Two different actions (one from a discovery job and one from a integration) were trying to apply changes on the same object on DB which lead to a deadlock.
UCMDB server restart was necessary to get the UCMDB up and running again.
This isn’t acceptable for a production environment, so UCMDB has to detect such a situation and solve it on their own, so that there isn’t the need for a restart and a system downtime.
There should be a way that UCMDB report such a situation or skip one action.
At the moment the only workaround is to reschedule the two jobs and hope that they won’t run at the same time in future which couldn’t be guarantee (after a restart or when a jobs take much more time than usual), so it could happen again.
Reference: QCIM1H125451 – ‘Write access is not possible due to deadlocks’
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.