Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.debian.user > #240710 > unrolled thread
| Started by | Reiner Buehl <reiner.buehl@gmail.com> |
|---|---|
| First post | 2021-10-01 11:20 +0200 |
| Last post | 2021-10-02 07:40 +0200 |
| Articles | 11 — 5 participants |
Back to article view | Back to linux.debian.user
New mdadm RAID1 gets renamed from md3 to md127 after each reboot Reiner Buehl <reiner.buehl@gmail.com> - 2021-10-01 11:20 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Linux-Fan <Ma_Sys.ma@web.de> - 2021-10-01 16:00 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Reiner Buehl <reiner.buehl@gmail.com> - 2021-10-01 17:40 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Felix Miata <mrmazda@earthlink.net> - 2021-10-01 16:20 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Reiner Buehl <reiner.buehl@gmail.com> - 2021-10-01 17:40 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Tim Woodall <debianuser@woodall.me.uk> - 2021-10-02 00:20 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Joe Pfeiffer <pfeiffer@cs.nmsu.edu> - 2021-10-02 02:00 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Reiner Buehl <reiner.buehl@gmail.com> - 2021-10-02 05:10 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Joe Pfeiffer <pfeiffer@cs.nmsu.edu> - 2021-10-04 04:10 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Reiner Buehl <reiner.buehl@gmail.com> - 2021-10-04 08:40 +0200
Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot Tim Woodall <debianuser@woodall.me.uk> - 2021-10-02 07:40 +0200
| From | Reiner Buehl <reiner.buehl@gmail.com> |
|---|---|
| Date | 2021-10-01 11:20 +0200 |
| Subject | New mdadm RAID1 gets renamed from md3 to md127 after each reboot |
| Message-ID | <D3mRX-3hh-7@gated-at.bofh.it> |
[Multipart message — attachments visible in raw view] — view raw
I created a new mdadm RAID 1 as /dev/md3. But after each reboot, it gets activated as md127. How can I fix this - preferably without haveing to delete the whole array again... The array is defined like this in /etc/mdadm: ARRAY /dev/md3 metadata=1.2 level=raid1 num-devices=1 UUID=41e0a87f:22a2205f:0187c73d:d8ffefea (yes, it only has one disk at the moment, the other will be added once this issue is fixed :-) ) I also did run update-initramfs -u to ensure that this is updated there too. If I manually stop and re-assemble it, it shows as md3, but once I reboot, it shows up as md127 again. In the mdadm --detail --scan output, it shows up like this: ARRAY /dev/md/<Host FQDN>:3 metadata=1.2 name=<Host FQDN>:3 UUID=41e0a87f:22a2205f:0187c73d:d8ffefea How can I stop the renaming? Best regards, Reiner
[toc] | [next] | [standalone]
| From | Linux-Fan <Ma_Sys.ma@web.de> |
|---|---|
| Date | 2021-10-01 16:00 +0200 |
| Subject | Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot |
| Message-ID | <D3reV-5LD-5@gated-at.bofh.it> |
| In reply to | #240710 |
[Multipart message — attachments visible in raw view] — view raw
Reiner Buehl writes: > I created a new mdadm RAID 1 as /dev/md3. But after each reboot, it gets > activated as md127. How can I fix this - preferably without haveing to delete > the whole array again... > The array is defined like this in /etc/mdadm: > > ARRAY /dev/md3 metadata=1.2 level=raid1 num-devices=1 UUID=41e0a87f: > 22a2205f:0187c73d:d8ffefea [...] I have observed this in the past, too and do not know how to "fix" it. Why is it necessary for the volume to appear under /dev/md3? Might it be possible to use its UUID instead, i.e. check the output of ls -l /dev/disk/by-uuid to find out if your md3/md127 can be accessed by an unique ID. You could then point the entries in /etc/fstab to the UUID rather than the "unstable" device name? HTH and YMMV Linux-Fan öö
[toc] | [prev] | [next] | [standalone]
| From | Reiner Buehl <reiner.buehl@gmail.com> |
|---|---|
| Date | 2021-10-01 17:40 +0200 |
| Message-ID | <D3sNJ-6Nj-17@gated-at.bofh.it> |
| In reply to | #240723 |
On 01.10.2021 15:50, Linux-Fan wrote: > I have observed this in the past, too and do not know how to "fix" it. > > Why is it necessary for the volume to appear under /dev/md3? Might it > be possible to use its UUID instead, i.e. check the output of > > ls -l /dev/disk/by-uuid > > to find out if your md3/md127 can be accessed by an unique ID. You > could then point the entries in /etc/fstab to the UUID rather than the > "unstable" device name? It would be beneficial to have the /dev/md3 since I have a few regular expressions that look for only one digit numbers in md arrays but the main reason for me is that I do not trust the array if it randomly renames itself. Mounting via UUID would be possible but I am not sure that a few of the monitoring tools that I use like the UUID.
[toc] | [prev] | [next] | [standalone]
| From | Felix Miata <mrmazda@earthlink.net> |
|---|---|
| Date | 2021-10-01 16:20 +0200 |
| Message-ID | <D3ryh-67d-1@gated-at.bofh.it> |
| In reply to | #240710 |
Reiner Buehl composed on 2021-10-01 11:11 (UTC+0200): > I created a new mdadm RAID 1 as /dev/md3. But after each reboot, it gets > activated as md127. How can I fix this - preferably without haveing to > delete the whole array again... > The array is defined like this in /etc/mdadm: > ARRAY /dev/md3 metadata=1.2 level=raid1 num-devices=1 > UUID=41e0a87f:22a2205f:0187c73d:d8ffefea ... Mine (old) is like so: # head -n3 /etc/mdadm.conf HOMEHOST <ignore> DEVICE containers partitions ARRAY /dev/md0 metadata=1.0 name=msi85:0tmp UUID=... -- Evolution as taught in public schools is, like religion, based on faith, not based on science. Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata
[toc] | [prev] | [next] | [standalone]
| From | Reiner Buehl <reiner.buehl@gmail.com> |
|---|---|
| Date | 2021-10-01 17:40 +0200 |
| Message-ID | <D3sNK-6Nj-39@gated-at.bofh.it> |
| In reply to | #240725 |
On 01.10.2021 16:11, Felix Miata wrote: > Mine (old) is like so: > # head -n3 /etc/mdadm.conf > HOMEHOST <ignore> > DEVICE containers partitions > ARRAY /dev/md0 metadata=1.0 name=msi85:0tmp UUID=... I tried changing the HOMEHOST from <system> to <ignore> but that did not help.
[toc] | [prev] | [next] | [standalone]
| From | Tim Woodall <debianuser@woodall.me.uk> |
|---|---|
| Date | 2021-10-02 00:20 +0200 |
| Subject | Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot |
| Message-ID | <D3z2N-2c0-3@gated-at.bofh.it> |
| In reply to | #240732 |
On Fri, 1 Oct 2021, Reiner Buehl wrote:
> On 01.10.2021 16:11, Felix Miata wrote:
>> Mine (old) is like so:
>> # head -n3 /etc/mdadm.conf
>> HOMEHOST <ignore>
>> DEVICE containers partitions
>> ARRAY /dev/md0 metadata=1.0 name=msi85:0tmp UUID=...
> I tried changing the HOMEHOST from <system> to <ignore> but that did not
> help.
>
>
I've solved this but I don't recall how.
cat /proc/mdstat
Personalities : [raid1] [linear] [multipath] [raid0] [raid6] [raid5]
[raid4] [raid10]
md1 : active raid1 sdb3[2] sda3[3]
468588347 blocks super 1.2 [2/2] [UU]
bitmap: 3/4 pages [12KB], 65536KB chunk
mdadm --misc -D /dev/md1
/dev/md1:
...
Name : xen.home.woodall.me.uk:1
cat /proc/mdstat
Personalities : [raid1] [linear] [multipath] [raid0] [raid6] [raid5]
[raid4] [raid10]
md0 : active raid1 sdb3[2] sda3[3]
468588356 blocks super 1.2 [2/2] [UU]
bitmap: 1/2 pages [4KB], 131072KB chunk
mdadm --misc -D /dev/md0
/dev/md0:
...
Name : boot.home.woodall.me.uk:0
I think it might be that :n at the end of the name.
Another machine doesn't have the :n and there its device md127.
(those machine names bring back memories. I'd rename the array except no
doubt everything would collapse in a heap...)
[toc] | [prev] | [next] | [standalone]
| From | Joe Pfeiffer <pfeiffer@cs.nmsu.edu> |
|---|---|
| Date | 2021-10-02 02:00 +0200 |
| Subject | Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot |
| Message-ID | <D3ABz-2Xh-3@gated-at.bofh.it> |
| In reply to | #240762 |
Tim Woodall <debianuser@woodall.me.uk> writes: > On Fri, 1 Oct 2021, Reiner Buehl wrote: > >> On 01.10.2021 16:11, Felix Miata wrote: >>> Mine (old) is like so: >>> # head -n3 /etc/mdadm.conf >>> HOMEHOST <ignore> >>> DEVICE containers partitions >>> ARRAY /dev/md0 metadata=1.0 name=msi85:0tmp UUID=... >> I tried changing the HOMEHOST from <system> to <ignore> but that did >> not help. >> >> > > I've solved this but I don't recall how. Same here... FWIW my /etc/mdadm/mdadm.conf contains the line ARRAY /dev/md/1 metadata=1.2 UUID=67d3c233:96a0737c:5f88ed9b:936ea3ae name=snowball:1 and my array is indeed /dev/md1 (I've got *no* recollection why I didn't make it md0!). Is it possible you need to update your init ramdisk? Maybe your changes to your mdadm.conf aren't being seen?
[toc] | [prev] | [next] | [standalone]
| From | Reiner Buehl <reiner.buehl@gmail.com> |
|---|---|
| Date | 2021-10-02 05:10 +0200 |
| Message-ID | <D3Dzr-4Zl-1@gated-at.bofh.it> |
| In reply to | #240766 |
On 02.10.2021 01:32, Joe Pfeiffer wrote: > Is it possible you need to update your init ramdisk? Maybe your changes > to your mdadm.conf aren't being seen? I do run update-initramfs -u after each change. Shouldn't that be enough to to update the mdadm.conf in the init ramdisk?
[toc] | [prev] | [next] | [standalone]
| From | Joe Pfeiffer <pfeiffer@cs.nmsu.edu> |
|---|---|
| Date | 2021-10-04 04:10 +0200 |
| Message-ID | <D4lAt-6Z0-1@gated-at.bofh.it> |
| In reply to | #240771 |
Reiner Buehl <reiner.buehl@gmail.com> writes: > On 02.10.2021 01:32, Joe Pfeiffer wrote: >> Is it possible you need to update your init ramdisk? Maybe your changes >> to your mdadm.conf aren't being seen? > > I do run update-initramfs -u after each change. Shouldn't that be > enough to to update the mdadm.conf in the init ramdisk? That's not it then... googling "md127" I found a *bunch* of possible causes.
[toc] | [prev] | [next] | [standalone]
| From | Reiner Buehl <reiner.buehl@gmail.com> |
|---|---|
| Date | 2021-10-04 08:40 +0200 |
| Message-ID | <D4pNL-18Y-5@gated-at.bofh.it> |
| In reply to | #240868 |
On 04.10.2021 03:41, Joe Pfeiffer wrote: > That's not it then... googling "md127" I found a *bunch* of possible > causes. I tried already updating the name attribute, setting the homehost to none and a few others. Unfortunately none helped.
[toc] | [prev] | [next] | [standalone]
| From | Tim Woodall <debianuser@woodall.me.uk> |
|---|---|
| Date | 2021-10-02 07:40 +0200 |
| Subject | Re: New mdadm RAID1 gets renamed from md3 to md127 after each reboot |
| Message-ID | <D3FUB-6sB-1@gated-at.bofh.it> |
| In reply to | #240766 |
On Fri, 1 Oct 2021, Joe Pfeiffer wrote: > Tim Woodall <debianuser@woodall.me.uk> writes: > >> On Fri, 1 Oct 2021, Reiner Buehl wrote: >> >>> On 01.10.2021 16:11, Felix Miata wrote: >>>> Mine (old) is like so: >>>> # head -n3 /etc/mdadm.conf >>>> HOMEHOST <ignore> >>>> DEVICE containers partitions >>>> ARRAY /dev/md0 metadata=1.0 name=msi85:0tmp UUID=... >>> I tried changing the HOMEHOST from <system> to <ignore> but that did >>> not help. >>> >>> >> >> I've solved this but I don't recall how. > > Same here... FWIW my /etc/mdadm/mdadm.conf contains the line > ARRAY /dev/md/1 metadata=1.2 UUID=67d3c233:96a0737c:5f88ed9b:936ea3ae name=snowball:1 > > and my array is indeed /dev/md1 (I've got *no* recollection why I didn't > make it md0!). > > Is it possible you need to update your init ramdisk? Maybe your changes > to your mdadm.conf aren't being seen? > Don't think it's anything to do with the config file: root@debootstrap17:~# mdadm --examine --scan ARRAY /dev/md/debootstrap17 metadata=1.2 UUID=5e7d0dca:9f579eef:afb81eec:b0060281 name=debootstrap17 ARRAY /dev/md/5 metadata=1.2 UUID=1cdac5e5:c9fcac40:5e738a68:0a650b23 name=test:5 ARRAY /dev/md/test5 metadata=1.2 UUID=46ced7db:4b422364:1938cce3:6561f99d name=test5 That's with no arrays defined in the config file. After these commands: dd if=/dev/zero of=mdadm.blk bs=1M count=50 losetup -f mdadm.blk mdadm --create /dev/md/test:5 --metadata=1.2 --bitmap=internal --force --level=1 --raid-devices=1 /dev/loop0 mdadm --stop /dev/md/test\:5 dd if=/dev/zero of=mdadm2.blk bs=1M count=50 losetup -f mdadm2.blk mdadm --create /dev/md/test5 --metadata=1.2 --bitmap=internal --force --level=1 --raid-devices=1 /dev/loop1 mdadm --stop /dev/md/test5 mdadm --examine --scan
[toc] | [prev] | [standalone]
Back to top | Article view | linux.debian.user
csiph-web