[solved] Workspace stuck in 0.55.9 | Twenty | Page 1

stable marsh Aug 17, 2025, 5:59 PM

#

Hi, all. I just migrated from v1.2.0 to 1.3.1 and suddenly, everything broke. I am getting a black screen without a sidebar menu.

I found these runtime server logs, but dunno what they actually mean (business user):

2025-08-17T18:09:31.376333572Z ICU features such as interpolation and plurals will not work properly for that message. 
2025-08-17T18:09:31.376338040Z 
2025-08-17T18:09:31.376342278Z Please compile your catalog first. 
2025-08-17T18:09:31.376347189Z 
2025-08-17T18:09:31.376351637Z Uncompiled message detected! Message:
2025-08-17T18:09:31.376355484Z 
2025-08-17T18:09:31.376358289Z > OsENGP
2025-08-17T18:09:31.376360825Z 
2025-08-17T18:09:31.376363831Z That means you use raw catalog or your catalog doesn't have a translation for the message and fallback was used.

stable marsh Aug 17, 2025, 7:01 PM

#

Rolling back to v1.2.0. killed everything:

waxen hawk Aug 17, 2025, 8:06 PM

#

Hello there, seems like you're having issue with translations. But it should not be a blocking issue

#

Could you please share your browser console logs

#

Especially failing network request responses

stable marsh Aug 17, 2025, 10:02 PM

#

stable marsh Hi, all. I just migrated from v1.2.0 to 1.3.1 and suddenly, everything broke. I ...

Looking at the admin panel, at first, all systems seem operational.

BUT: When wanting to see more detail for the worker, I get the error "Queue information is not available because the worker is supposedly down".

stable marsh Aug 18, 2025, 6:20 AM

#

waxen hawk Hello there, seems like you're having issue with translations. But it should not...

I am getting 500 errors from some graphql requests. I attached the request.
And here's the response:

{
    "errors": [
        {
            "message": "Unable to merge GraphQL type \"Query\": Field \"apiKeys\" already defined with a different type. Declared as \"ApiKey\", but you tried to override with \"ApiKeyConnection\""
        }
    ]
}

📎 message.txt

waxen hawk Aug 18, 2025, 6:25 AM

#

Seems like your missing API key migration

#

What s your workspace version in db ?

stable marsh Aug 18, 2025, 6:30 AM

#

waxen hawk What s your workspace version in db ?

Great question. How do I find out 😂?

#

I went back up to the version 1.3.1, if that's what you mean with "workspace".

#

With the AI's help, here's what I found: workspace metadataVersion = 64.

waxen hawk Aug 18, 2025, 6:44 AM

#

Not the metadata but version at core.workspace.version

stable marsh Aug 18, 2025, 6:57 AM

#

0.55.9 appears correct

waxen hawk Aug 18, 2025, 8:13 AM

#

Seems like your workspace is blocked in 0.55.9
Please try fixing your instance by downgrading it to 0.55.9 and re-attempt the upgrade process sharing any encountered errors

stable marsh Aug 18, 2025, 9:10 AM

#

But I am already on 0.55.9, so to what version should I downgrade it to? And are there any guides or so on how to do that specifically? Or does that just involve rolling back to 1.3.0 or 1.2.1 and sharing the errors popping up there?

#

Oh, hold on. YOu mean roll back all the way to v0.55.9? But I was never on that version. I joined to project when it was on v1.0 or so.

waxen hawk Aug 18, 2025, 9:28 AM

#

0.55.9 is the version just before 1.0.0

waxen hawk Aug 18, 2025, 9:28 AM

#

stable marsh Oh, hold on. YOu mean roll back all the way to v0.55.9? But I was never on that ...

And yes you need to 0.55 -> 1.0 -> 1.1 -> 1.2 -> 1.3

stable marsh Aug 18, 2025, 9:47 AM

#

waxen hawk Seems like your workspace is blocked in `0.55.9` Please try fixing your instance...

So here are the errors I encountered when going to v0.55 in my postgres container:

#

📎 message.txt

#

And this happened in the worker container when moving to v1.0:

#

📎 message.txt

#

In both cases the app would not load and just show this

#

Here are the error logs after switching to 1.1

#

Worker container:

📎 message.txt

#

The postgres container also showed some errors:

2025-08-18T09:54:00.147845194Z 2025-08-18 09:54:00.147 UTC [125] ERROR:  relation "core.settingPermission" does not exist at character 2853
2025-08-18T09:54:00.147870382Z 2025-08-18 09:54:00.147 UTC [125] STATEMENT:  SELECT "RoleEntity"."id" AS "RoleEntity_id", "RoleEntity"."label" AS "RoleEntity_label", "RoleEntity"."canUpdateAllSettings" AS "RoleEntity_canUpdateAllSettings", "RoleEntity"."canReadAllObjectRecords" AS "RoleEntity_canReadAllObjectRecords", "RoleEntity"."canUpdateAllObjectRecords" AS "RoleEntity_canUpdateAllObjectRecords",

#

But while the app is still not working there aren't any GraphQL errors with v1.1 (forgot to check the versions before)

#

And now version 1.2. App is still not functioning (being sent to the welcome page, but seein nothing), but postgres and redis are ready to accept connections. Not seeing any other errors strangely.

#

And for the grand finale: v1.3. Again: postgres and redis seem fine and are waiting for connections. And I'm not seeing any errors in the server or worker container either directly after installation.
But when going to the app, there's 500 error:

#

📎 message.txt

#

So despite all that, I seem to have landed right back where I started. I am still stuck on 0.55.9 according to my db.

waxen hawk Aug 18, 2025, 12:44 PM

#

Please do not upgrade to a next version until your workspace version in db equal the installed twenty instance

#

If you do you will miss migration resulting across time in blocking issue

#

Seem to be a duplicated of https://discord.com/channels/1130383047699738754/1405234431085772891

#

Will have a look within the day

stable marsh Aug 18, 2025, 3:44 PM

#

waxen hawk Please do not upgrade to a next version until your workspace version in db equal...

OK, so when / how does the workspace version change if not by installing the next version? I'm back down to 0.55.9 now.

waxen hawk Aug 18, 2025, 3:48 PM

#

When upgrading to the next version, if the upgrade and migrate are successfull your workspace version will get updated to the currently run twenty instance version

stable marsh Aug 18, 2025, 3:49 PM

#

In that case even the first upgrade appears to be failing. Unsure why though. And to top it off: Nothing is working anymore right now either

waxen hawk Aug 18, 2025, 3:50 PM

#

You might have corrupted your workspace

#

What's not working ? Please share errors of the current failing version upgrade

#

Do you have any backups ?

stable marsh Aug 18, 2025, 3:54 PM

#

Nothing is. Getting this screen. Only error I can see is in the postgres container

#

📎 message.txt

#

No backups unfortunately

stable marsh Aug 18, 2025, 10:36 PM

#

Nevermind all this. I just reinstalled everything and am now on v1.3.0. It was probably less pain than trying to figure out how to fix this stuff.

raw sigil Aug 20, 2025, 12:00 PM

#

@stable marsh Can you clarify what steps you took to remedy this?

stable marsh Aug 20, 2025, 4:54 PM

#

Will, like I said: I deleted everything and started from scratch

raw sigil Aug 21, 2025, 9:19 AM

#

Do you mean you manually re-entered data into a blank database?

meager niche Aug 21, 2025, 11:26 AM

#

facing similar issue. i'm on 1.2.1 (everything is fine) and when i upgrade to 1.3.0 or 1.3.1 i get the same erros OP got. both the graphql 500 aswell as the translation issues. did you find a solution that does not include copying data manually? or there any configs that need to be set properly for this update?

meager niche Aug 21, 2025, 12:00 PM

#

workspace version is 1.2.1

waxen hawk Aug 21, 2025, 1:00 PM

#

Hello @meager niche not the same issue here than you
No need to reset your instance please try running translation codegen commands

fast star Aug 21, 2025, 3:08 PM

#

meager niche facing similar issue. i'm on 1.2.1 (everything is fine) and when i upgrade to 1....

Hey! The way i solved it was by first checking docker logs, i noticed there was a DB error: 'collation version mismatch' manually, first fixed that by accessing twenty-db container and excecuting the commands suggested by postgre. ALTER DATABASE <db name> REFRESH COLLATION VERSION; on all the databases. But this only solved db errorrs on the container.

Then i downgraded back to 1.2.0, which first showed me an error, but then i tested on incognito and it worked so I deleted cache on the browser. And afterwards i started doing the upgrade in steps. v1.3.0 then v1.3.1

analog dust Aug 21, 2025, 9:21 PM

#

I have the same problem as you. I decided to delete everything too. I'll come back soon and reset everything and start over, because I honestly can't.

quaint shuttle Aug 22, 2025, 6:04 AM

#

You can ignore the translation message, it's just a warning. Could you check what your browser network tab?

#

I think the server is sending back an error to the FE but not logging it. There was a breaking change in views between 1.2 and 1.3, it might be it

meager niche Aug 22, 2025, 7:15 AM

#

getting this 500 graphql error OP also got:
#1406698879575130254 message

placid robin Aug 22, 2025, 7:28 AM

#

I might have the same issue (https://discord.com/channels/1130383047699738754/1408348100581527616)

quaint shuttle Aug 22, 2025, 7:40 AM

#

Could you both share the response from your network tab?

#

let's not mix the thread too, it's not helping :p

meager niche Aug 22, 2025, 8:31 AM

#

**Request URL: **https://xxx/graphql
**Request Method: **POST
Status Code: 500 Internal Server Error

Payload: (removed parts of the query bc of msg length limit)

operationName: "FindManyViews"
query: "query FindManyViews($filter: ViewFilterInput, $orderBy: [ViewOrderByInput], $lastCursor: String, $limit: Int) 
variables: {}

Response:

{
    "errors": [
        {
            "message": "Unable to merge GraphQL type \"Query\": Field \"apiKeys\" already defined with a different type. Declared as \"ApiKey\", but you tried to override with \"ApiKeyConnection\""
        }
    ]
}

#

and login does not properly work (the redirect) but if i then change the url to just baseurl i am able to go to the admin dashboard which shows everything as operational but all collections aswell as views dont show up

waxen hawk Aug 22, 2025, 8:46 AM

#

Hello guys it's really hard to follow several issue within the same thread, we may have been sharing the same translations error but it has never been blocking
It seems like you all have différents root cause for the moment
Please review your workspace version in database in core.workspace.version
If you workspace version is lower than the current installed version it means you should downgrade your instance and retry upgrade of the current workspace version
It should be failing for some reasons and this is your error

waxen hawk Aug 22, 2025, 8:46 AM

#

meager niche **Request URL: **https://xxx/graphql **Request Method: **POST **Status Code:** 5...

Seems like your missing priors migrations around api keys please do as above

#

I'm locking this thread for the moment please feel free to reopen new ones with the minimal required information:

twenty installed version
workspace version
any error logs related to an upgrade process ( translations bring a false positive )

#[solved] Workspace stuck in 0.55.9