通过ADDM嗅到存储硬盘故障

今天ADDM巡检发现出现问题：Finding  The throughput of the I/O subsystem was significantly lower than expected

该问题从来未出现过，立即引起笔者的警觉，展开 如下 相关项发现多个裸设备同时出现IO异常的告警，而按笔者所在的业务系统，该时段显然未进入一天的业务最高锋，而这个问题是以往哪怕是节前最高峰也从未出现的。马上要求系统工程师确认存储子系统有无问题，答复是“远程管理口未接上”。当天下班后笔者强烈的直觉感觉到可能存在存储异常状况，决定前往IDC机房巡检查看存储系统。到IDC居然发现由于临时太急，存储的钥匙也未带上，后通过存储柜门的小孔透视发现一块磁盘亮黄灯。于是立即向系统工程师反馈这一故障，当然我们的存储由于RAID+HOTSPARE结构,即使坏两块盘也不丢数据 。

最后分析应该是该块磁盘故障导致IO临时异常，提醒大家,ADDM中观测到大量的裸设备或文件系统异常时一定要关注磁盘有无异常状况。

后续改进措施:要求存储系统接上远程管理口，便于远程检查，以笔者所在机房为例，打车28元，时间至少半个小时以上，如果有远程管理口，这部分时间和金钱显然可以省下来

 

Finding  The throughput of the I/O subsystem was significantly lower than expected.
Impact (minutes)  32.2
Impact (%)  27.5

 Recommendations



Show All Details | Hide All Details

Details Category Benefit (%)
Hide Host Configuration  27.5
Action  Consider increasing the throughput of the I/O subsystem. Oracle's recommended solution is to stripe all data file using the SAME methodology. You might also need to increase the number of disks for better performance. Alternatively, consider using Oracle's Automatic Storage Management solution.

Rationale  During the analysis period, the average data files' I/O throughput was 898 K per second for reads and 40 K per second for writes. The average response time for single block reads was 19 milliseconds.


Hide Host Configuration  24.2
Action  The performance of file /dev/rgaza_disk was significantly worse than other files. If striping all files using the SAME methodology is not possible, consider striping this file over multiple disks.

Rationale  The average response time for single block reads for this file was 112 milliseconds.


Hide Host Configuration  1
Action  The performance of file /dev/rsystem_disk was significantly worse than other files. If striping all files using the SAME methodology is not possible, consider striping this file over multiple disks.

Rationale  The average response time for single block reads for this file was 206 milliseconds.


Hide Host Configuration  0.8
Action  The performance of file /dev/rdata35_disk was significantly worse than other files. If striping all files using the SAME methodology is not possible, consider striping this file over multiple disks.

Rationale  The average response time for single block reads for this file was 527 milliseconds.


Hide Host Configuration  0.6
Action  The performance of file /dev/rtemp1_disk was significantly worse than other files. If striping all files using the SAME methodology is not possible, consider striping this file over multiple disks.

Rationale  The average response time for single block reads for this file was 34 milliseconds.



 Findings Path