directadmin process to check task.queue to ensure it is being processed

Version 1.421


The task.queue relies on crond to call the dataskq binary every minute. If something is broken, the task.queue will grow and backup jobs won't get processed. Often times, the Admin does not know the task.queue isn't being processed. This change will have the directadmin process check for the task.queue. The internal default directadmin.conf setting: check_task_queue=2048 which means if the task.queue is more than 2048 bytes in size, and is more than 5 minutes in age, then the message can be generated. If you wish to disable this check, add the following to your directadmin.conf, and restart directadmin: check_task_queue=0 You can change the number in case a 2Kb task.queue, older than 5 minutes, is normal on your system. The fstat variable st_atime is used.. so just make sure you're not looking at the file, else the check will reset. We use st_atime because the modified time could be newer due to the append, so we want the oldest possible time related to the file (as the dataskq renames and deletes it) The check is trigged after a login by an Admin. The message repeat shares the service check interval in the admin.conf: service_email_last_sent= which is otherwise used to limit the notices if a service is down. This is shared because it's essentially monitoring if crond is down, but via other means.

Interested to try DirectAdmin? Get a 30-day Free Trial!