Cronjob of Surfer App stopped?

[Kubernetes]

Hello,

one of my Apps (Surfer App) does have a cronjob configured to execute every minute a https "ping" to healthcheck.io

* * * * * curl https://hc-ping.com/XXXXXXXX-0ad4-XXXXXX-bb0a-3XXXXXXXXX30

That was running a few weeks without any problems, but since yesterday it seems the cronjob is not executed anymore.

Are any issues similar to this known? I didn't really change anything - you know

After rebooting the Cloudron Server the cronjob seems to work again.

Best,
Michael

[girish]

For good measure, maybe you should provide some curl timeouts. Maybe it is getting stuck? Currently, the logic will just not run cron again if the previous one did not complete.

I think --connect-timeout and --max-time

[girish]

In that cron job, you can also prefix something like echo "==. Pinging healthcheck.io" && curl xxx . This way it will generate that statement in app logs.

[Kubernetes]

@girish maybe the reason, I don't see any timeouts between last successfull ping and the first missing one. But I will follow your recommendation and put in an echo for logging. Thanks!

[Kubernetes]

@girish is there a way to restart the Cronjob "Manager" without rebooting the server? It happend again this morning that the Cronjob stopped running. In the logs I see a last echo, but no "OK" response after curl command execution. Still not sure why that happening.

However I think it would be good to get a Notification if Cloudron Cronjob decides to stop Cronjobs for Containers?

[girish]

@Kubernetes systemctl restart box should restart the cron jobs as well. Server reboot is not required.

[girish]

I have added a debug message now if a cron job is running too long - https://git.cloudron.io/cloudron/box/-/commit/c978e3b7ea60f86f3629cf4550f89924d5ead2c1 . Maybe you can apply this on your server to test what's going on.

[Kubernetes]

Great, I have applied the debug message and restarted the box. I will let you know when the problem reoccurs.

[Kubernetes]

@girish
15 minutes ago the last ping arrived at healthcheck, but the logs show the last check misses an OK and there is no debug message?

2024-02-09T20:32:57.000Z Send Healthcheck
2024-02-09T20:33:57.000Z OK
2024-02-09T20:33:57.000Z Send Healthcheck
2024-02-09T20:34:57.000Z OK
2024-02-09T20:34:57.000Z Send Healthcheck
2024-02-09T20:35:57.000Z OK
2024-02-09T20:35:57.000Z Send Healthcheck
2024-02-09T20:36:57.000Z OK
2024-02-09T20:36:57.000Z Send Healthcheck
2024-02-09T20:37:57.000Z OK
2024-02-09T20:37:57.000Z Send Healthcheck
2024-02-09T20:38:57.000Z OK
2024-02-09T20:38:57.000Z Send Healthcheck

[girish]

So, "Send Healthcheck" and "OK" are printed before and after your cron ? If that's the case, looks like it didn't complete. If you do docker inspect , you will find a container there like crontab.0 or something which is your container. You can exec into it and ps and check if that curl is stuck.