What causes "adopt tenant bootstrap credential failed: %w" in iflytek/astron-agent?

Inspect the wrapped driver error in tenant service logs (deadlock vs timeout vs connection loss).

adopt tenant bootstrap credential failed

Error message

adopt tenant bootstrap credential failed: %w

What it means

adoptTenantBootstrapCredential wraps a failure of the UPDATE that marks an existing (older-release, manually created via the public API) tb_auth credential row as tenant-managed by setting extend=tenantBootstrapManagedMarker. Any MySQL/driver error from ExecContext is wrapped with %w. The reconciliation is aborted so the credential is not silently left half-adopted.

Solutions

Inspect the wrapped driver error in tenant service logs (deadlock vs timeout vs connection loss).
For deadlock/lock-wait-timeout, simply retry: the whole reconcile runs in a transaction and is idempotent.
Ensure the reconcile transaction is executed against the primary, not a read-only replica.
Check MySQL error log for InnoDB deadlock details and serialize concurrent bootstrap runs (e.g. advisory lock).
Verify connection-pool sizing (max_open_conns) is adequate for the number of concurrent reconcilers.

Example fix

// before
if err := reconcileTenantBootstrapTransaction(ctx, tx, creds); err != nil {
    return err
}

// after: retry transient DB failures
for i := 0; i < 3; i++ {
    err := reconcileTenantBootstrapTransaction(ctx, tx, creds)
    if err == nil || !isRetryableDBError(err) {
        return err
    }
    time.Sleep(time.Duration(1<<i) * 100 * time.Millisecond)
}

Defensive patterns

Strategy: retry

Try / catch

if err := reconcile(ctx, tx, creds); err != nil {
    if isDeadlockOrTimeout(err) {
        return retryWithBackoff(ctx, func() error { return reconcile(ctx, tx, creds) })
    }
    return err
}

Prevention

Keep reconcile transactions short to reduce lock contention.
Run only one reconciler per deployment (leader election).
Point writes at the primary, never a read replica.
Retry idempotent transactional operations on transient DB errors.

When it happens

Trigger: transaction.ExecContext("UPDATE tb_auth SET extend=?, update_time=? WHERE app_id=? AND api_key=? AND api_secret=? AND is_delete=0 AND COALESCE(extend,'')<>?" ) returns a non-nil error during reconcileTenantBootstrapTransaction — e.g. deadlocked with another writer, lock wait timeout, connection reset, or read-only replica.

Common situations: Concurrent tenant bootstrap runs deadlock on the same tb_auth rows; the DB connection pool is exhausted; the statement is routed to a read-only replica; the table is corrupted or under schema migration (ALTER TABLE metadata lock).

Related errors

AI-assisted analysis of iflytek/astron-agent@5e758547a8 (2026-09-12). Data as JSON: /api/errors/22bf4f4b3aca8a6f. Report an issue: GitHub.

Appendix: source

Thrown at core/tenant/tools/database/bootstrap_credentials.go:229

) error {
	// A strong pair explicitly configured by the deployment may already have
	// been created through Tenant's public API on an older release. Because it
	// belongs to the reserved app and exactly matches the current deployment
	// Secret, adopt it into managed ownership so a later rotation can retire it.
	if _, err := transaction.ExecContext(
		ctx,
		`UPDATE tb_auth
SET extend = ?, update_time = ?
WHERE app_id = ? AND api_key = ? AND api_secret = ? AND is_delete = 0
  AND COALESCE(extend, '') <> ?`,
		tenantBootstrapManagedMarker,
		now,
		credentials.TenantID,
		credentials.APIKey,
		credentials.Secret,
		tenantBootstrapManagedMarker,
	); err != nil {
		return fmt.Errorf("adopt tenant bootstrap credential failed: %w", err)
	}
	return nil
}

func rotateTenantBootstrapCredentials(
	ctx context.Context,
	transaction bootstrapTransaction,
	credentials config.TenantBootstrapCredentials,
	now string,
) error {
	if _, err := transaction.ExecContext(
		ctx,
		`UPDATE tb_auth
SET is_delete = 1, update_time = ?
WHERE api_key = ? AND api_secret = ?`,
		now,
		config.LegacyTenantKey,
		config.LegacyTenantSecret,

View on GitHub (pinned to 5e758547a8)