Having an issue that's very strange. There are 4 jobs each configured for two of my SQL servers, so 8 jobs. The Mon-Thu Full Database backup jobs run to one library partition, the friday jobs run to the other partition for weekly, offsite storage. There are also TLog jobs configured and those are set to backup and flush the logs later in the same overnight.
The issue is that the 4 TLog jobs do not run. The are scheduled to run tonight, just like they were last week. There is no lob jog or error message in BE or Wndows logs. It's almost like they get kicked off and just reschedule themselves for the next day.
One thing I noticed is that I cannot configure the job to be a Full backup of the logs. When I configure the job as full, the icon is a full orange cylinder, as expected. Then when I choose the log/truncate option, I see the icon change to a half orange cylinder. When I save the config, the icon changes to the quarter orange cylinder, showing it listed as an Inc job. If there's no Full job for an Inc job to refer to, wouldn't that cause this? I'm guessing at this point.
Since the issue is happening to both servers' job definitions, I know it's not some error or fluke, but I am perplexed.
v14.1 r1786(.1093) 64b
You cannot define a log backup job without a full backup, so showing the log backup job as an incremental job is correct. See my article below on how to define a log backup job.
You should post some screenshots of your job definitions.
2 screenshots attached, 1 shows schedule and 2 shows SQL settings. This is one of the 2 four-part jobs that have similar configs.
I did try re-running the TLog job manually and the job runs. How about that..?
Given this I could probably script a Start-BEJob command and call it when the full job completes. But that's not ideal.
Any other ideas after reviewing the screen grabs? Thanks for helping out
Why do you set your log backups to start 5 minutes after the full backups? Do your full backups finish so fast?
It could be that the log backups exceeded their wait time and is rescheduled automatically. You either schedule them to start later or increase the wait time (the option above the one circled in red in the screenshot below).
We schedule them that way, and they queue and run in order so they all have the same start date (this is a practice that started in BE2012 for reasons I'm not going to get into, but are related to the disasterous UI changes). That's not the issue; the setting you're curious about is set to reschedule after 24 hours, cancel box is unchecked. If the job missed the window, I would get a missed job alert and would see it in the job history as Missed. As I previously stated, that doesn't happen.