Discourse instance stops running in every midnight

Product Support

AI-generated summary

A Discourse instance, ‘app2’, stops running every midnight, but can be restarted with ./launcher rebuild app2. keithkhl initially thought backups were the cause, but the issue persisted after disabling them. NateDhaliwal suggested making it a multisite installation, linking to a multisite configuration guide, but RGJ questioned why this would resolve the issue, seeking further explanation.

1

I have recently copied a discourse instance to another folder in my server and assigned a new domain.

The new one has nearly identical settings, but runs with different logo, categories, and policy, but other than that, both are the same.

The old instance’s name is ‘app’, as for all ordinary Discourse instances, and the new one is ‘app2’. I have changed the ‘app.yml’ in /containers/ to ‘app2.yml’, and run rebuild, when I initially set up this copy website.

But the new site stops running in every midnight. The command ‘./launcer rebuild app2’ makes it working until the day’s midnight.

I initially thought it is due to backup, but even after I killed all backups, I still experience the same issue.

It could be Redis, as I can see that, inside the container, the processor cannot find redis.

I can certainly try re-installing from scratch or make the original one a multisite, but would like to find a way to resolve this.

2

I would make it a multisite installation. The processes terminate at the end of the day, but multisite should make them stay up.

3

:thinking: why would that make a difference?