- 最后登录
- 2013-11-9
- 在线时间
- 9 小时
- 威望
- 16
- 金钱
- 176
- 注册时间
- 2012-6-17
- 阅读权限
- 10
- 帖子
- 36
- 精华
- 0
- 积分
- 16
- UID
- 514
|
2#
发表于 2013-3-13 18:56:04
天哪? Usable_file_MB 为负值,难道会经不起一个disk的损坏吗?
Goal
Checking the value of USABLE_FILE_MB in v$asm_diskgroup it is noticed that the value has a negative value. What could this mean and what could be done given the circumstances?
Fix
Checking the definition in the documentation:
USABLE_FILE_MB = Amount of free space that can be safely utilized taking mirroring into account, and yet be able to restore redundancy after a failgroup failure.
When this value is negative and there is only one disk in your failure group, a disk failure could be critical for the system because it means that there is not enough free space to tolerate a disk failure. If a disk fails, the subsequent rebalance would run out of space before full redundancy could be restored to all files.
As such, this negative value is just a signal that the diskgroup's failgroups need more space in order to tolerate a possible failure of the entire failure group.
Starting 10gR2 this column is used to determine the "Used % of Safely Usable" for a Disk group. This metric generates a warning alert if the Disk Group is using 90% of the Safely usable space and Critical warning for 100%. The threshold limit can be changed to generate alerts at different values.
NOTE: For customers using Oracle Exadata Storage Servers, by default a failure group is a complete cell. This value may be negative as we recommend having enough free space to cover the loss of one disk (quarter and half racks) and two disks for full racks. You do not need to keep the freespace for an entire cell (the value you will see in USABLE_FILE_MB).
|
|