My MySQL instance is stuck during startup with this error:

pgsql

CopyEdit

[ERROR] [MY-013379] [Server] Server upgrade started with version 80404, but server upgrade of version 80404 is still pending.

It looks like an incomplete data dictionary upgrade. I do not have filesystem or shell access to run recovery commands (e.g., innodb-force-recovery or manual upgrade finalization).

Could you help finalize or recover this upgrade? Alternatively, could you provide me with a data volume snapshot or raw data files so I can try to recover manually on another instance?

What did you do to the MySQL instance? Please share the exact steps you took.

I have the free plan, my instance exceed the limit of 500mb and stopped itself. I purchase the pro plan and try to rebuild the instance, but failed. When it try to rebuild I have this error logs
2025-06-19T20:46:26.101804Z 4 [System] [MY-013381] [Server] Server upgrade from '80404' to '80405' started.

2025-06-19T20:47:17.307531Z 4 [ERROR] [MY-013178] [Server] Execution of server-side SQL statement 'alter table db comment='Database privileges'; ' failed with error code = 3664, error message = 'Failed to delete SDI 'mysql.db' in tablespace 'mysql'.'.

2025-06-19T20:47:17.313834Z 0 [ERROR] [MY-013380] [Server] Failed to upgrade server.

2025-06-19T20:47:17.313895Z 0 [ERROR] [MY-010119] [Server] Aborting

2025-06-19T20:47:17.904278Z 0 [System] [MY-010910] [Server] /usr/sbin/mysqld: Shutdown complete (mysqld 8.4.5) MySQL Community Server - GPL.

2025-06-19T20:47:17.904295Z 0 [System] [MY-015016] [Server] MySQL Server - end.

You may need to modify the start command to force the upgrade since it looks pending

docker-entrypoint.sh mysqld --innodb-use-native-aio=0 --disable-log-bin --upgrade=FORCE

Here

https://railway.com/project/ba16fb93-58ed-455e-bbc1-664363cf910b/service/2adeb165-c202-4b9e-b300-232f2afdf296/settings?environmentId=6a2e9036-44be-4661-bf26-b2d4c6c4ed57#deploy-start-command

I have this same error logs after modify the start command and restart the instance

2025-07-08T07:44:12.671164Z 0 [System] [MY-015015] [Server] MySQL Server - start.

2025-07-08T07:44:12.951755Z 0 [System] [MY-010116] [Server] /usr/sbin/mysqld (mysqld 8.4.4) starting as process 1

2025-07-08T07:44:12.958028Z 1 [System] [MY-013576] [InnoDB] InnoDB initialization has started.

2025-07-08T07:44:13.653008Z 1 [System] [MY-013577] [InnoDB] InnoDB initialization has ended.

2025-07-08T07:44:13.661487Z 1 [ERROR] [MY-013379] [Server] Server upgrade started with version 80404, but server upgrade of version 80404 is still pending.

2025-07-08T07:44:13.661722Z 0 [ERROR] [MY-010020] [Server] Data Dictionary initialization failed.

2025-07-08T07:44:13.661763Z 0 [ERROR] [MY-010119] [Server] Aborting

2025-07-08T07:44:13.864406Z 0 [System] [MY-010910] [Server] /usr/sbin/mysqld: Shutdown complete (mysqld 8.4.4) MySQL Community Server - GPL.

2025-07-08T07:44:13.864422Z 0 [System] [MY-015016] [Server] MySQL Server - end.

Just to clarify:

. I purchase the pro plan and try to rebuild the instance, but failed. When it try to rebuild I have this error logs

All you did was re-deploy this, right?

Yes

What happens if you use this start command?

docker-entrypoint.sh mysqld --innodb-use-native-aio=0 --disable-log-bin --upgrade=NONE

Also, you've been changing the base image of the service multiple times. Please do not do that because you're switching between incompatible versions that may cause data corruption.

I have this error:

Mounting volume on: /var/lib/containers/railwayapp/bind-mounts/4851fd40-12e5-45c7-8824-30a9b78f7822/vol_27sggnpaupdhqp44

2025-07-14 08:43:38+00:00 [Note] [Entrypoint]: Entrypoint script for MySQL Server 8.4.4-1.el9 started.

2025-07-14 08:43:38+00:00 [Note] [Entrypoint]: Switching to dedicated user 'mysql'

2025-07-14 08:43:38+00:00 [Note] [Entrypoint]: Entrypoint script for MySQL Server 8.4.4-1.el9 started.

'/var/lib/mysql/mysql.sock' -> '/var/run/mysqld/mysqld.sock'

2025-07-14T08:43:38.896566Z 0 [System] [MY-015015] [Server] MySQL Server - start.

2025-07-14T08:43:39.155401Z 0 [System] [MY-010116] [Server] /usr/sbin/mysqld (mysqld 8.4.4) starting as process 1

2025-07-14T08:43:39.161330Z 1 [System] [MY-013576] [InnoDB] InnoDB initialization has started.

2025-07-14T08:43:39.770961Z 1 [System] [MY-013577] [InnoDB] InnoDB initialization has ended.

2025-07-14T08:43:39.771062Z 1 [ERROR] [MY-013377] [Server] Server shutting down because upgrade is required, yet prohibited by the command line option '--upgrade=NONE'.

2025-07-14T08:43:39.771073Z 1 [ERROR] [MY-010334] [Server] Failed to initialize DD Storage Engine

2025-07-14T08:43:39.771235Z 0 [ERROR] [MY-010020] [Server] Data Dictionary initialization failed.

2025-07-14T08:43:39.771257Z 0 [ERROR] [MY-010119] [Server] Aborting

2025-07-14T08:43:39.937177Z 0 [System] [MY-010910] [Server] /usr/sbin/mysqld: Shutdown complete (mysqld 8.4.4) MySQL Community Server - GPL.

2025-07-14T08:43:39.937193Z 0 [System] [MY-015016] [Server] MySQL Server - end.

🛠️ The internal ticket MySQL borked after redeploying has been marked as backlog.

I'm out of ideas on this. You changed the base image multiple times and that likely triggered some update in the database (especially if you changed a major version). Sorry we can't help you as this is application-specific issue that wasn't caused by us

✅ The ticket Database issue after redeployment has been marked as completed.