another RAID / md question

jan128 · 11-11-2020, 08:26 PM

I received series of attached messaged during recent "upgrade".
I am not currently working on cleaning-up my RAID arrays, however I am not sure
which file should be "master" . I was under the impression that /etc/mdadm/mdadm.conf is.

If that is the case - /usr/share/mdadm/mkconf cannot be matching
/etc/mdadm/mdadm.conf - because mdadm.conf is missing some configuration.

No, I have not , not yet, compared the files.
But if md127 is not in mdadm.conf the error is to be expected - so if I manually add to mdadm.conf the error will go away,right ?

Not really a question to the forum , just confirmation would be sufficient.

Code:

W: mdadm: the array /dev/md127 with UUID 5afe531e:3c41ecd9:7057ae6e:98985049
W: mdadm: is currently active, but it is not listed in mdadm.conf. if
W: mdadm: it is needed for boot, then YOUR SYSTEM IS NOW UNBOOTABLE!
W: mdadm: please inspect the output of /usr/share/mdadm/mkconf, compare
W: mdadm: it to /etc/mdadm/mdadm.conf, and make the necessary changes.

syg00 · 11-11-2020, 08:32 PM

And ?.

computersavvy · 11-12-2020, 11:00 AM

If you follow instructions given by the system the best that can happen is you fix it and learn something at the same time. The worst is you totally bork it and need to reinstall. Doing nothing except ask questions gains you nothing - either in system status or knowledge.

rnturn · 11-14-2020, 12:16 PM

Quote:

Originally Posted by computersavvy

If you follow instructions given by the system the best that can happen is you fix it and learn something at the same time. The worst is you totally bork it and need to reinstall.

Re-installing seems like overkill. Couldn't our OP boot from a rescue CD or DVD and make corrections to "/etc/mdadm.conf". Heck, you could read the "mdadm.conf" manpage and use the output of "mdadm --detail /dev/mdN" to write it from scratch. The contents of "mdadm.conf" aren't critical for booting, anyway (at least not on Slackware---the contents of that file on one of my SW systems -- that boots from /dev/md0 -- is all comments) but is handy to remind yourself six months after building it what disks/partitions make up a given RAIDset and, if memory serves, it might need to be correct if you ever need to manipulate the RAIDset(s) using mdadm.

The OP would find it informative to read and re-read the mdadm(8) manpage. There's a lot of good material on that page.