Skip to content

logging: classify generic http 5xx errors as transient #812

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ben-dz merged 2 commits into from
Aug 31, 2026
Merged

logging: classify generic http 5xx errors as transient #812

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
75d21dd
logging: classify generic http 5xx errors as transient
ben-dz Aug 30, 2026
da53e1f
logging: decide on the request's own status and scope aws matching to…
ben-dz Aug 31, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions indexer/pkg/solingest/logger_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,12 @@ func TestTemporalLoggerErrorLevel(t *testing.T) {
err: errors.New("Code: 62. DB::Exception: Syntax error"),
want: slog.LevelError,
},
{
name: "third-party http 503 demoted to warn",
msg: "Activity error.",
err: errors.New("validatorsapp refresh: failed to get validators: unexpected status code 503: <html><body><h1>503 Service Unavailable</h1></body></html>"),
want: slog.LevelWarn,
},
{
name: "context cancellation demoted to warn",
msg: "Activity error.",
Expand Down
11 changes: 11 additions & 0 deletions utils/pkg/dberror/dberror.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,12 @@ var eofRe = regexp.MustCompile(`\beof\b`)
// longer digit run (e.g. "statuscode: 2001").
var awsRespErrRe = regexp.MustCompile(`https response error statuscode: (200|500|502|503|504)\b`)
Comment thread
ben-dz marked this conversation as resolved.
Outdated

// httpStatus5xxRe matches the "status code NNN" shape non-AWS HTTP clients use
// (e.g. validators.app's "unexpected status code 503") for the same retryable
// set as awsRespErrRe. 501/505 are permanent endpoint failures and 4xx are
// actionable, so all keep paging.
var httpStatus5xxRe = regexp.MustCompile(`status[ _]?code[:= ]?\s*(500|502|503|504)\b`)
Comment thread
ben-dz marked this conversation as resolved.
Outdated

// ErrTransient is a sentinel that explicitly marks an error as transient for
// IsTransient, independent of its message. Wrap a return with it (e.g. via
// errors.Join or fmt.Errorf("...: %w", ErrTransient)) when the caller knows a
Expand Down Expand Up @@ -117,6 +123,11 @@ func Classify(err error) ErrorType {
return ErrorTypeConnectivity
}

// Retryable 5xx from any other HTTP upstream.
if httpStatus5xxRe.MatchString(errStr) {
return ErrorTypeConnectivity
}

// Connection/connectivity patterns
connectivityPatterns := []string{
"connectivityerror",
Expand Down
9 changes: 9 additions & 0 deletions utils/pkg/dberror/dberror_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,6 +59,15 @@ func TestClassifyAndIsTransient(t *testing.T) {
// A non-AWS message mentioning statuscode: 200 without the SDK prefix must not match.
{"non-aws statuscode 200", errors.New("handler returned statuscode: 200 but body was empty"), dberror.ErrorTypeUnknown, false},

// Non-AWS HTTP clients format retryable 5xx as "status code NNN" (the
// validators.app shape that paged on 2026-08-30). Same 5xx set as the AWS
// shape; 4xx and 501 stay actionable.
{"validatorsapp 503", errors.New(`validatorsapp refresh: failed to get validators: unexpected status code 503: <html><body><h1>503 Service Unavailable</h1></body></html>`), dberror.ErrorTypeConnectivity, true},
{"generic status code 500", errors.New("unexpected status code 500: internal server error"), dberror.ErrorTypeConnectivity, true},
{"generic status_code variant", errors.New("request failed: status_code=502"), dberror.ErrorTypeConnectivity, true},
{"generic status code 400 stays actionable", errors.New("unexpected status code 400: bad request"), dberror.ErrorTypeUnknown, false},
{"generic status code 501 stays actionable", errors.New("unexpected status code 501: not implemented"), dberror.ErrorTypeUnknown, false},

// Non-transient: real, actionable failures should still escalate to ERROR.
{"syntax error", errors.New("Code: 62. DB::Exception: Syntax error"), dberror.ErrorTypeQuery, false},
{"access denied", errors.New("access denied for user"), dberror.ErrorTypeAuth, false},
Expand Down
Loading