Join us for an expert-led overview of the tools and concepts you'll need to pass exam PL-300. The first session starts on June 11th. See you there!
Get registeredJoin us at FabCon Vienna from September 15-18, 2025, for the ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM. Get registered
Hi,
I have a workspace with Warehouse and Lakehouse in there and until this morning, the workspace was nicely integrated with git. That is sync to and update from git were just working fine.
But all of a sudden, the git inetgration is not working with the workspace now. Neither can sync from or sync to git. No clear error but just " Something went wrong " or "Unable to update workspace" .
I have tried mapping the branch to a new workspace and still the same issue.
If I drop the Lakehouse from Git branch ( only a meta file anyways) , commit and then try to sync , then workspace gets updated.
So, it seems some kind of release or on-going-work by the fabric dev team, has broken the git inetgration with workspaces having lakehouses.
Is someone else facing the same issue . It started for me this morning GMT at around 9 AM
Thanks.
AK
Solved! Go to Solution.
Think , this issue is surfacing in different regions one by one. I have not faced this issue anymore since last Thursday. So it tells me that the fix or rollback (whatevr) is also getting released region by region. Strange that no one from microsoft has replied on this thread to assure everyone on the issue. It wasted one day or my time and it seems many people are trying to address it by lengthy workarounds which may become redundant once the issue is fixed.
Today I found this known issue in Fabric https://learn.microsoft.com/en-us/fabric/get-started/known-issues/known-issue-1005-git-deployment-pi... that more or less describes the issue. As per the page, there is no workarrounds for it and appears that also deployment pipelines are affected by this.
Still having this issue in North Europe as of 28/01/2025, known issue on the github:
I get the error message that the lakehouse item name is reserved when trying to sync through the UI and when syncing via the API and checking the long running operation status I get this:
{"status":"Failed","createdTimeUtc":"2025-01-28T15:12:20.2216078","lastUpdatedTimeUtc":"2025-01-28T15:12:23.5966656","percentComplete":null,"error":{"errorCode":"GitSyncFailed","moreDetails":[{"errorCode":"Git_InvalidResponseFromWorkload","message":"An error occurred while processing the operation",
Anyone know of an ETA on this fix?
For us it is solved now.
It's the same issue in my side too. Does anyone know if Microsoft is seriously addressing this issue?
It seems the issue has been resolved. It is working as expected today.
Can this be the rootcause for the current problems:
Azure DevOps has switched CDN providers, started on 8-1-2025
Azure DevOps - Event
Did this impact the internal Fabric <-> Azure Devops communitcation?
A fix for this is being investigated by Fabric engineers. I got a EOD 24th as a date from the support.
Yeah, some users should now be able to sync there workspaces back with Azure DevOps and Github. They have implemented the fix for some "clusters" already. This was not a DevOps issue (since they had a recent update), I tried cloning my repo to a github account and connected to it there as well, same issue.
As of now, I still have this issue.
My workspaces are syncing with Azure DevOps again.
I was just able to synchronize the Lakehouses with my Github repository again. Apparently something is happening and an improvement is noticeable.
Same issue for the North Europe zone. When trying to update a workspace from Git using PowerShell, we get the error ""InvalidDefinitionFormat ,message":"Requested item definition format is invalid".
When trying to sync a new empty workspace with a Fabric repo containing a Lakehouse, we get the error "Git_InvalidResponseFromWorkload".
We're seeing it in North Europe as well
This issue pertains to the lakehouses in your repository. Although deleting the lakehouse, as only the metadata is version-controlled and resynchronized, temporarily resolves the problem, this is not an ideal solution. I have contacted the support team, but they have not provided an estimated time for a fix.
If you need a quick workaround for the moment, you can delete the lakehouse folder in your repository. Additionally, if you have any "Hidden" items, such as "StagingLakehouse," they should also be deleted. However, this should be done by someone familiar with navigating the Fabric environment, as it requires expertise to ensure everything integrates correctly. It is not recommended if you do not have oversight or control over how all components come together.
Don't know how removing the lakehouse folders from GIT will solve the sync issue. Currently, my workspace is connected to GIT but not synched due to this error. All the artefacts that are on GIT are also inside the workspace. If I remove the lakehouse folders then reconnect the workspace with GIT and try the sync, I get the following
Also, just be reading how the initialize should take place (meaning Sync content from this workspace into Git or Sync content from Git into this workspace), if there is something in the workspace versioned controlled like a lakehouse that is no longer in the GIT repository, then, during sync, these artefacts will be deleted if I select the second option, or commited and pushed to GIT if I select the first one. For the second option, I uderstand that the sync will imply the deletion of the lakehouse from the workspace and this implies data loss.
The workaround works when you delete the folder from the remote repository before you sync the branch to your workspace.
We have the same issue now in Sweden it seems, has anyone created a ticket regarding this i powerapps and got it resolved without waiting out 1 day?
Same here, is there a working workarround for this, or we just need to wait and hopefully the functionality will be restored at some point with no time frame at the horizon?
For me it started to fail like this aprox 24 hours ago (at the moment I am writing this message). Also, in the developer console of the browser I have intercepted a message Invalid artifact definition format: LakehouseGit. In all the metadata and .platform files stored in GIT I don't have this type written.
Also, what I observed is that, now, the Lakehosue and Warehouse created when a dataflow gen 2 is created, are no longer visible in the Source Control tab when there is a dataflow present in the workspace. Could these be related?
Previously working Fabric workspaces with Lakehouses integrated to Azure DevOps Git repos are failing with the 'Something went wrong' error message. My tenant is in 'North Central US (Illinois)'. It started with one workspace and then eventually all workspaces in the tenant integrated to Git are failing to sync. Looking at the network calls in the browser dev tools, the Fabric UI git status response was a 400 after selecting the 'Sync branch' button.
I noticed the workspace/git/status API calls were failing and returned a 400 code. I tried to call git/status myself using the Fabric API and the error message said I needed to initialize the workspace. So I tried.
https://api.fabric.microsoft.com/v1/workspaces/<my workspace id>/git/initializeConnection
body payload
{
"initializationStrategy": "PreferRemote"
}
Here is the 400 response
{
"requestId": "06c9683e-e02c-433d-8797-d6f6a9704708",
"errorCode": "InvalidDefinitionFormat",
"message": "Requested item definition format is invalid"
}
Maybe I do not have the correct body payload parameters. If I do, then Microsoft should fix this API call.
Think , this issue is surfacing in different regions one by one. I have not faced this issue anymore since last Thursday. So it tells me that the fix or rollback (whatevr) is also getting released region by region. Strange that no one from microsoft has replied on this thread to assure everyone on the issue. It wasted one day or my time and it seems many people are trying to address it by lengthy workarounds which may become redundant once the issue is fixed.
Hi @Amit_K,
Thank you for confirming that your issue has been resolved. I sincerely apologize for the inconvenience you experienced. Since you haven't encountered the issue since last Thursday, and the resolution is being gradually deployed across different regions. I request you to please accept this post as a solution to help other community members find it more quickly, and don't forget to give a Kudos – it would be greatly appreciated!
Thanks and regards,
Anjan Kumar Chippa
User | Count |
---|---|
13 | |
4 | |
3 | |
3 | |
3 |
User | Count |
---|---|
8 | |
8 | |
7 | |
6 | |
5 |