1 |
On Sun, 2012-01-08 at 15:03 -0500, Jeff Cranmer wrote: |
2 |
> On Sun, 2012-01-08 at 12:31 -0600, Paul Hartman wrote: |
3 |
> > > |
4 |
> > > What is going on here? |
5 |
> > |
6 |
> > (I didn't read this whole thread, sorry if I'm repeating someone else's |
7 |
> > advice) |
8 |
> > |
9 |
> > kernel autodetection only works on old superblock version 0.90, you're |
10 |
> > using 1.2. Not a big deal, we use mdadm to do it. |
11 |
> > |
12 |
> > Define your arrays in /etc/mdadm.conf and start /etc/init.d/mdadm in |
13 |
> > your boot runscripts with "rc-update add mdadm boot", it will bring up |
14 |
> > the array at boot time. |
15 |
> > |
16 |
> > In my mdadm.conf i have a line like this: |
17 |
> > |
18 |
> > ARRAY /dev/md1 metadata=1.01 name=black:1 |
19 |
> > UUID=8e653e72:9d5df6ba:bb66ea8b:02f1c317 |
20 |
> > |
21 |
> > (might be word-wrapped, should be all one line) |
22 |
> > |
23 |
> > That's all that was needed to bring it up automatically at boot time. |
24 |
> > |
25 |
> > Also AFAIR there was a "gotcha" about the hostname stored in the array's |
26 |
> > metadata must match your machine's hostname or else mdadm auto-assemble |
27 |
> > won't accept it (to protect you in case you're plugging disks from |
28 |
> > another machine for recovery, you don't want it to use them as your main |
29 |
> > drives), so in that case you must specify it explicitly or set the AUTO |
30 |
> > parameter in mdadm.conf to accept this condition. If you created the |
31 |
> > array from within a LiveCD or on another machine, the hostname might not |
32 |
> > match your system. |
33 |
> > |
34 |
> > See the mdadm manpage for more info. |
35 |
> |
36 |
> mdadm was added to the default level, not boot. |
37 |
> My /etc/mdadm.conf file has two active lines |
38 |
> DEVICE /dev/sd[bcd]1 |
39 |
> ARRAY dev/md0 metadata=1.2 spares=1 name=office-desktop:0 |
40 |
> devices=/dev/sdb1,dev/sdc1,/dev/sdd1 |
41 |
> |
42 |
> It looks like I'm having trouble with a faulty /dev/sdc1, so what I'd |
43 |
> like to do is wipe out the existing array and try starting a RAID1 array |
44 |
> just with sdb1 and sdd1. |
45 |
> |
46 |
> I got rid of the old array by using the commands |
47 |
> mdadm --manage --fail /dev/md0 |
48 |
> mdadm --manage --stop /dev/md0 |
49 |
> |
50 |
> I then used mdadm --verbose --create /dev/md0 --level=1 |
51 |
> --raid-devices=2 /dev/sdb1 /dev/sdd1 |
52 |
> |
53 |
> The result of this command was |
54 |
> dadm: /dev/sdb1 appears to be part of a raid array: |
55 |
> level=raid5 devices=3 ctime=Sat Jan 7 08:16:00 2012 |
56 |
> mdadm: partition table exists on /dev/sdb1 but will be lost or |
57 |
> meaningless after creating array |
58 |
> mdadm: Note: this array has metadata at the start and |
59 |
> may not be suitable as a boot device. If you plan to |
60 |
> store '/boot' on this device please ensure that |
61 |
> your boot-loader understands md/v1.x metadata, or use |
62 |
> --metadata=0.90 |
63 |
> mdadm: /dev/sdd1 appears to be part of a raid array: |
64 |
> level=raid5 devices=3 ctime=Sat Jan 7 08:16:00 2012 |
65 |
> mdadm: size set to 1465136400K |
66 |
> Continue creating array? y |
67 |
> mdadm: Defaulting to version 1.2 metadata |
68 |
> mdadm: array /dev/md0 started. |
69 |
> |
70 |
> The results of cat /proc/mdstat are |
71 |
> Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] |
72 |
> [raid4] [multipath] |
73 |
> md0 : active raid1 sdd1[1] sdb1[0] |
74 |
> 1465136400 blocks super 1.2 [2/2] [UU] |
75 |
> [>....................] resync = 2.1% (31838144/1465136400) |
76 |
> finish=269.7min speed=88551K/sec |
77 |
> |
78 |
> unused devices: <none> |
79 |
> |
80 |
> Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] |
81 |
> [raid4] [multipath] |
82 |
> md0 : active raid1 sdd1[1] sdb1[0] |
83 |
> 1465136400 blocks super 1.2 [2/2] [UU] |
84 |
> [>....................] resync = 2.1% (31838144/1465136400) |
85 |
> finish=269.7min speed=88551K/sec |
86 |
> |
87 |
> unused devices: <none> |
88 |
> |
89 |
> The results of mdadm --detail /dev/md0 are |
90 |
> /dev/md0: |
91 |
> Version : 1.2 |
92 |
> Creation Time : Sun Jan 8 14:47:43 2012 |
93 |
> Raid Level : raid1 |
94 |
> Array Size : 1465136400 (1397.26 GiB 1500.30 GB) |
95 |
> Used Dev Size : 1465136400 (1397.26 GiB 1500.30 GB) |
96 |
> Raid Devices : 2 |
97 |
> Total Devices : 2 |
98 |
> Persistence : Superblock is persistent |
99 |
> |
100 |
> Update Time : Sun Jan 8 14:48:54 2012 |
101 |
> State : active, resyncing |
102 |
> Active Devices : 2 |
103 |
> Working Devices : 2 |
104 |
> Failed Devices : 0 |
105 |
> Spare Devices : 0 |
106 |
> |
107 |
> Rebuild Status : 2% complete |
108 |
> |
109 |
> Name : office-desktop:0 (local to host office-desktop) |
110 |
> UUID : bfc16c6e:4e8cb910:96ff7ed2:6fec32bc |
111 |
> Events : 1 |
112 |
> |
113 |
> Number Major Minor RaidDevice State |
114 |
> 0 8 17 0 active sync /dev/sdb1 |
115 |
> 1 8 49 1 active sync /dev/sdd1 |
116 |
> |
117 |
> When I try to mount this drive, however, I get |
118 |
> mount: /dev/md0: can't read superblock |
119 |
> |
120 |
> What do I need to do to complete the process? |
121 |
> |
122 |
> Thanks |
123 |
> |
124 |
> Jeff |
125 |
> |
126 |
> |
127 |
> |
128 |
Success - I managed to get a raid1 device operating. |
129 |
I created the final filesystem by using mkfs.xfs -f /dev/md0, then |
130 |
waited for the rebuild to complete before rebooting the system. |
131 |
|
132 |
It appears to be created successfully. Now I'll try the same sequence |
133 |
with sdb and sdc to see if sdc is a good disk. If that works, I'll |
134 |
retry a raid5 array tomorrow night. |
135 |
|
136 |
Jeff |