business hours) and excludes on hold periods... so as you would need it to do.

When I said that date/time when start and stop of fix timers trigger might not necessarily coincide with h_datelogged and h_dateresolved is because the timers are actually managed by "Start" and "Stop" timer nodes in the BP process which, depending on configuration, can trigger at various points during request lifetime...

 

46 minutes ago, Drew Davies said:

the average time that it has calculated for November is drastically higher than expected

Is there any possibility that the numbers are actually correct? Maybe we/you can have a report of all requests resolved in Nov and see from the list if the numbers are correct or not?

[Steven Boardman]

@Drew Davies if you are using an average based measure, and then creating a widget from this, you can convert the seconds into minutes / hours using the Data Format option on your widget - in this case a Measured Sample, using the average measure. 

[Drew Davies]

Hi both thank you for the replies, the 'average' time that it's calculating is over 500 hours. If this was 500 of our business hours that equates to over 2 months and we've only been live ~6 weeks! Checking the last month opened vs resolve dates I'd say our average is closer to 10 hours or so.

 

I think I might try to filter the measure by a specific user for now to look for a reason as to what's causing the discrepancy