Options

Monitoring DRBD / HA Cluster

jibbajabbajibbajabba Member Posts: 4,317 ■■■■■■■■□□
I run a Linux DRBD / HA cluster with several resources, which include IPs and disks.

Now I am a bit stuck here .. not with the setup itself that is, but how I monitor the setup.

At the moment I use an IP Monitor to monitor the individual nodes via IP, the VIPs and also using a test webpage / database using the VIP of apache.

Trouble with this is now if DRBD has a problem. If DRBD fails altogether, sure enough the monitoring for the VIPs (ping and http) fail.

However, I'd like to know if there is a failover. Lets assume the resources are now on the secondary node, all looks fine via the monitoring etc. but there is obviously a reason why the failover occured and I'd like to be notified when and if this is the case.

A "cheap" method I am using at the moment is touch the output of /proc/drbd into a webpage via a cron job and using our monitoring software to check for a string, in this case "Primary/Secondary" or "Secondary/Primary" respectively.

Is there a better, cleaner method to monitor DRBD ?
My own knowledge base made public: http://open902.com :p
Sign In or Register to comment.