The sysdb code that adds the timestamp entries tried to be smart in the sense that it tried to create the timestamp entry along when the persistent cache entry is created and then just update the timestamp entry when then persistent cache entry is updated.
This breaks when some persistent cache entry already exists, but no timestamp entry exists, because the timestamp code would just try to update an entry which would fail with LDB_NO_SUCH_OBJECT (or similar).
We should catch this error and recover by adding the timestamp entry even during update.
owner: somebody => pcech
milestone: NEEDS_TRIAGE => SSSD 1.14.3
rhbz: => todo
Metadata Update from @jhrozek:
- Issue assigned to pcech
- Issue set to the milestone: SSSD 1.14.3
to comment on this ticket.
Copyright © 2014-2017 Red Hat
3.10.1 — Documentation