Roundcube suddenly crashed

[imc67]

Thanks to Netdata I got a critical warning that in the NGNIX logs more than normal errors where found. After looking into Cloudron it seemed two instances of Roundcube (latest version) where crashed with:

Nginx Error: Error reloading nginx: reload exited with code 1 signal null

In the box log I saw 2 unresponsive apps but got no single warning from Cloudron itself! Shouldn’t I have gotten an email?

Anyway retrying the last task (strange because I haven’t touched the apps!) it is solved.

Any idea what went wrong?

[warg]

Similar issue here: I never got any email for an app that was down short-term nor when healthcheck fails. Even when healthcheck failed, app was fine so this doesn't sound like it works properly.

[girish]

We don't have downtime email notifications . These have to come from some external monitoring setup. This comes to the yet untackled notifcation problem. We used to send it all admins back in the day, but this was considered back then, too much.

[girish]

@imc67 said in Roundcube suddenly crashed:

Nginx Error: Error reloading nginx: reload exited with code 1 signal null

This indicates nginx (the reverse proxy in front of all apps) had some trouble reloading and not roundcube itself crashing. Roundcube itseful internally uses php/apache. Did you see anything else in the nginx logs around this?