feat(preprod): Add NOT_RAN to SizeAnalysisState

[chromy]

Now we can not run size analysis for two legitimate reasons:

• Not having enough quota
• Skipping size because we only want to run build distribution

It's important to be able to tell these apart in the UI.
Currently we don't create a PreprodArtifactSizeMetrics row in this case which
makes them hard to tell apart.

This adds a new state for the row: NOT_RAN and also two new 'error' states:

• NO_QUOTA
• SKIPPED

We make a new top level state to be able to easily tell this case apart from real failures
but reuse error_code and error_message rather than introduce two new columns.

[cursor]

PutSize model missing NOT_RAN state variant

High Severity

The endpoint adds a case handler for NOT_RAN state, but the PutSize request model in launchpad.py was not updated to include a PutSizeNotRan variant. The PutSize union type only accepts PutSizeFailed | PutSizePending | PutSizeProcessing. Any API request attempting to set state to NOT_RAN will fail Pydantic validation before reaching the match statement, making this feature completely non-functional.

 

[chromy]

PutSize - by design. Launchpad will never post back for a not_started analysis

[cursor]

Duplicated match cases for FAILED and NOT_RAN states

Low Severity

The FAILED and NOT_RAN cases have identical bodies (setting state, error_code, and error_message). Python's pattern matching supports combining cases with |, which is already used elsewhere in the codebase (e.g., event_parser.py). These two cases could be combined into a single case like case ... FAILED | ... NOT_RAN: to reduce code duplication and ensure consistent handling.

 

[cursor]

Duplicated match cases incrementing errored_count

Low Severity

The FAILED and NOT_RAN cases both simply increment errored_count. These could be combined using Python's | pattern matching syntax (case ... FAILED | ... NOT_RAN:) to eliminate the duplication and make the identical treatment explicit.

 

[chromy]

status check to be handled in follow up.

[sentry]

Bug: The get_size_analysis_response function fails to handle the new NOT_RAN state, causing an incorrect 500 error with a misleading message instead of a 422 with error details.
_{Severity: MEDIUM}

Suggested Fix

In the get_size_analysis_response function, add a conditional branch to explicitly handle the NOT_RAN state. This branch should return a 422 status code and include the specific error_code and error_message associated with the metric, similar to how the FAILED state is currently handled.

Prompt for AI Agent

Review the code at the location below. A potential bug has been identified by an AI
agent.
Verify if this is a real issue. If it is, propose a fix; if not, explain why it's not
valid.

Location:
src/sentry/preprod/api/models/project_preprod_build_details_models.py#L135-L141

Potential issue: The function `get_size_analysis_response` does not handle the newly
introduced `NOT_RAN` state for size analysis metrics. When a metric has this state, the
function bypasses existing checks for `PENDING`, `PROCESSING`, and `FAILED` states. It
then attempts to access the `analysis_file_id`, which is `None` for `NOT_RAN` metrics.
This results in a `SizeAnalysisResultsUnavailableError` being raised, leading to a 500
HTTP status code with the misleading message "Size analysis completed but results are
unavailable". The expected behavior is to return a 422 status code with a specific error
message explaining why the analysis was not run, such as `NO_QUOTA` or `SKIPPED`.

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with Cloud Agents, enable Autofix in the Cursor dashboard.}

[cursor]

NOT_RAN state causes inconsistent status check behavior

Medium Severity

The PR adds NOT_RAN handling in format_status_check_messages (incrementing errored_count), but the _compute_overall_status function in tasks.py wasn't updated. That function returns StatusCheckStatus.IN_PROGRESS for any state that isn't FAILED or COMPLETED. This causes NOT_RAN metrics to produce a status check that's permanently stuck at "in progress" while displaying a message saying "X app(s) errored" - an inconsistent state.

 

[chromy]

status check to be handled in follow up.