We have a job exists to give us success/fail of all backups that have occurred in the environment. It polls the api every hour for all jobs that have an end time > our last known end time. It appears that backuptasksessions do not get committed to the API in the order that they end.
My question is..
1. Is this just a timing issue where it take a few seconds longer for some jobs to commit to the database than others, so they are just mildy out of order?
2. Is this the result of backuptasksessions not flushing to the database til the parent backupsession is completed?
If it is number 1, then I can just overlap my queries by an hour to resolve this problem. Though I will have to begin checking against existing records to throw out dupes.
If it is number 2, then I have to change my logic to grab backupsessions that have ended since our last known endtime.
If it is using some other behavior to commit backuptasksessions please let me know the logic so i can plan appropriately.
- Mike D