On Monday September 4, bothie@xxxxxx wrote: > Neil Brown <neilb@xxxxxxx> wrote: > > > On Sunday September 3, bothie@xxxxxx wrote: > > > >> I have a really really big problem. In fact, the problem is the output of > >> mdadm --examine as shown on http://nomorepasting.com/paste.php?pasteID=68021 > > > > Please explain why you think that output is a problem. It looks fine > > to me. > > 1. craida1 claims about itself to be raidh-raidh1 ... Thanks. I didn't notice that (there was a lot of detail to wade through). Presumably /dev/mapper/raidh-raidh1 has major/minor of 254,20, but that seems odd.. What does ls -l /dev/mapper/raidh-raidh1 /dev/mapper/craida1 show? > > 2. according to craida1 the raid is composed of raidh-raidh1, a faulty disk > (that was craid1b), craidb1, craidc1 and craida2, but not craida1 itself, > of course ^^^^^^^ For craida1 I see: 49 50 Number Major Minor RaidDevice State 51 this 0 254 20 0 active sync /dev/mapper/raidh-raidh1 52 53 0 0 254 20 0 active sync /dev/mapper/raidh-raidh1 54 1 1 0 0 1 faulty removed 55 2 2 254 22 2 active sync /dev/mapper/craidb1 56 3 3 254 23 3 active sync /dev/mapper/craidc1 57 4 4 254 24 4 active sync /dev/mapper/craidd1 so the raidh-raidh1 - craida1 confusion is still there, but I don't see craida2 at all. > > 2b. Continue with craida2. It claims to be craide1 and being part of > craide1, craidh1, craida2, craidc2, craidd2. Ok, I see the issue now. The device names reported by 'mdadm -E' are based on the major/minor numbers that are stored in the superblock. They are only meaningful if major/minor device numbers are stable. Presumably dm minor numbers are very unstable, in that if you take down a target and the reassemble it could very well have a different minor number. So the information in the superblock is out-of-date. Don't worry. The major/minor information in the superblock is ignored except by "mdadm -E". It is not used when assembling arrays. So this output does look weird, but it doesn't indicate a problem. > > PS: You may stop putting me CC as I finally managed successfully to join the > list ;) I didn't Cc you. I sent the mail to you, and Cced the list, as I always do. (It is, in general, the safest thing to do). Hopefully getting two copies of the reply wont be a major inconvenience, and if it is I believe procmail can be configured to remove such duplicates. NeilBrown -- VGER BF report: H 0.0361865 - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
