Self-hosted update process

[mitodrummer]

Hello team Twenty!

I've been running a self hosted instance of twenty for the past couple months. Since version 1.3.0.

I just updated from 1.4.0 to 1.6.7 but am seeing a largely blank CRM when i log into a workspace.
e.g.

The only commands I run after updating my fork is:

• nvm use
• yarn
• npx nx run twenty-server:database:migrate:prod
• npx nx start

The DB migrations seemed to complete no problem, and the server starts up fine locally, but it feels like i'm missing some steps or there is an issue migrating from 1.3.0 to 1.6.7

TY!

[mitodrummer]

Only error I can see in my log is:

[mitodrummer]

Could this be a result of skipping over a release? e.g. not migrating to 1.5.0 and going straight to 1.6.7?

[mitodrummer]

I reset things back to my 1.4.0 version, to try and migrate to 1.5.0 first.

I also just became aware of this command: npx nx run twenty-server:command workspace:sync-metadata which I did not run after migrating from our initial deployment of 1.3.0 to 1.4.0. Running it now against my live 1.4.0 deployment did succeed, but I wonder if workspaces created after the move to 1.4.0 (prior to running the metadata sync for 1.4.0) could be in a state that causes the sync for 1.5.0 to fail (as seen below):

When i do npx nx run twenty-server:command workspace:sync-metadata on 1.5.0 release, I am seeing many errors:
e.g.
[Nest] 81107 - 2025-09-24, 12:43:40 p.m. ERROR [SyncWorkspaceMetadataCommand] Error in workspace f63bb5d5-406e-4bf7-a470-f099e0b86d1a: Cannot convert RELATION to column type.

Our setup has many workspaces created, and we have live customers using it atm, so it's imperative I find an upgrade path for this, otherwise it could result in us abandoning twenty, which I'd like to avoid.

[mitodrummer]

I assume this error is coming from field-metadata-type-to-column-type.util.ts . It seems as if type RELATION isn't being handled here. I have customized the opportunity object to have an extra relation to a person, but not sure if that is what is blowing this up.

[mitodrummer]

September 24, 2025, 22:23 | Error: Could not run migration aborting | server
-- | -- | --
September 24, 2025, 22:23 | at UpgradeCommand.runMigrationCommand (/app/packages/twenty-server/dist/src/database/commands/upgrade-version-command/upgrade.command.js:142:19) | server
September 24, 2025, 22:23 | at process.processTicksAndRejections (node:internal/process/task_queues:105:5) | server
September 24, 2025, 22:23 | at async UpgradeCommand.run (/app/packages/twenty-server/dist/src/database/commands/command-runners/migration.command-runner.js:43:13) | server
September 24, 2025, 22:23 | at async Command.<anonymous> (/app/node_modules/nest-commander/src/command-runner.service.js:162:24) | server
September 24, 2025, 22:23 | at async Command.parseAsync (/app/node_modules/nest-commander/node_modules/commander/lib/command.js:936:5) | server
September 24, 2025, 22:23 | at async CommandRunnerService.run (/app/node_modules/nest-commander/src/command-runner.service.js:202:9) | server
September 24, 2025, 22:23 | at async CommandFactory.runApplication (/app/node_modules/nest-commander/src/command.factory.js:37:9) | server
September 24, 2025, 22:23 | at async bootstrap (/app/packages/twenty-server/dist/src/command/command.js:33:5) | server
September 24, 2025, 22:23 | ] | server
September 24, 2025, 22:23 | �[31m[Nest] 65 - �[39m09/25/2025, 5:23:21 AM �[31m ERROR�[39m �[38;5;3m[UpgradeCommand] �[39m�[31mCommand failed�[39m | server
September 24, 2025, 22:23 | �[31m[Nest] 65 - �[39m09/25/2025, 5:23:21 AM �[31m ERROR�[39m �[38;5;3m[UpgradeCommand] �[39m�[31mundefined�[39m | server
September 24, 2025, 22:23 | �[32m[Nest] 65 - �[39m09/25/2025, 5:23:21 AM �[32m LOG�[39m �[38;5;3m[UpgradeCommand] �[39m�[32mCommand completed!�[39m | server
September 24, 2025, 22:23 | �[31m[Nest] 65 - �[39m09/25/2025, 5:23:21 AM �[31m ERROR�[39m �[38;5;3m[Error] �[39m�[31mCould not run migration aborting�[39m | server
September 24, 2025, 22:23 | �[32m[Nest] 65 - �[39m09/25/2025, 5:23:21 AM �[32m LOG�[39m �[38;5;3m[DatabaseMigrationService] �[39m�[32mFound 32 active or suspended workspaces that are below version 0.53.0


^ taken from my ecs deployment of a docker image built from v1.4.0 source 

[prastoin]

Hey @mitodrummer you're missing 1.5 upgrade
Please upgrade your twenty incrementally
You can review you workspace version in database at core.workspace.version I guess you're still in 1.4

[mitodrummer]

So, I think what happened, was my custom built docker image was missing APP_VERSION build arg, and so all my workspaces had a NULL version. I've manually set them to 1.3.0 (since that was the version I started with), and built a proper 1.4.0 image with APP_VERSION set. This seems to have resolved my issue, so I am going to attempt the update to 1.5.0 and 1.6.7 incrementally. Wish me luck!

[prastoin]

Ok nice ! that's the way ! Please feel free to reach out if you have any issue
Closing from now on