OLS flapping/down alerts via Prometheus

bocagua

Hi everyone,

I use Prometheus to monitor my Enhance VPSs. I’m finding that on several servers, I keep getting alerts that OLS is down, but it recovers immediately.

Is anyone else experiencing this? I haven’t received any complaints from clients yet, but it seems like OpenLiteSpeed is continuously failing or restarting.

Thanks!

SharedGrid

I think OLS reloads every time there is an update to the .htaccess file, so this may explain what you're seeing.

Adam

Reload of OLS is graceful so it shouldn't cause any downtime. OLS will also reload whenever a virtualhost is changed, this is normal behaviour.

If you run systemctl status openlitespeed, has the systemd unit restarted? Do you have any 3rd party addons which might be restarting openlitespeed?

bocagua

Thanks for the reply! You nailed it.

I checked the status and the service has been active for 6 days, so it is indeed not crashing. However, looking at the logs, I see very frequent reloads (sometimes multiple times within a minute).

It seems Prometheus is picking up these "Graceful Reloads" as downtime. Here is the output:

Active: active (running) since Fri 2026-01-23 12:56:22 UTC; 6 days ago ... Jan 30 09:04:15 server1-1 systemd[1]: Reloaded OpenLiteSpeed HTTP Server. Jan 30 09:04:20 server1-1 systemd[1]: Reloading OpenLiteSpeed HTTP Server... Jan 30 09:04:23 server1-1 systemd[1]: Reloaded OpenLiteSpeed HTTP Server. Jan 30 09:06:06 server1-1 systemd[1]: Reloading OpenLiteSpeed HTTP Server... Jan 30 09:06:09 server1-1 systemd[1]: Reloaded OpenLiteSpeed HTTP Server. Jan 30 09:06:14 server1-1 systemd[1]: Reloading OpenLiteSpeed HTTP Server...

Is this frequency normal? I understand OLS reloads when a vHost changes, but I am not making changes or adding domains at this rate. It seems to be reloading every few seconds on its own.

I'm also using cpfence

Since these reloads are happening so close together, my monitoring check likely coincides with the reload process, triggering the alert.

Thanks for the help!