Stop source-name worker looping on permanent failures
The access-review source-name worker never marked a source synced when name resolution errored, so it re-claimed the source on every poll and retried at vendor-latency cadence. Two permanently-failing sources generated millions of error logs (Brex /v2/company 403 and Cloudflare /accounts 400) and hammered vendor APIs (8.6M 403s to Brex in 30 days) -- a ban risk, all for best-effort display metadata. Generalize the Google-403 special case: name resolvers now classify a non-2xx response through nameStatusError, which wraps ErrTerminalNameResolution for permanent client errors (400, 401, 403, 404) and returns a plain, retryable error for everything else (5xx, network). The worker treats a terminal error as done -- it keeps the generic name and marks the source synced -- while transient failures keep retrying as before. Also fix the Cloudflare name resolver's own bug: it requested per_page=1, but Cloudflare's List Accounts endpoint requires per_page in 5..50 and 400s otherwise (the driver already uses 50). That 400 was the sole cause of the Cloudflare retry storm; bump it to 50. Signed-off-by: Aurélien Sibiril <81782+aureliensibiril@users.noreply.github.com>
This commit is contained in:
@@ -41,6 +41,28 @@ type NameResolver interface {
|
||||
ResolveInstanceName(ctx context.Context) (string, error)
|
||||
}
|
||||
|
||||
// ErrTerminalNameResolution marks a name-resolution failure as permanent:
|
||||
// an auth or bad-request response that retrying cannot fix. The
|
||||
// source-name worker treats it as terminal — it keeps the generic source
|
||||
// name and marks the source synced instead of re-claiming it every poll.
|
||||
// Transient failures (5xx, network errors) are returned as plain errors so
|
||||
// they keep retrying. Name resolution is best-effort display metadata, so a
|
||||
// permanent failure must never wedge the worker in a retry loop (a single
|
||||
// unauthorized source otherwise produced millions of error logs in prod).
|
||||
var ErrTerminalNameResolution = errors.New("terminal name resolution failure")
|
||||
|
||||
// nameStatusError classifies a non-2xx response from a name-resolution
|
||||
// request. Permanent client errors (400, 401, 403, 404) wrap
|
||||
// ErrTerminalNameResolution; everything else (notably 5xx) stays retryable.
|
||||
func nameStatusError(what string, statusCode int) error {
|
||||
switch statusCode {
|
||||
case http.StatusBadRequest, http.StatusUnauthorized, http.StatusForbidden, http.StatusNotFound:
|
||||
return fmt.Errorf("cannot fetch %s: unexpected status %d: %w", what, statusCode, ErrTerminalNameResolution)
|
||||
default:
|
||||
return fmt.Errorf("cannot fetch %s: unexpected status %d", what, statusCode)
|
||||
}
|
||||
}
|
||||
|
||||
// slackNameResolver resolves the Slack workspace name via auth.test.
|
||||
type slackNameResolver struct {
|
||||
httpClient *http.Client
|
||||
@@ -143,7 +165,7 @@ func (r *linearNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch linear organization: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("linear organization", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -184,7 +206,7 @@ func (r *cloudflareNameResolver) ResolveInstanceName(ctx context.Context) (strin
|
||||
|
||||
q := cfURL.Query()
|
||||
q.Set("page", "1")
|
||||
q.Set("per_page", "1")
|
||||
q.Set("per_page", "50")
|
||||
cfURL.RawQuery = q.Encode()
|
||||
|
||||
req, err := http.NewRequestWithContext(ctx, http.MethodGet, cfURL.String(), nil)
|
||||
@@ -202,7 +224,7 @@ func (r *cloudflareNameResolver) ResolveInstanceName(ctx context.Context) (strin
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch cloudflare accounts: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("cloudflare accounts", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -251,7 +273,7 @@ func (r *brexNameResolver) ResolveInstanceName(ctx context.Context) (string, err
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch brex company: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("brex company", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -298,7 +320,7 @@ func (r *tallyNameResolver) ResolveInstanceName(ctx context.Context) (string, er
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch tally organization: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("tally organization", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -505,7 +527,7 @@ func (r *hubspotNameResolver) ResolveInstanceName(ctx context.Context) (string,
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch hubspot account info: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("hubspot account info", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -736,7 +758,7 @@ func (r *sentryNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
}
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch sentry organization: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("sentry organization", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -784,7 +806,7 @@ func (r *githubNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch github organization: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("github organization", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -916,7 +938,7 @@ func (r *gitlabNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch gitlab group: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("gitlab group", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -969,7 +991,7 @@ func (r *bitbucketNameResolver) ResolveInstanceName(ctx context.Context) (string
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch bitbucket workspace: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("bitbucket workspace", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -1028,7 +1050,7 @@ func (r *herokuNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch heroku team: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("heroku team", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -1178,7 +1200,7 @@ func (r *asanaNameResolver) ResolveInstanceName(ctx context.Context) (string, er
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch asana workspace: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("asana workspace", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -1228,7 +1250,7 @@ func (r *netlifyNameResolver) ResolveInstanceName(ctx context.Context) (string,
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch netlify account: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("netlify account", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -1276,7 +1298,7 @@ func (r *clickupNameResolver) ResolveInstanceName(ctx context.Context) (string,
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch clickup team: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("clickup team", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -1345,7 +1367,7 @@ func (r *vercelNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
}
|
||||
|
||||
if teamResp.StatusCode != http.StatusNotFound {
|
||||
return "", fmt.Errorf("cannot fetch vercel team: unexpected status %d", teamResp.StatusCode)
|
||||
return "", nameStatusError("vercel team", teamResp.StatusCode)
|
||||
}
|
||||
|
||||
// Personal-account fallback: /v2/teams/<uid> returns 404, but
|
||||
@@ -1399,7 +1421,7 @@ func (r *mondayNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch monday account: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("monday account", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -1452,7 +1474,7 @@ func (r *notionNameResolver) ResolveInstanceName(ctx context.Context) (string, e
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch notion users/me: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("notion users/me", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
@@ -1502,7 +1524,7 @@ func (r *microsoft365NameResolver) ResolveInstanceName(ctx context.Context) (str
|
||||
defer func() { _ = httpResp.Body.Close() }()
|
||||
|
||||
if httpResp.StatusCode < 200 || httpResp.StatusCode >= 300 {
|
||||
return "", fmt.Errorf("cannot fetch microsoft 365 organization: unexpected status %d", httpResp.StatusCode)
|
||||
return "", nameStatusError("microsoft 365 organization", httpResp.StatusCode)
|
||||
}
|
||||
|
||||
var resp struct {
|
||||
|
||||
@@ -22,6 +22,7 @@ package drivers
|
||||
|
||||
import (
|
||||
"context"
|
||||
"errors"
|
||||
"net/http"
|
||||
"net/http/httptest"
|
||||
"net/url"
|
||||
@@ -31,6 +32,35 @@ import (
|
||||
"github.com/stretchr/testify/require"
|
||||
)
|
||||
|
||||
func TestNameStatusError(t *testing.T) {
|
||||
t.Parallel()
|
||||
|
||||
terminal := []int{
|
||||
http.StatusBadRequest,
|
||||
http.StatusUnauthorized,
|
||||
http.StatusForbidden,
|
||||
http.StatusNotFound,
|
||||
}
|
||||
for _, code := range terminal {
|
||||
err := nameStatusError("thing", code)
|
||||
require.Error(t, err)
|
||||
assert.ErrorIs(t, err, ErrTerminalNameResolution, "status %d must be terminal", code)
|
||||
}
|
||||
|
||||
retryable := []int{
|
||||
http.StatusTooManyRequests,
|
||||
http.StatusInternalServerError,
|
||||
http.StatusBadGateway,
|
||||
http.StatusServiceUnavailable,
|
||||
http.StatusGatewayTimeout,
|
||||
}
|
||||
for _, code := range retryable {
|
||||
err := nameStatusError("thing", code)
|
||||
require.Error(t, err)
|
||||
assert.False(t, errors.Is(err, ErrTerminalNameResolution), "status %d must be retryable", code)
|
||||
}
|
||||
}
|
||||
|
||||
// hostRewriter redirects requests to the configured target host so that
|
||||
// resolvers with hardcoded production URLs (api.notion.com, etc.) can be
|
||||
// pointed at an httptest server.
|
||||
|
||||
@@ -172,6 +172,23 @@ func (h *sourceNameHandler) Process(ctx context.Context, source coredata.AccessR
|
||||
|
||||
instanceName, err := resolver.ResolveInstanceName(resolveCtx)
|
||||
if err != nil {
|
||||
// A permanent failure (auth/bad-request) cannot be fixed by
|
||||
// retrying: keep the generic name and mark the source synced so the
|
||||
// worker stops re-claiming it every poll. Returning the error here
|
||||
// would leave name_synced_at NULL and re-enqueue the source forever
|
||||
// (a single unauthorized source produced millions of error logs).
|
||||
if errors.Is(err, drivers.ErrTerminalNameResolution) {
|
||||
h.logger.WarnCtx(
|
||||
ctx,
|
||||
"permanent name resolution failure, keeping generic name",
|
||||
log.String("source_id", source.ID.String()),
|
||||
log.String("provider", dbConnector.Provider.String()),
|
||||
log.Error(err),
|
||||
)
|
||||
|
||||
return h.markNameSynced(ctx, &source)
|
||||
}
|
||||
|
||||
h.logger.WarnCtx(
|
||||
ctx,
|
||||
"cannot resolve instance name",
|
||||
|
||||
Reference in New Issue
Block a user