The lsa session_key test does not cleanup secret objects correctly. The test is known to be failing if sealing is turned on in some circumstances. In this case a secret is created and then the function dcerpc_fetch_session_key() fails. The secret is not removed! We use torturesecret-%08x with random() to fill in the number. Sometimes it happens that random() returns a number we already used. So we end up trying to create a secret for an entry which already exists and run into a collision This change makes sure we always cleanup behind us and do not leave secret objects we created.
Created attachment 12687 [details] patch for 4.5
Created attachment 12688 [details] patch for 4.4
Re-assigning to Karolin for inclusion in 4.5.next, 4.4.next.
(In reply to Jeremy Allison from comment #3) Pushed to autobuild-v4-{5,4}-test.
(In reply to Karolin Seeger from comment #4) Pushed to both branches. Closing out bug report. Thanks!