Ok, I’ve used the process management software, monit, since at least 2004, and it is simply an indespensible tool in my sysadmin cache. Basically it watches a process, say like Apache, and restarts it if it dies. But wait, that’s not all, it does tons of other things. Want it to watch it and restart it at a certain time? Sure. How about if it uses 50% of system memory in 5 cycles (cycles are checks, 120 seconds by default)? Yep, it’ll take care of that. How about watching a file and stopping a service and/or issuing an alert by email or web if the file’s UID, permission, or whatever has changed? No problem. Disk space is greater than 90% on one partition you want an email to go out to the admin? Easy. Seriously, once you start using monit you’ll be amazed at what you can cover, it’s truly one of the best tools I’ve ever used, and of course it’s GPL’d open source.
So, this week we had an issue where a some of our sites were down, and the monitor that watches them were internal to our network, and relied on some of the same resources; which is lees than ideal. I have a remote server running at one of our partner’s sites, so it’s the perfect canidate to watch our sites from a ‘real world’ view.
Home