Route error spans through MapError to reduce UnknownFailure bucket#7241
Merged
rajeshkamal5050 merged 1 commit intomainfrom Mar 24, 2026
Merged
Route error spans through MapError to reduce UnknownFailure bucket#7241rajeshkamal5050 merged 1 commit intomainfrom
rajeshkamal5050 merged 1 commit intomainfrom
Conversation
#7239) Several high-volume code paths used EndWithStatus(err) which bypasses MapError() — the central error classification function. This caused errors to appear with raw Go type names or empty descriptions in telemetry, inflating the UnknownFailure bucket. Changes: - cmd/mcp.go: MCP tool handler now calls MapError(err, span) instead of EndWithStatus(err), giving MCP tool failures proper telemetry codes (e.g., service.arm.400 instead of azcore_ResponseError) - internal/agent/copilot_agent.go: Initialize() and ensureSession() now route errors through cmd.MapError for structured classification - internal/tracing/tracer.go: EndWithStatus() now prefixes error descriptions with 'internal.' to match MapError's catch-all convention, ensuring remaining callers in pkg/ (container_helper, extensions/manager) produce consistent internal.* codes Fixes #7239 Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
spboyer
requested review from
JeffreyCA,
hemarina,
vhvb1989 and
weikanglim
as code owners
Contributor
There was a problem hiding this comment.
Pull request overview
Routes more error paths through the centralized internal/cmd.MapError classification so telemetry spans emit structured result codes (reducing the UnknownFailure bucket) instead of raw Go type names.
Changes:
- Updated MCP tool handler spans to call
internal/cmd.MapError(err, span)before ending spans. - Updated Copilot agent spans in
Initialize()andensureSession()to usecmd.MapErrorfor error classification. - Adjusted
Span.EndWithStatus(err)fallback to prefix raw error type descriptions withinternal.for callers that can’t useinternal/cmd.MapError.
Reviewed changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.
| File | Description |
|---|---|
| cli/azd/internal/tracing/tracer.go | Prefixes EndWithStatus error status descriptions with internal. to align with the repo’s catch-all convention. |
| cli/azd/internal/agent/copilot_agent.go | Routes Copilot agent span errors through cmd.MapError and ends spans explicitly. |
| cli/azd/cmd/mcp.go | Routes MCP tool handler span errors through internal/cmd.MapError and ends spans explicitly. |
| cli/azd/.vscode/cspell.yaml | Adds internalcmd to cspell overrides for cmd/mcp.go. |
Collaborator
Azure Dev CLI Install InstructionsInstall scriptsMacOS/Linux
bash: pwsh: WindowsPowerShell install MSI install Standalone Binary
MSI
Documentationlearn.microsoft.com documentationtitle: Azure Developer CLI reference
|
Member
Author
Telemetry Context: 1,017 Opaque Errors This PR Will Help ClassifyDeep dive into 1,017 errors (21.6%) have Where the opaque errors come from
Container Apps (all variants) = 47% of the opaque errors. Why this matters beyond classificationThe ARM response body contains an Currently, 366 unique machines hit these opaque errors. 65% retry and fail again — they have no information to guide a fix. |
JeffreyCA
approved these changes
Mar 24, 2026
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
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.
Summary
Fixes #7239
The
UnknownFailureresult code is the single largest error bucket in telemetry — 2.35M errors/28d with zero classification. Root cause: several high-volume code paths usedspan.EndWithStatus(err)which bypassesMapError(), the central error classification function.Root Cause
EndWithStatus(err)callserrorDescription(err)which produces a raw Go type name (e.g.,errors_errorString) as the span status description. When this is empty or unrecognized, the telemetry converter falls back to"UnknownFailure":Meanwhile,
MapError()produces structured codes likeservice.arm.400,auth.login_required,tool.docker.failed— but these paths bypass it.Changes
1. MCP tool handler (
cmd/mcp.go)Replaced
span.EndWithStatus(err)withinternalcmd.MapError(err, span)+span.End(). Every MCP tool failure now gets proper telemetry classification.2. Copilot agent (
internal/agent/copilot_agent.go)Both
Initialize()andensureSession()now route errors throughcmd.MapErrorinstead ofEndWithStatus.3.
EndWithStatusfallback (internal/tracing/tracer.go)For remaining callers in
pkg/(container_helper.go, extensions/manager.go) that cannot importinternal/cmd,EndWithStatus()now prefixes error descriptions withinternal.to matchMapError's catch-all convention. This ensures consistentinternal.*codes instead of bare type names.Expected Impact
UnknownFailureor raw type nameservice.arm.400,auth.login_required, etc.errors_errorStringinternal.errors_errorString(consistent convention)This should significantly reduce the 2.35M
UnknownFailurebucket by routing errors through the existing classification infrastructure.