I requested an account for one user, when it is fulfilled on target system and task completed in Saviynt, its status is "manually provisioned". However, I realized the user should not have such account, so I requested to delete this account. the delete account task is generated but the provisioning job did not process it and hung there until application restart. Nothing shown in log entries and the task is pending for long time although I run the same job many times.
However, when I check target system, that account is actually deleted on target system. but Saviynt did not have any update and keep pending
This is a bug need to be investigated or what is correct behavior? Can I delete an account immediately after creation and before next import/reconcile?
No, it shouldnt be the case when account is newly provisioned the status is manually provisioned . ( in that time you can check in target to see actually if the account is created or not ).
Now if you want to delete it you can delete it and the status would be manually deprovisioned.
Also on which connector you are able to see this ? is this specific or for all connectors you are seeing this.
The behavior is on Jar Connector only. when the remove task generated and WSRetry job started, the job is hang there. it cannot stop until application restart. after restart, run the WSRetry job again, the task is failed status because it was actually deleted on target by the hang job already.
If its specific to jar connector then when you hit the delete/remove account the response may not come directly and system is still waiting for the response and may be in hung state hence the job would have been stuck.
Anyways you can open a FD ticket and our team will assist you further.
Thanks Darshan for your reminder. in our case, we use instant provisioning, all other operations such as create account, update account and change password is processed very quick and get it done in few seconds. I'll try to reproduce the issue and work with Support team.
Thank you very much!
Thanks Rushikesh. looks we need open a support ticket to see what the root cause is. the worst thing is job hang there. and no way to stop it unless restart the server/application which is not acceptable in prod env.