Active Alerts

Active alerts arrive without you having to look them up. They are very important, and usually represent system shenanigans.

 

Make sure you are receving alerts BEFORE:

• out of disk on main, catalog, and log mounts;
• memory goes into swap;
• a server is at max load.

 

Make sure you are receving alerts WHEN:

• an unexpected number of processes are spawned;
• any zope/zeo/apache/haproxy/any service instance shuts down or restarts;
• any directory that has queued files exceeds a certain limit. E.g. if you are using MaildropHost, make sure that your spool never gets larger than 20 files. It probably means something is erroring out somewhere and the service isn't running batches correctly. Same for any async processing.

 

 

VERIFY that your system defaults are working:

• you are receiving nightly logwatch reports;
• your cron jobs are reporting to you when finished where applicable (hint: knowing which zeodbbackups you need is a good thing to save).