闪回恢复区中添加或删除文件等变化都将记录在数据库的 alert 日志中,Oracle 10g 也针对该新特性提供了一个新的视图, DBA_OUTSTANDING_ALERTS,通过该视图可以得到相关的信息。
SQL> DESC dba_outstanding_alerts
名称 是否为空? 类型
----------------------------------------- -------- ----------------------------
SEQUENCE_ID NUMBER
REASON_ID NOT NULL NUMBER
OWNER VARCHAR2(30)
OBJECT_NAME VARCHAR2(513)
SUBOBJECT_NAME VARCHAR2(30)
OBJECT_TYPE VARCHAR2(64)
REASON VARCHAR2(4000)
TIME_SUGGESTED TIMESTAMP(6) WITH TIME ZONE
CREATION_TIME TIMESTAMP(6) WITH TIME ZONE
SUGGESTED_ACTION VARCHAR2(4000)
ADVISOR_NAME VARCHAR2(30)
METRIC_VALUE NUMBER
MESSAGE_TYPE VARCHAR2(12)
MESSAGE_GROUP VARCHAR2(64)
MESSAGE_LEVEL NUMBER
HOSTING_CLIENT_ID VARCHAR2(64)
MODULE_ID VARCHAR2(64)
PROCESS_ID VARCHAR2(128)
HOST_ID VARCHAR2(256)
HOST_NW_ADDR VARCHAR2(256)
INSTANCE_NAME VARCHAR2(16)
INSTANCE_NUMBER NUMBER
USER_ID VARCHAR2(30)
EXECUTION_CONTEXT_ID VARCHAR2(128)
ERROR_INSTANCE_ID VARCHAR2(142)
在闪回恢复区中的空间使用超过 85% 的时候,数据库将会向 alert 文件中写入告警信息。而当超过 97% 的时候将会写入严重告警信息。当闪回恢复区空间不够的时候,Oracle将报告如下类似的错误:
ORA-19809: limit exceeded for recovery files
ORA-19804: cannot reclaim 52428800 bytes disk space from 1258291200 limit
这个时候查询 dba_outstanding_alerts:
SQL> select reason,object_type,suggested_action from dba_outstanding_alerts;
REASON OBJECT_TYPE SUGGESTED_ACTION
------------------------------ -------------------- ----------------------------------------
db_recovery_file_dest_size of RECOVERY AREA Add disk space and increase db_recovery_
1258291200 bytes is 88.20% use file_dest_size, backup files to tertiary
d and has 148509184 remaining device, delete files from recovery area
bytes available. using RMAN, consider changing RMAN rete
ntion policy or consider changing RMAN a
rchivelog deletion policy.
同时,oracle在alert中还会给出解决该问题的建议
************************************************************************
You have following choices to free up space from flash recovery area:
1. Consider changing RMAN RETENTION POLICY. If you are using Data Guard,then consider changing RMAN ARCHIVELOG DELETION POLICY.
2. Back up files to tertiary device such as tape using RMANB ACKUP RECOVERY AREA command.
3. Add disk space and increase db_recovery_file_dest_size parameter to reflect the new space.
4. Delete unnecessary files using RMAN DELETE command. If an operating system command was used to delete files, then use RMAN CROSSCHECK and DELETE EXPIRED commands.
************************************************************************
V$RECOVERY_FILE_DEST视图 包含闪回恢复区的相关信息:
SQL> desc V$RECOVERY_FILE_DEST
名称 是否为空? 类型
----------------------------------------- -------- ---------------------
NAME VARCHAR2(513)
SPACE_LIMIT NUMBER
SPACE_USED NUMBER
SPACE_RECLAIMABLE NUMBER
NUMBER_OF_FILES NUMBER
SQL> select * from v$recovery_file_dest;
NAME SPACE_LIMIT SPACE_USED SPACE_RECLAIMABLE NUMBER_OF_FILES
---------- ----------- ---------- ----------------- ---------------
4039114752 0 0 0
通过查询视图v$flash_recovery_area_usage,可以获得当前闪回恢复区的空间使用情况,并且可以知道是哪些文件占中了空间,据此可以做出相应的处理,或者加大闪回恢复区,或者移走相应的文件。
SQL> select * from v$flash_recovery_area_usage;
FILE_TYPE PERCENT_SPACE_USED PERCENT_SPACE_RECLAIMABLE
-------------------- ------------------ -------------------------
CONTROL FILE 0 0
REDO LOG 0 0
ARCHIVED LOG 0 0
BACKUP PIECE 0 0
IMAGE COPY 0 0
FLASHBACK LOG 0 0
FOREIGN ARCHIVED LOG 0 0
如果闪回恢复区空间耗尽,且归档路径设置到了闪回恢复区中,则由于日志无法归档,数据库会hang住。所以,对于生产库,如果将归档放到闪回恢复区中,需要密切关注闪回恢复区的空间使用情况,否则一旦闪回恢复区的空间用尽,将导致数据库无法提供服务。