备库缺少用于接收主库日志的备用重做日志文件-恢复dg

### 1. 当前 SCN 状态

- **备库 SCN**:
  ```sql
  SELECT dbid, checkpoint_change# FROM v$database;
  ```

- **备库当前 SCN**:
  ```sql
  SELECT TO_CHAR(dbms_flashback.get_system_change_number, 'FM9999999999999999999999999999') get_system_change_number FROM dual;
  ```

- **主库 SCN**:
  ```sql
  SELECT dbid, TO_CHAR(CHECKPOINT_CHANGE#, 'FM9999999999999999999999999999') AS CHECKPOINT_CHANGE FROM v$database;
  SELECT TO_CHAR(dbms_flashback.get_system_change_number, 'FM9999999999999999999999999999') get_system_change_number FROM dual;
  ```

### 2. 备库日志状态

- 备库在日志应用时出现以下信息：
  ```
  RFS[5]: No standby redo logfiles created for thread 1
  Media Recovery Waiting for thread 1 sequence 327076 (in transit)
  ```
  这表示备库缺少用于接收主库日志的备用重做日志文件。

### 3. 处理 SCN 差异

由于备库 SCN 落后于主库 SCN，您可以采取以下步骤进行处理：

1. **停止备库的日志应用**:
   ```sql
   ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
   ```

2. **在主库上进行增量备份**:
   使用 RMAN 从主库进行增量备份：
   ```sql
   RMAN> BACKUP DEVICE TYPE DISK INCREMENTAL FROM SCN 14630127306418 DATABASE FORMAT '/oradata/bak/orclbk_%U.bk';
   ```

3. **将备份和控制文件传到备库**:
   ```bash
   scp /path/to/backup/* oracle@<standby_host>:/oradata/backup
   ```

4. **在备库上创建新的控制文件**:
   ```sql
   ALTER DATABASE CREATE STANDBY CONTROLFILE AS '/oradata/bak/stb.ctl';
   ```

5. **在备库上启动到 NOMOUNT 状态**:
   ```sql
   STARTUP NOMOUNT;
   ```

6. **恢复控制文件**:
   ```sql
   RMAN> RESTORE CONTROLFILE FROM '/oradata/bak/stb.ctl';
   ```

7. **挂载数据库**:
   ```sql
   RMAN> ALTER DATABASE MOUNT;
   ```

8. **注册备份文件**:
   ```sql
   RMAN> CATALOG START WITH '/oradata/bak/';
   ```

9. **恢复数据文件**:
   ```sql
   RMAN> RECOVER DATABASE NOREDO;
   ```

10. **重新启动备库的日志应用**:
    ```sql
    ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;
    ```

### 4. 日志与归档状态检查

- 检查主库的归档日志状态:
  ```sql
  SELECT SEQUENCE#, APPLIED FROM V$ARCHIVED_LOG ORDER BY SEQUENCE# ASC;
  ```

- 检查备库的归档日志状态:
  ```sql
  SELECT SEQUENCE#, APPLIED FROM V$ARCHIVED_LOG ORDER BY SEQUENCE# ASC;
  ```

### 5. 监控和维护

- 设置定时任务清理老旧归档日志。
- 修改 `control_file_record_keep_time` 为 14，确保控制文件能够记录更多的日志信息。
  ```sql
  ALTER SYSTEM SET control_file_record_keep_time=14;
  ```

### 6. 定时删除归档
```bash
#!/bin/bash

$ORACLE_HOME/bin/rman log=/oradata/rman/deleteArch.log <<EOF
CONNECT TARGET /
RUN {
  CROSSCHECK ARCHIVELOG ALL;
  DELETE NOPROMPT OBSOLETE;
  DELETE FORCE NOPROMPT ARCHIVELOG ALL COMPLETED BEFORE 'SYSDATE-7';
}
EXIT;
```

添加定时任务至 crontab：
```bash
0 1 * * * sh /oradata/rman/deletearch.sh
```

FYF

导航1234

最近发表

友情链接