[AI-3] Report the health of the LLM connection - #24889
Open
tangopium wants to merge 1 commit into
Open
Conversation
3 tasks
|
Caution The provided work package version does not match the core version Details:
Please make sure that:
|
|
Caution The Enterprise plan field is not set on the work package Details:
Please make sure that:
|
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 0191122c9c
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Comment on lines
+70
to
+71
| report = Llm::Validators::ConnectionValidator.new(connection).call | ||
| report.save! |
There was a problem hiding this comment.
When a manual deep check records an inference failure but /models remains reachable, the next six-hour job omits the inference group and saves its shallow report as the connection's latest report. Because the side panel and full status page only read latest_health_report, that later report can replace the failure with a green status without ever retrying the operation that failed; scheduled results need to retain the latest inference verdict or not supersede a deep report.
Useful? React with 👍 / 👎.
Collaborator
Author
There was a problem hiding this comment.
Deliberately left open: fixing this means either carrying old inference results into new reports or making a deep report sticky, which changes what the latest report means, and re-running inference on the schedule would spend a billed completion. Deferring the semantics to review.
tangopium
force-pushed
the
feature/66020-pr11-health-status
branch
from
August 22, 2026 15:05
0191122 to
df3f328
Compare
tangopium
changed the title
tangopium
force-pushed
the
feature/66020-pr11-health-status
branch
from
August 22, 2026 15:26
df3f328 to
bed276b
Compare
|
Warning Flaky specs
🤖 Ask Copilot to investigateCopy the prompt below into a new comment on this PR to delegate the investigation to GitHub Copilot. It will look into the flakiness and open a separate pull request with you as reviewer. |
tangopium
force-pushed
the
feature/66020-pr11-health-status
branch
from
August 22, 2026 15:51
bed276b to
ceca821
Compare
|
Warning Flaky specs
🤖 Ask Copilot to investigateCopy the prompt below into a new comment on this PR to delegate the investigation to GitHub Copilot. It will look into the flakiness and open a separate pull request with you as reviewer. |
Adds the health side panel and report page, reusing the HealthReports framework and preparing it for a third consumer. Six validator groups check the configuration, the server, the model catalogue, the capability verdicts, the feature bindings and, only when an administrator explicitly asks, a real inference call, so a scheduled check never spends a billed completion. The scheduled Llm::HealthCheckJob runs only while a connection is enabled, toggled from the update service and the disconnect action. Reports are kept as an audit trail with their own lifecycle, distinct from verdicts, and pruned by a cron job because nothing else prunes health_reports. Part 11 of the AI-3 stack. https://community.openproject.org/work_packages/66020
tangopium
force-pushed
the
feature/66020-pr11-health-status
branch
from
August 22, 2026 16:16
ceca821 to
761b2d4
Compare
|
Warning Flaky specs
🤖 Ask Copilot to investigateCopy the prompt below into a new comment on this PR to delegate the investigation to GitHub Copilot. It will look into the flakiness and open a separate pull request with you as reviewer. |
Deploying openproject with ⚡ PullPreview
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Ticket
AI-3
What are you trying to accomplish?
PR 11 of 11 in the AI-3 stack. Adds the health side panel and report page, reusing the HealthReports framework and preparing it for a third consumer. Six validator groups check the configuration, the server, the model catalogue, the capability verdicts, the feature bindings and, only when an administrator explicitly asks, a real inference call, so the scheduled check never spends a billed completion. The scheduled job runs only while a connection is enabled, reports are kept as an audit trail with their own lifecycle distinct from verdicts, and a cron job prunes them because nothing else prunes
health_reports.Merge checklist
llm_connectionfeature flagStacked on #24888.