[Question]: After I finished uploading the files, the generate button remained gray and the knowledge graph couldn't be enabled.I have configured the chat model.

[jikangding]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (Language Policy).
• Non-english title submitions will be closed directly ( 非英文标题的提交将会被直接关闭 ) (Language Policy).
• Please do not modify this template :) and fill in all the required fields.

Describe your problem

After I finished uploading the files, the generate button remained gray and the knowledge graph couldn't be enabled.I have configured the chat model.
After I parsed the file, the log page was empty.
It works fine on my Windows computer, but it doesn't work on the Linux server.

[jikangding]

[dosubot]

Hi @jikangding! I'm Dosu and I’m helping the ragflow team.

@jikangding The grayed-out generate button means your files have zero chunks — the button only enables when chunk_count > 0 [1]. The empty log page confirms that file parsing is likely failing silently on your Linux server.

A few things to check:

1. If you're on v0.25.0, there was a KeyError during metadata generation that caused parsing failures — this was fixed in PR #14383 [2]. Make sure you have this fix applied.

2. Knowledge graph config timing: There's a known issue where file processing configurations are locked at import time [3]. If you enabled knowledge graph settings after uploading files, you'll need to delete and re-import the files with knowledge graph already enabled [4].

3. Check your Docker/task worker logs on the Linux server (e.g., docker logs for the ragflow container) — there may be errors during parsing that aren't surfaced in the UI.

Could you share which RAGFlow version you're running on the Linux server and whether the version matches the one on your Windows machine?

To reply, just mention @dosu.

---

Share context across your team and agents. Try Dosu.

[jikangding]

Both my server and my local device are running version v0.25.0. I haven't seen a dedicated function to enable the knowledge graph. Even when I tried to create a new knowledge base, it didn't work

[dosubot]

@jikangding Since you're on v0.25.0 on both machines, the core problem is very likely a metadata generation bug that was just fixed in PR #14383 (merged April 27) [1]. This bug causes a KeyError: 'properties' during file parsing, which silently fails — resulting in zero chunks, an empty log page, and a grayed-out generate button. It particularly affects Linux deployments and upgrades from v0.24.0 [2].

To fix the parsing issue, you'll need to update your Linux server to include the PR #14383 fix — either by pulling the latest code and rebuilding the Docker image, or by volume-mounting the patched rag/svr/task_executor.py file [2].

Regarding the knowledge graph — there isn't a toggle to "enable" it at knowledge base creation. The workflow in v0.25.0 is:

1. Configure entity types in your dataset's Configuration page [3]
2. Parse your files (this is what's currently broken for you)
3. Generate — on the Files page, click the "Generate" button and select "Knowledge graph" [4]
4. Enable in retrieval — toggle "Use Knowledge Graph in search" when querying [4]

But none of this will work until file parsing succeeds, so applying the PR #14383 fix should be your first step.

To reply, just mention @dosu.

---

Share context across your team and agents. Try Dosu.

[buua436]

@jikangding
This issue has been fixed in this PR:
#14459

The fix will be included in the upcoming v0.25.1 release.