Before labelling this as a bug; can you give more information:

• how many monitors are you running
• what type of monitors
• what are the system specs

I'm running roughly 100 monitors which are a mix of HTTP and PING. My server is a 4 core VM with 8GB of RAM.

On the same server I'm also running an instance of Observium which is monitoring 100+ devices through SNMP and it has no issues with performance.

Here is another screenshot from htop:

System shouldnt be a problem for 100 monitors.
Whats the frequency of your monitors?

20 seconds.

I'm not sure if this could be a contributing factor but I'm getting some weird graphs for my HTTP monitors.

You running it on a VM...? Then thats probably the issue?

20 seconds.

It maybe too low for 100 monitors, try to increase it to 60 or higher.

I've just finished changing them all to 60 seconds but it made no difference. I'm now going to clear the heartbeats and events as well.

After clearing all the historical data the CPU usage has come right down and will occasional peak above 10%.

My system seems to be stable again so I'm going to close this issue now...

My system seems to be stable again so I'm going to close this issue now...

I understand it became stable after cleaning the historical data, but that seems pretty weird. How did it behave after a few months? I'm having the same issue all of a sudden, but I didn't try cleaning anything yet.

In v1, reducing either retention/ping-rate/.. or moving to a faster storage solution is the only option to get around this problem.
In v2, we have likely resolved this problem. See #4500 for further details of what needs to happen before we can release this version.