From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bill Davidsen Subject: Re: Deleting mdadm RAID arrays Date: Fri, 08 Feb 2008 07:44:18 -0500 Message-ID: <47AC4EA2.4010805@tmr.com> References: <200802051142.19625.admin@domeny.pl> <200802071056.33221.admin@domeny.pl> <47AB79B1.6000503@tmr.com> <200802081035.25722.admin@domeny.pl> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <200802081035.25722.admin@domeny.pl> Sender: linux-raid-owner@vger.kernel.org To: Marcin Krol Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids Marcin Krol wrote: > Thursday 07 February 2008 22:35:45 Bill Davidsen napisa=C5=82(a): > =20 >>> As you may remember, I have configured udev to associate /dev/d_* d= evices with >>> serial numbers (to keep them from changing depending on boot module= loading=20 >>> sequence).=20 >>> =20 > > =20 >> Why do you care?=20 >> =20 > > Because /dev/sd* devices get swapped randomly depending on boot modul= e insertion > sequence, as I explained earlier. > > =20 So there's no functional problem, just cosmetic? >> If you are using UUID for all the arrays and mounts =20 >> does this buy you anything?=20 >> =20 > > This is exactly what is not clear for me: what is it that identifies = drive/partition as part of=20 > the array? /dev/sd name? UUID as part of superblock? /dev/d_n? > > If it's UUID I should be safe regardless of /dev/sd* designation? Yes= or no? > > =20 Yes, absolutely. >> And more to the point, the first time a =20 >> drive fails and you replace it, will it cause you a problem? Require= =20 >> maintaining the serial to name data manually? >> =20 > > That's not the problem. I just want my array to be intact. > > =20 >> I miss the benefit of forcing this instead of just building the=20 >> information at boot time and dropping it in a file. >> =20 > > I would prefer that, too - if it worked. I was getting both arrays me= ssed=20 > up randomly on boot. "messed up" in the sense of arrays being compose= d > of different /dev/sd devices. > > =20 Different devices? Or just different names for the same devices? I=20 assume just the names change, and I still don't see why you care...=20 subtle beyond my understanding. > =20 >>> And I made *damn* sure I zeroed all the superblocks before reassemb= ling=20 >>> the arrays. Yet it still shows the old partitions on those arrays! >>> =20 >>> =20 >> As I noted before, you said you had these on whole devices before, d= id=20 >> you zero the superblocks on the whole devices or the partitions? Fro= m=20 >> what I read, it was the partitions. >> =20 > > I tried it both ways actually (rebuilt arrays a few times, just udev = didn't want > to associate WD-serialnumber-part1 as /dev/d_1p1 as it was told, it s= till claimed > it was /dev/d_1).=20 > =20 I'm not talking about building the array, but zeroing the superblocks.=20 Did you use the partition name, /dev/sdb1, when you ran mdadm with=20 "zero-super" or did you zero the whole device, /dev/sdb, which is what=20 you were using when you first built the array with whole devices. If yo= u=20 didn't zero the superblock for the whole device it may explain why a=20 superblock is still found. --=20 Bill Davidsen "Woe unto the statesman who makes war without a reason that will stil= l be valid when the war is over..." Otto von Bismark=20 - To unsubscribe from this list: send the line "unsubscribe linux-raid" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html