1、問題現象:
ceph -s
HEALTH_WARN 14 requests are blocked > 32 sec; 11 osds have slow requests
2、問題分析:
ceph health detail 查看有哪些OSD或者PGs卡住。
HEALTH_WARN 14 requests are blocked > 32 sec; 11 osds have slow requests
7 ops are blocked > 536871 sec
2 ops are blocked > 268435 sec
2 ops are blocked > 67108.9 sec
3 ops are blocked > 33554.4 sec
1 ops are blocked > 536871 sec on osd.0
1 ops are blocked > 536871 sec on osd.10
2 ops are blocked > 536871 sec on osd.12
2 ops are blocked > 268435 sec on osd.18
1 ops are blocked > 536871 sec on osd.31
1 ops are blocked > 536871 sec on osd.38
1 ops are blocked > 67108.9 sec on osd.38
1 ops are blocked > 33554.4 sec on osd.48
1 ops are blocked > 67108.9 sec on osd.52
1 ops are blocked > 536871 sec on osd.63
1 ops are blocked > 33554.4 sec on osd.64
1 ops are blocked > 33554.4 sec on osd.69
11 osds have slow requests
3、處理過程:
對上述OSD進行重啓,切記需要一個一個重啓,每重啓一個OSD,待數據平衡完成之後,再執行下一個OSD重啓。