I'm currently running jobs on job clusters and would like these jobs to time out after 168 hours (7 days), at which point a new job cluster will be assigned. This timeout is specifically to ensure that jobs don't run on the same cluster for too long, allowing a fresh new cluster to take over. However, I don't want to receive alerts for these timeoutsโonly for actual job failures. I've noticed that enabling failure alerts also triggers alerts for timeouts, which creates unnecessary noise. How can I configure the system to notify me only of job failures and not timeouts?