I’m in the process of setting up ChirpStack on kubernetes (locally on minikube for the time being, going to a full fledged on-premises cluster later) by using docker-compose-env as a reference, as suggested in another post.
All components run as deployments, with the exception of PostgreSQL and Redis which run as statefulsets.
When I try to access the application server UI, everytime with the exception of twice I’ve been met with the error message pq: relation “user” does not exist (code: 2). After the successful logins I’ve tried to add a network server, but after some idle-time I got kicked back to the login page, where the same error message occurred again.
The logs from the application server also yields some errors:
msg=“sync remote fragmentation setup error” error=“select error: pq: relation “remote_fragmentation_session” does not exist”
msg=“sync remote multicast class-c session error” error=“select error: pq: relation “remote_multicast_class_c_session” does not exist”
msg=“fuota deployment error” error=“select error: pq: relation “fuota_deployment” does not exist”
msg=“send gateway ping error: get gateway for ping error: select error: pq: relation “gateway” does not exist”
And the network server logs yields the following error:
msg=“multicast scheduler error” ctx_id=53504d3d-9770-4a37-a13a-57a4900380fd error=“get multicast queue-items error: select error: pq: relation “multicast_queue” does not exist”
The database does in fact contain these, so I don’t know why both components complain about not finding them. The logs indicate that connecting to the DBs was successful, and I’ve verified that the DB users and tables are created.
I’d like to add that I’m quite new to both LoRa and kubernetes, so please bear with me here. Any help is greatly appreciated