我在 mdadm 镜像中设置了 2 个驱动器(每个驱动器上有 3 个分区)。上周,第二个驱动器 (sdb) 由于一些 I/O 错误而退出了阵列,我仍在调查此问题。服务器重启后,/proc/mdstat 报告阵列降级,正如预期的那样。
Personalities : [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
md2 : active raid1 sda3[0]
1410613056 blocks super 1.0 [2/1] [U_]
bitmap: 11/11 pages [44KB], 65536KB chunk
md1 : active (auto-read-only) raid1 sda2[0]
2097088 blocks super 1.0 [2/1] [U_]
bitmap: 1/1 pages [4KB], 65536KB chunk
md0 : active raid1 sda1[0]
52426624 blocks super 1.0 [2/1] [U_]
bitmap: 1/1 pages [4KB], 65536KB chunk
我已将前两个分区重新添加到相应的阵列中,因为它们很小(50Gb 和 2Gb),并省略了第三个分区,因为我不想等待同步(1.5Tb 同步持续 > 5 小时)。第三个分区的 /var/log/messages 的 grep 如下所示:
2014-08-29T07:02:09.168903+02:00 cube kernel: [ 11.460362] md: bind<sdb3>
2014-08-29T07:02:09.168908+02:00 cube kernel: [ 11.681545] md: bind<sda3>
2014-08-29T07:02:09.168913+02:00 cube kernel: [ 11.692178] md: kicking non-fresh sdb3 from array!
2014-08-29T07:02:09.168917+02:00 cube kernel: [ 11.692209] md: unbind<sdb3>
2014-08-29T07:02:09.168922+02:00 cube kernel: [ 11.700679] md: export_rdev(sdb3)
2014-08-29T07:02:09.168942+02:00 cube kernel: [ 11.706311] md/raid1:md2: active with 1 out of 2 mirrors
2014-08-29T07:02:09.168949+02:00 cube kernel: [ 11.829345] created bitmap (11 pages) for device md2
2014-08-29T07:02:09.168954+02:00 cube kernel: [ 11.877779] md2: bitmap initialized from disk: read 1 pages, set 859 of 21525 bits
2014-08-29T07:02:09.168959+02:00 cube kernel: [ 11.915548] md2: detected capacity change from 0 to 1444467769344
2014-08-28T20:17:30.195301+02:00 cube kernel: [ 10.713443] RAID1 conf printout:
2014-08-28T20:17:30.195306+02:00 cube kernel: [ 10.713459] --- wd:1 rd:2
2014-08-28T20:17:30.195333+02:00 cube kernel: [ 10.713469] disk 0, wo:0, o:1, dev:sda3
2014-08-28T20:17:30.195342+02:00 cube kernel: [ 10.810753] md2: unknown partition table
今天我启动了服务器,mdadm 自行重新添加了第三个分区。一开始我以为我做到了,但我的命令历史记录中没有发出此命令。这可能是一些隐藏的 mdadm 选项/功能吗?今天 /var/log/messages 中的 grep 非常奇怪:
2014-08-31T11:31:42.388226+02:00 cube kernel: [ 13.240221] md: bind<sda3>
2014-08-31T11:31:42.388234+02:00 cube kernel: [ 13.294663] md/raid1:md2: active with 1 out of 2 mirrors
2014-08-31T11:31:42.388238+02:00 cube kernel: [ 13.382804] created bitmap (11 pages) for device md2
2014-08-31T11:31:42.388243+02:00 cube kernel: [ 13.387596] md2: bitmap initialized from disk: read 1 pages, set 859 of 21525 bits
2014-08-31T11:31:42.388248+02:00 cube kernel: [ 13.433121] md2: detected capacity change from 0 to 1444467769344
2014-08-31T11:31:42.388253+02:00 cube kernel: [ 13.445851] md2: unknown partition table
2014-08-31T11:31:42.388258+02:00 cube kernel: [ 13.448115] md: bind<sdb3>
2014-08-31T11:31:42.388276+02:00 cube kernel: [ 13.448230] RAID1 conf printout:
2014-08-31T11:31:42.388283+02:00 cube kernel: [ 13.448237] --- wd:1 rd:2
2014-08-31T11:31:42.388288+02:00 cube kernel: [ 13.448245] disk 0, wo:0, o:1, dev:sda3
2014-08-31T11:31:42.388292+02:00 cube kernel: [ 13.448252] disk 1, wo:1, o:1, dev:sdb3
2014-08-31T11:31:42.388297+02:00 cube kernel: [ 13.448266] RAID1 conf printout:
2014-08-31T11:31:42.388302+02:00 cube kernel: [ 13.448271] --- wd:2 rd:2
2014-08-31T11:31:42.388306+02:00 cube kernel: [ 13.448277] disk 0, wo:0, o:1, dev:sda3
2014-08-31T11:31:42.388374+02:00 cube kernel: [ 13.448284] disk 1, wo:0, o:1, dev:sdb3