Pasted 9 months, 1 week ago — Expires in 82 days
Hi Allan, Benedict and JT, I'm running a server at home and a few small VPS VMs on Digital Ocean and Vultr and I would like to monitor them so I can fix them when they break. After a few incidents where my servers were down for days and I didn't know about it I setup Nagios on my server at home. It worked reasonably well but the configuration files were atrocious and everyone complains about how awful it is. Then I switched to Zabbix and it's easier to configure although it's through a Web GUI and this isn't stored in my puppet git repo like everything else. It's charting and collection of performance data is way better than Nagios. My problem however is that I run Zabbix at home and recently for some reason my network connection seems to have become less stable, so now enough of my Zabbix connections are dropped that Zabbix has become so noisy with false positives that I can't have the emails go to my inbox anymore which defeats the purpose of having monitoring. My home connection also doesn't support IPv6. I'd like to move my monitoring up to my cloud servers. I'm trying to keep my costs low so I use the smallest servers I can get and run everything in jails. My servers only have 512 MB of RAM and looking at the RAM usage on my zabbix jail at home MySQL itself is using over 600 MB of RAM. Perhaps I could get this down with some tuning, but Zabbix seems a bit too heavy to run on a 512MB VM, with other jails/services . I've started looking at Sensu but it seems a bit too complicated for my needs. I want something small and lightweight that will just email me when something is really down. I'll probably just go back to Nagios and maybe keep Zabbix around just for the CPU and Memory charts. What are you using for server monitoring? What are your opinions on monitoring tools? I've been looking around and there are a lot of tools and it's hard to figure out what the right one is. Thanks for the great show. Keep up the good work.