T
Twenty5mo ago
Rob

Upgrading from v0.12 to v0.20

Hi Twenty team, congratulations on the new release! I am trying to upgrade my self hosted docker compose instance and have ran in to some errors. Could you please let me know if these are critical or not? I am migrating from v0.12.2 1. I upgraded my twenty and twenty-postgres images to use v0.20.0. They built and executed without issue. 2. I restored my database back to the new postgres image database. 3. I launched the twenty docker image and it automatically ran a database setup and migrations. The last message was "Successfuly migrated DB!" I followed instructions from: https://discord.com/channels/1130383047699738754/1130383048173682821/1250864881561894994 and ran 3. docker exec -it twenty_1 yarn database:migrate:prod This returned "No migrations are pending" 4. docker exec -it twenty_1 yarn command:prod workspace:sync-metadata -w 6e278d07-a88a-45d3-86c9-daea30a9b0e5 -f through the following warning (complete warning attached as txt)
[Nest] 224 - 06/14/2024, 4:13:14 AM WARN [SyncWorkspaceMetadataCommand] Workspace health check failed with error, but sync has been forced.
[Nest] 224 - 06/14/2024, 4:13:14 AM WARN [SyncWorkspaceMetadataCommand] EntityNotFoundError: Could not find any entity of type "DataSourceEntity" matching:
[Nest] 224 - 06/14/2024, 4:13:14 AM WARN [SyncWorkspaceMetadataCommand] Workspace health check failed with error, but sync has been forced.
[Nest] 224 - 06/14/2024, 4:13:14 AM WARN [SyncWorkspaceMetadataCommand] EntityNotFoundError: Could not find any entity of type "DataSourceEntity" matching:
11 Replies
Rob
RobOP5mo ago
5. docker exec -it twenty_1 yarn command:prod migrate-0.20:update-message-channel-sync-status-enum the final message was [UpdateMessageChannelSyncStatusEnumCommand] No workspace found 6. docker exec -it home-server_twenty_1 yarn command:prod migrate-0.20:update-message-channel-visibility-enum the final message was [UpdateMessageChannelSyncStatusEnumCommand] No workspace found 7. docker exec -it home-server_twenty_1 yarn command:prod migrate-0.20:backfill-record-position  -w 6e278d07-a88a-45d3-86c9-daea30a9b0e5 throws the error error: required option '-w, --workspace-id [workspace_id]' not specified Thanks for assisting with this upgrade issue. Do you have any suggestions as to what I might be doing wrong and how I can fix this?
charles
charles5mo ago
It looks like there is no workspace found. If you connect to your database, can you see your workspace in core.workspace? in metadata.dataSource? do you see the corresponding postgres schema?
Rob
RobOP5mo ago
Thanks Charles. Maybe I restored my database incorrectly. I’ll try again and check the workspace as you suggest. I’ll report back here. Thanks
charles
charles5mo ago
ok!
Rob
RobOP5mo ago
Hi @charles , I'm pleased to report back that the previous errors were mainly my fault for making a mistake when i restored the database. However, the final recommended command still produces an error for me. When I run yarn command:prod migrate-0.20:backfill-record-position I still get the error error: required option '-w, --workspace-id [workspace_id]' not specified. If I include -w [workspace_id] I still get the same error. Additionaly, there are two workspace health check errors. - LOG [WorkspaceHealthCommand] Workspace is not healthy, found 2 issues Could this be the cause of my issue above? The web application appears to be working correctly and I'm loving the new features in v0.20, but I'd like to ensure my upgrade went correct before I move across. Is 'backfill-record-position' the correct command to run? Am I doing something obviously wrong? I’ve noticed that new changes I make appear in the cool new timeline. But changes to records from before the update aren’t in the timeline. Could this be because the backfill command didn’t run successfully?
Unknown User
Unknown User5mo ago
Message Not Public
Sign In & Join Server To View
Rob
RobOP5mo ago
Thanks @Félix , and is it a concern that I’m getting this error? https://discord.com/channels/1130383047699738754/1251031346957254727/1251828189815373894
Weiko
Weiko5mo ago
Hi @Rob, the health check errors are probably COLUMN_NULLABILITY_CONFLICT on FULL_NAME columns. You can read the logs in /logs folder inside your server package. If so you can ignore them as they are simply warnings. Actually I'm pretty sure that's what you have since I also have exactly 2 errors. We will fix them soon. To your first error, you still have ... [workspace_id]' not specified? Did you remove the brackets when you typed your workspace_id in the command?
Rob
RobOP5mo ago
Hi @Weiko thanks for getting back to me. For the error ... [workspace_id]' not specified , originally your instructions did not include the -w. Now that I run it with the updated command it works. Thanks! Let's mark this as resolved 🙂
charles
charles5mo ago
great! I feel we are improving in upgrading between versions but still some room for improvements to reach smoothness!
Rob
RobOP5mo ago
Pretty smooth to me! Congratulations on a great release
Want results from more Discord servers?
Add your server