Colleague Maintenance

Summary: Once a month Colleague runs OS updates from 2AM to 7AM. The Redshift extract of Colleague starts at 12AM, so there is technically a chance that if the extract job runs past 2AM some tables in stage_colg_dbo will fail to update. This has not happened yet, however this would mean that whatever tables in Redshift related to the tables in stage_colg_dbo would be a day behind of where we would normally expect them. This would then propagate to reports that pull data from Redshift. The solution is to start the load again after 7AM from where it failed and refresh any affected reports.

Actions: In the event that Colleague staging tables are not loaded into Redshift correctly, the prod load should be resumed at or after 7AM from where it failed. Then the PowerBI Dataflows and Reports that were affected should be refreshed.

Impact: Any Redshift tables based on the colleague tables that failed to load would be missing data from the previous day until the load can be resumed. As a result, Power BI Dataflows and Reports based on those Redshift Tables would also be missing data from the previous day.

Risk: This scheduled maintenance has not been a problem in the past, but the load was moved from starting at 1AM to 12AM to give us more breathing room there.