Today, if you have say 2 nodes configured behind a Nodebalancer, and one of those nodes fails its health check, the Nodebalancer will drop it from its list and stop routing traffic to it. Perfect.

But, it never bothers to tell anyone about this. No notifications are sent.

This seems like a pretty significant issue to me; I've opened a support ticket about this, but the response was basically "you should have some other monitoring solution". Very frustrating - this is a pretty basic requirement for a device that's sitting between me and my users, no? Ideally there would be lots of configuration available here - but initially, just the simplest thing would be great. For example - just notify me via email if one of my nodes changes status in the Nodebalancer.

I _could_ poll http://www.linode.com/api/nodebalancer/ ... ode%2Elist every 30 seconds, but that doesn't seem to be a good use of anyone's resources.

In Linode's defense, they did say they'd pass it along to their developers. But I'm hoping I'm not the only one who needs this...a groundswell of support here might help things along. So...anyone?

At my day job, it's the same story, the monitoring of a server is completely separate from the load balancer.

What you should be doing is polling the same page that the nodebalancer is, whatever that healthcheck page is.

It would be nice if the nodebalancer could send an email when something happens like the very useful reboot notifications or the other alerts.

But Linode are exactly right when they say "you should have some other monitoring solution". You really, really should for anything important. I recommend nagios.