SLA false alert

Products

• iDash Workload Automation

Issue/Introduction

SLA FBS_WM_CDN_JGLSAV1S_STOP has received an at risk alert for the past 2 days at 2:37am and 2:46am, before the job has started running at approx 2:50am and finishing at 3:44am and 3:55am.

setup is as follows:
deadline: 8:30am
Start tracking: 360m
At risk: 240m
Run validity: 360m
Past hours: 6h

Alert details:
iDash:SLA ID=SLA_NAME_STOP^CPA-20220223-0830 : Deadline=2022-02-23 08:30 -0500 : SLA=SLA_NAME_STOP^CPA : Next forecasted completion:2022-02-23 05:18:Warning:240 Minutes

Please advise on why the alert is forecasting for 5:18am and receiving false alerts.

 

Environment

Release :

Component :

Resolution

You can go into the Admin Tool and create a runtime override for these jobs. iDash will use that value for forecasting, ignoring the calculated average runtime.

This is only as accurate as the time you enter and remains in effect until you delete it..

More drastically, we can delete the average runtime entries for these jobs from the database. iDash will start from scratch for these two jobs and begin building a new average runtime.

The first SLA may also alarm, as iDash will assume a 5 minute runtime for any job it needs to predict and does not have a calculated value for. If you again see long runs, they will again affect the forecast.