t8y2/dbx · warning

failed to restore Oracle segment attributes: %w

Error message

failed to restore Oracle segment attributes: %w

What it means

This is the deferred cleanup path of the segment-attributes helper: after the operation runs, the driver re-enables Oracle segment attributes. If that restore statement fails, the driver marks the connection bad (driver.ErrBadConn) and, when the main operation did not already fail, returns 'failed to restore Oracle segment attributes'. The Oracle session may be left with segment attributes disabled.

Source

Thrown at agents/drivers/oracle-go/main.go:3148

		return s.appendTableDependentDDL(schema, table, fallback), nil
	}
	return "", err
}

func withOraclePortableMetadataSession(db *sql.DB, operation func(*sql.Conn) error) (err error) {
	conn, err := db.Conn(context.Background())
	if err != nil {
		return err
	}
	if _, err = conn.ExecContext(context.Background(), oracleDisableSegmentAttributesSQL); err != nil {
		_ = conn.Close()
		return fmt.Errorf("failed to disable Oracle segment attributes: %w", err)
	}
	defer func() {
		if _, resetErr := conn.ExecContext(context.Background(), oracleEnableSegmentAttributesSQL); resetErr != nil {
			_ = conn.Raw(func(any) error { return driver.ErrBadConn })
			if err == nil {
				err = fmt.Errorf("failed to restore Oracle segment attributes: %w", resetErr)
			}
		}
		_ = conn.Close()
	}()
	return operation(conn)
}

func (s *server) appendTableDependentDDL(schema, table, tableDDL string) string {
	indexDDLs, _ := s.loadTableIndexDDLs(schema, table)
	return s.appendTableDependentDDLWithIndexes(schema, table, tableDDL, indexDDLs)
}

func (s *server) appendTableDependentDDLWithIndexes(schema, table, tableDDL string, indexDDLs []string) string {
	var builder strings.Builder
	baseDDL := strings.TrimSpace(tableDDL)
	builder.WriteString(baseDDL)
	dependentAppended := false
	appendDependent := func(ddl string) {

View on GitHub (pinned to c0390bff16)

Solutions

1. Reconnect and manually run the enable-segment-attributes SQL to restore session state
2. Retry the whole operation; the driver invalidates the bad connection so a retry gets a fresh one
3. Investigate the wrapped reset error (check ORA- code) — typically a dropped connection
4. Shorten operation duration or raise Oracle/network idle timeouts so the conn survives to cleanup

Example fix

// before
conn.ExecContext(ctx, oracleDisableSegmentAttributesSQL) // long op, conn dies before reset
// after
ctx, cancel := context.WithTimeout(context.Background(), 10*time.Minute) // keep-alive long enough for op + reset
defer cancel()

if err := db.PingContext(ctx); err != nil { return err } // conn health before long operation

err := withSegmentAttributesDisabled(db, op)
if err != nil && strings.Contains(err.Error(), "restore Oracle segment attributes") {
    // reconnect and manually re-enable attributes before proceeding
    if rdb, e2 := sql.Open("oracle", dsn); e2 == nil {
        rdb.ExecContext(ctx, oracleEnableSegmentAttributesSQL)
        rdb.Close()
    }
}

When it happens

Trigger: Connection dropped or session killed between the operation and the deferred reset; privileges changed mid-flight; the enable SQL fails for version/edition reasons even though the disable succeeded.

Common situations: Long operations whose connection times out before cleanup; network interruption during a large export; DBA killing the session; failover to a node where the session state is gone.

Related errors

• begin manual transaction: %w
• strings.Join(failures, "; ")
• ${strings.Join(failures, "; ")}
• agent session not found: %s
• reserve connection for manual transaction: %w

AI-assisted analysis of t8y2/dbx@c0390bff16 (2026-09-05). Data as JSON: /api/errors/fc803d065f2f6e38. Report an issue: GitHub.