Not thinking it through - Part 3
Continuing the saga of the index rebuild job, ( see "not thinking it through" and ditto part 2 ) the job has been rewritten into three steps
step 1 puts the database into simple recovery and disables tran log backups
step 2 does the index rebuild stuff
step 3 puts the database mode to full recovery and enables the transaction log backups.
Sadly the step sequence is that each relies on success of the previous so that putting the database back into full recovery requires success of step 2. Step 2 failed this weekend so the production database was left in simple recovery with disabled transaction log backups, which wouldn't have worked anyway as you can't make a log backup for a database in simple recovery!
The job sequence is also still missing a full backup after the database mode is returned to full, this job runs early morning but some 17 hours before the next scheduled full backup.
Once again the logic has not been thought through and a critical database left in a perilous state.
The saga continues.