Upgrade not Possible: Manifest Unknown
I'm self-hosting and tried to update v 1.7.0 and 1.8.0 (currently on 1.6.7), but I keep getting this error:
What's causing this and what can I do?
34 Replies
Hey @BigKahuna Please upgrade to latest 1.7 patch version
Please also share the whole logs
Using Coolify and these are all the logs I am getting on deploying:
That's becasue 1.7.0 does not exist
Please consunme the 1.7 mirror patch agnostical
from docker hub
Where do I find this?
ah
https://hub.docker.com/r/twentycrm/twenty/tags
Please review your workspace version in database before and after the upgrade
That's inconvenient. I cannot switch installation methods. I commited to docker compose via Coolify and using a Docker Image basically forces me to create a new app and re-invite all users etc.
Coolify consumes the docker hub images under the hood
Interesting. Not that tech-savvy. How can I use the patch version without creating a new app?
Can I use a particular tag?
I'm not familiar with coolify
You should be able to update the TAG version of your .env
To be pointing on 1.7 instead of 1.7.0
( the same way you've followed to achieve to the point you're atm => eg consuming 1.7.0 )
That's easy enough. I'll try 1.7
That worked. Thanks for the help. Can I now switch to 1.7.0 and 1.8.0?
Please try avoiding consuming patched version, stay abstract and keep to consuming 1.7 and 1.8
Please double check that your workspace version in database is correct
My workspace version actually show 1.7.7 instead of 1.7. But just for my understanding: The production-ready version always 2 digits and 3 digits is the "patched" version?
When you consume the patch agnostical it will find the latest 1.7 version
So everything's good here
The upgrade was successful
You can attempt to 1.8 upgrade if you want to
OK, that broke things unfortuntely. My logs:
Server:
Worker:
Does this require a ticket on Github or it it a know issue with a quick fix?
Please share the whole logs seems to be truncated
They are due to message constraints
Server
Worker
My workspace version actually show 1.7.7 instead of 1.7.How are you viewing this please ? Seems like you're missing a previous migration
That changed when I deployed 1.7. It was 1.6.7 before that. Both values were identical
Where do you read the info please
In the UI
That's why, please try inspecting your database at
core.workspace.version
Unfortunately, not very intuitive they are not the same infoTook me a second to get it: 1.6.7
Ok the upgrade failed please inspect error I suspect the same as https://discord.com/channels/1130383047699738754/1430251005534011523 please refer to this channel in order to resolve the issue
So I went back to 1.6.7 and upgraded to 1.7. Here's what I'm seeing: No errors in my deployment or runtime logs of server, worker or db, but this line here is weird
My enviroment variables are actually set to this:
It's a "double negation" meaning I want to enable cron jobs and migrations. COuld this be the issue?
You should not have to touch to that
Please share your whole twenty-server logs
Upgrade command is run at the very first steps of your twenty-server bootstrap
Server runtime
Duplicated of https://discord.com/channels/1130383047699738754/1430251005534011523
Hey @Félix do you have more information regarding the
createHandoffFromDefaultAgent introduction
Could this migration have been introduced in a patch version of twenty ?
@BigKahuna I've removed the messages that might have contained private information about your coolify instance
createHandoffFromDefaultAgent seems to be quite hacky in the way it has been associated with the agent definition
The sync metadata seems to be expecting a column for it whereas it just contextual metadata for standard agent creation
@Félix createHandoffFromDefaultAgent does not seem to be consumed in the end or I'm misunderstanding something here
My guess would be that:
- createHandoffFromDefaultAgent: true, has been introduced 4 weeks ago in standard roles declaration
- but extraction of the given field has been introduced two weeks ago
There's a time lapse where some selft hosters might have been able to introduce roles with the data while proceeding to a sync metadata ( might also occur in production too )
But overall this does not seem to be a consumed property at all within the codebase if it's not even declare the role entitySo is there a solution for this issue?
Could you please dm me all your agent in database located at
core.agent ? @BigKahunaAny news here? Same issue.
[Nest] 142 - 11/04/2025, 8:57:15 PM ERROR [WorkspaceSyncMetadataService] Sync of standard objects failed with:
[Nest] 142 - 11/04/2025, 8:57:15 PM ERROR [WorkspaceSyncMetadataService] EntityPropertyNotFoundError: Property "createHandoffFromDefaultAgent" was not found in "AgentEntity". Make sure your query is correct.
[Nest] 142 - 11/04/2025, 8:57:17 PM WARN [UpgradeCommand] Error in workspace 10b90d20-3545-4e75-96be-fd9ed807af5c: Property "createHandoffFromDefaultAgent" was not found in "AgentEntity". Make sure your query is correct.
PromiseMemoizer Event: A WorkspaceDataSource for workspace 10b90d20-3545-4e75-96be-fd9ed807af5c is being cleared. Actual pool closure managed by PgPoolSharedService. Not calling dataSource.destroy().
[Nest] 142 - 11/04/2025, 8:57:17 PM ERROR [UpgradeCommand] Error in workspace 10b90d20-3545-4e75-96be-fd9ed807af5c: Property "createHandoffFromDefaultAgent" was not found in "AgentEntity". Make sure your query is correct.
[Nest] 142 - 11/04/2025, 8:57:17 PM ERROR [UpgradeCommand] undefined
[Nest] 142 - 11/04/2025, 8:57:17 PM LOG [UpgradeCommand] Command completed!