get current bazaar index failed: %w; get legacy bazaar index

Error message

get current bazaar index failed: %w; get legacy bazaar index failed: %v

What it means

fetchBazaarIndex tries the current CDN bazaar index path first and falls back to the legacy path. This error wraps both failures: the primary fetch error (%w) and the legacy fetch error (%v). It means neither the current nor the legacy index endpoint could be reached or parsed, so no bazaar index snapshot is available.

Source

Thrown at kernel/bazaar/index.go:177

}

func snapshotBazaarIndex(now time.Time) (snapshot *bazaarIndexSnapshot, fresh, canRetry bool) {
	bazaarIndexState.mu.RLock()
	defer bazaarIndexState.mu.RUnlock()
	snapshot = bazaarIndexState.snapshot
	fresh = nil != snapshot && now.Before(bazaarIndexState.expiresAt)
	canRetry = bazaarIndexState.retryAt.IsZero() || !now.Before(bazaarIndexState.retryAt)
	return
}

func fetchBazaarIndex(ctx context.Context) (ret *bazaarIndexSnapshot, err error) {
	ret, err = fetchBazaarIndexPath(ctx, bazaarIndexPath)
	if nil == err {
		return
	}
	legacy, legacyErr := fetchBazaarIndexPath(ctx, bazaarLegacyIndexPath)
	if nil != legacyErr {
		return nil, fmt.Errorf("get current bazaar index failed: %w; get legacy bazaar index failed: %v", err, legacyErr)
	}
	return legacy, nil
}

func fetchBazaarIndexPath(ctx context.Context, indexPath string) (ret *bazaarIndexSnapshot, err error) {
	timeBucket := bazaarIndexNow().Unix() / int64(bazaarIndexCDNBucket/time.Second)
	u := fmt.Sprintf("%s%s?t=%d", bazaarIndexStatServer, indexPath, timeBucket)
	buf := &bytes.Buffer{}
	resp, err := httpclient.NewBrowserRequest().SetRetryCount(0).SetContext(ctx).SetOutput(buf).Get(u)
	if nil != err {
		return nil, err
	}
	if 200 != resp.StatusCode {
		return nil, fmt.Errorf("unexpected status code: %d", resp.StatusCode)
	}
	return parseBazaarIndex(buf.Bytes())
}

View on GitHub (pinned to 8641553a1f)

Solutions

Fix the underlying connectivity problem so at least one of the two index URLs is reachable
Check whether the bazaar server is publishing a valid index; a single fixed server-side index resolves this client-side
Rely on the cached snapshot if one exists (snapshotBazaarIndex serves stale data while refresh fails) and retry after network recovery
Inspect the wrapped errors (%w primary and %v legacy) separately — the primary error usually names the real cause (status code, timeout, parse failure)

Defensive patterns

Strategy: retry

Try / catch

snapshot, err := fetchBazaarIndex(ctx)
if err != nil {
    var netErr net.Error
    if errors.As(err, &netErr) || strings.Contains(err.Error(), "unexpected status code") {
        // schedule a retry with backoff; serve stale cached snapshot meanwhile
    }
}

Prevention

Serve the last cached snapshot while refreshes fail instead of erroring to users
Retry with exponential backoff — CDN outages are usually transient
Alert on server-side index publication failures, since both paths failing usually means the publisher is broken

When it happens

Trigger: The background index refresh calls fetchBazaarIndex and fetchBazaarIndexPath fails for both bazaarIndexPath and bazaarLegacyIndexPath — e.g. network is down, both CDN URLs return non-200 (see index.go:191), or both payloads fail parseBazaarIndex validation.

Common situations: Total loss of internet connectivity; bazaar CDN outage affecting both index URLs; a server-side change publishing invalid index JSON to both paths; DNS or proxy blocking the bazaar host; unit tests forcing both paths to fail.

Related errors

AI-assisted analysis of siyuan-note/siyuan@8641553a1f (2026-09-11). Data as JSON: /api/errors/9a7744e380d677d5. Report an issue: GitHub.