Your "PMS" Postgres service crashed because its volume ran out of disk space - the volume is currently only 500 MB. You can resize it up to 5 GB (the Hobby plan maximum) from the volume settings in your project, then redeploy the service to clear the crash loop.

after changed pro version also still not able to connect DB 2026-06-17 05:37:58.542 UTC [2873] LOG: invalid record length at 0/B3C16800: expected at least 24, got 0

2026-06-17 05:37:58.542 UTC [2873] LOG: redo done at 0/B3C167C8 system usage: CPU: user: 0.00 s, system: 0.00 s, elapsed: 0.00 s

2026-06-17 05:37:58.551 UTC [2874] LOG: checkpoint starting: end-of-recovery immediate wait

2026-06-17 05:37:58.551 UTC [2874] PANIC: could not write to file "pg_logical/replorigin_checkpoint.tmp": No space left on device

2026-06-17 05:37:58.552 UTC [18] LOG: checkpointer process (PID 2874) was terminated by signal 6: Aborted

2026-06-17 05:37:58.552 UTC [18] LOG: terminating any other active server processes

2026-06-17 05:37:58.553 UTC [18] LOG: all server processes terminated; reinitializing

2026-06-17 05:37:58.572 UTC [2880] LOG: database system shutdown was interrupted; last known up at 2026-06-17 05:37:58 UTC

Your "PMS" volume is still 500 MB, even after upgrading to Pro. Upgrading the plan raises the maximum allowed size but does not automatically resize existing volumes. Go to the volume settings for your PMS service, click "Live Resize," and increase it to at least 1 GB (Pro allows up to 1 TB), then redeploy the service.

navigation steps to change?

In your project, click on the "PMS" service, then click the "Volumes" tab on the left sidebar. Select the attached volume, click "Live Resize," increase it from 500 MB to at least 1 GB, and redeploy the service afterward.