"ovirt-master_change-queue" job is failing due to timeout, which occurs when waiting for the postfix service on the slave to start.
The mentioned error seems to be around these lines in the job log:
This does not seem to be the job timing out, instead, it seems like the slave got disconnected.
The error seems to have nothing to do with the CQ code as its failing during global_setup,sh.
Moving the ticket over to as this seems to be an infra issue with the slave, though I doubt much information about this could be recovered at this point.
3 months old, if the same issue will reappear we'll debug it.