[00:15] <FuZi0N> I have a folder with around 4500 subdirectories. How can I delete all subdirectories that are smaller than 100 MB?
[01:01] <JanC> I'm pretty sure all those subdirectories are smaller than 100MB  :)
[01:02] <JanC> but I guess you mean the files inside them
[01:03] <JanC> you could script it in bash using a tool like 'du'
[08:25] <zetheroo> I need help with getting an mdadm raid to boot. power was lost on the server and when restored the OS would not boot. After POST all I see is a blinking cursor on a black screen.
[08:28] <zetheroo> I did 'grub-install' on both disks from an Ubuntu live session but now get errors when booting and end up in grub rescue
[08:30] <RoyK> zetheroo: first up, which versions of things are you running? can you see both drives if booting up on a rescue usb thing? do you see any issues in smart data on the drives?
[08:30] <RoyK> also, do you have a backup?
[08:31] <zetheroo>  RoyK: I have both disks (SSDs) attached to my PC via USB dock at the moment
[08:31] <zetheroo> https://pastebin.com/K62C604U
[08:31] <zetheroo> the disks are both less than a month old
[08:32] <zetheroo> one of the errors when booting is 'unknown filesystem'
[08:36] <zetheroo> no, we don't have a current backup
[08:38] <RoyK> what does lsblk have to say?
[08:38] <RoyK> the age of the drives is generally irrelvant - drives can fail at any time
[08:39] <zetheroo> https://pastebin.com/ePn3BQ6t
[08:39] <RoyK> please pastebin output of "smartctl -a /dev/sdX"
[08:40] <RoyK> erm - nothing on the raid? is lvm installed on the machine?
[08:40] <RoyK> also, please don't use pastebin.com - it's spammy - https://paste.ubuntu.com/ or similar is quieter and works well ;)
[08:41] <zetheroo> https://paste.ubuntu.com/p/vff6PvF8bW/
[08:42] <zetheroo> not sure about the LVM question - but very likely.
[08:42] <RoyK> type 'vgs'
[08:43] <zetheroo> no output
[08:43] <RoyK> looks like the drives are new-ish - only been running for a week or so
[08:43] <zetheroo> lvm2 was not installed on my PC so just installed it
[08:43] <zetheroo> yep
[08:44] <RoyK> pvscan
[08:44] <zetheroo>   No matching physical volumes found
[08:44] <RoyK> just to check if there's some lvm stuff there
[08:44] <RoyK> oh
[08:44] <RoyK> could it be zfs?
[08:44] <zetheroo> hmm ...
[08:44] <RoyK> not likely
[08:44] <RoyK> since there's md0
[08:45] <RoyK> there really should be something visible there
[08:45] <RoyK> was this setup with crypto or something?
[08:45] <zetheroo> no crypto
[08:51] <zetheroo> is there a way to mount the disk?
[08:52] <zetheroo> maybe it's better to get all the critical data out and just do a fresh install
[08:52] <RoyK> there should be, but before that, you should see partitions/filesystems
[08:52] <RoyK> and it's very strange that lsblk shows none
[08:53] <RoyK> erm
[08:53] <RoyK> osprober-linux-sda1
[08:53] <RoyK> what is that?
[08:54] <zetheroo> no idea ... looking it up
[08:54] <zetheroo> https://github.com/openzfs/zfs/discussions/9801
[08:54] <zetheroo> ZOL ...?
[08:55] <RoyK> I hope not
[08:55] <RoyK> zfs on top of mdraid sounds like a bad dream
[08:55] <RoyK> or a badtrip
[08:56] <RoyK> what does 'file -s /dev/md0' have to say?
[08:57] <zetheroo> /dev/md0: Linux rev 1.0 ext4 filesystem data, UUID=99f7fdf9-6f9a-4069-856e-4db4c529741e (needs journal recovery) (extents) (64bit) (large files) (huge files)
[08:57] <RoyK> hm - try to mount it somewhere
[08:58] <zetheroo> should I mount md0 or the individual disks?
[08:58] <RoyK> md0
[08:58] <RoyK> the mirror seems to be in sync after all
[09:01] <zetheroo> mount: /mnt/raidarray: cannot mount /dev/md0 read-only.
[09:01] <zetheroo> mounting via fstab with:
[09:01] <zetheroo> UUID=99f7fdf9-6f9a-4069-856e-4db4c529741e  /mnt/raidarray ext4 defaults 0 0
[09:02] <oerheks> read only, sounds like a dirty filesystem
[09:04] <RoyK> oerheks: it is - it said above - power failure
[09:04] <zetheroo> https://paste.ubuntu.com/p/J7BwF4RMm6/
[09:05] <RoyK> zetheroo: any valuable data on it?
[09:05] <zetheroo> yep  -  it was our monitoring server :(
[09:05] <RoyK> zetheroo: not to rub it in, but try to use a backup next time ;)
[09:05] <zetheroo> and it had no redundant PSU 😲
[09:05] <RoyK> zetheroo: keep a backup anyway
[09:06] <RoyK> two backups, preferably, at least
[09:06] <zetheroo> yes, I know ... very bad practice with that :(
[09:06] <RoyK> anyway
[09:06] <RoyK> what does /proc/mdstat have to say now?
[09:07] <zetheroo> https://paste.ubuntu.com/p/gp4KnDYSSs/
[09:07] <RoyK> you might need to issue a raid sync/repair
[09:07] <RoyK> echo repair > /sys/block/md0/md/sync_action
[09:08] <zetheroo> bash: /sys/block/md0/md/sync_action: Permission denied
[09:08] <RoyK> 'sudo -i' first
[09:08] <zetheroo> ok, no output
[09:08] <RoyK> check mdstat
[09:09] <RoyK> it shouldn't take too long on a small ssd mirror
[09:09] <zetheroo> https://paste.ubuntu.com/p/sNhRTr4PNh/
[09:10] <zetheroo> https://paste.ubuntu.com/p/7WcdJM8xPT/
[09:12] <RoyK> try to stop it and reassemble
[09:12] <RoyK> mdadm --stop /dev/md0 ; mdadm --asseble --scan
[09:12] <zetheroo> mdadm: /dev/md/0 has been started with 2 drives.
[09:13] <RoyK> checkk mdstat
[09:13] <zetheroo> md0 : active (read-only)  ...
[09:13] <RoyK> anything interesting in dmesg?
[09:13] <RoyK> dmesg -T to get proper timestamps
[09:14] <zetheroo> https://paste.ubuntu.com/p/ZGPRcFTDvC/
[09:17] <RoyK> stop it again and try 'mdadm --assemble /dev/md0 --run'
[09:17] <RoyK> never seen it go into read-only mode like that before
[09:17] <zetheroo> oh ... I got it to mount
[09:17] <zetheroo> at least I think so ... it's mounted read-only though
[09:18] <zetheroo> UUID=99f7fdf9-6f9a-4069-856e-4db4c529741e	/mnt/raidarray	ext4	noload	0	0
[09:18] <zetheroo> used the noload option
[09:18] <zetheroo> is that ok, or should I unmount it?
[09:18] <RoyK> good, get the data off the system first
[09:18] <zetheroo> after mount - a I see 'mount: /mnt/raidarray: WARNING: device write-protected, mounted read-only.'
[09:19] <RoyK> did you try with --run?
[09:19] <zetheroo> no, should I unmount it and do that?
[09:20] <RoyK> try to umount, stop and reassemble with --run to see if it gets up without readonly
[09:20] <zetheroo> ok
[09:21] <zetheroo> md0 : active (read-only)
[09:21] <RoyK> esc
[09:21] <RoyK> sec
[09:21] <zetheroo> k
[09:23] <RoyK> hm - you might want to try mdadm -w /dev/md0
[09:23] <zetheroo> mdadm: failed to set writable for /dev/md0: Read-only file system
[09:24] <RoyK> do you have another disk around to test if drives plugged into that thing become writable?
[09:24] <zetheroo> oh, you mean the USB dock might be the problem?
[09:25] <RoyK> I have no idea, but I've never seen disks struggle with getting read-write like this
[09:25] <zetheroo> ok, will try another disk in the same USB dock ...
[09:27] <zetheroo> plugged in another disk and it seems to also be read-only
[09:28] <zetheroo> although root can write to it ...
[09:29] <RoyK> weird
[09:30] <RoyK> what's in dmesg?
[09:31] <zetheroo> https://paste.ubuntu.com/p/bJhhhSQ9K6/
[09:31] <zetheroo> I plugged in the two md0 SSDs again
[09:31] <RoyK> [Mi Dez 23 10:30:06 2020] sd 1:0:0:1: [sdd] Write Protect is off
[09:32] <RoyK> that's good
[09:32] <RoyK> what happens if you reassemble the raid?
[09:33] <zetheroo> not sure how to do that ... looking it up ...
[09:34] <RoyK> mdadm --assemble /dev/md0
[09:34] <RoyK> or mdadm --assemble --scan
[09:35] <RoyK> if there's an md0 already, --stop it
[09:36] <zetheroo> hey
[09:36] <zetheroo> md0 : active raid1 sdd1[2] sdc1[3]
[09:36] <RoyK> :)
[09:36] <RoyK> now try to mount it, get the data off and fsck it
[09:37] <RoyK> normally just mounting it should take care of those things missing
[09:37] <RoyK> it's jouranlled after all
[09:37] <zetheroo> https://paste.ubuntu.com/p/khs8B4FQfD/
[09:37] <zetheroo> ok
[09:37] <RoyK> did it mount?
[09:38] <oerheks> clean, active ync, sounds good
[09:39] <zetheroo> yes, it mounted with 'defaults' in fstab
[09:40] <RoyK> you may want to umount it and fsck it just for kicks
[09:40] <zetheroo> but first get the data off, right? or is fsck safe ...?
[09:40] <RoyK> should be safe
[09:41] <RoyK> can you pastebin lsblk output again?
[09:43] <zetheroo> https://paste.ubuntu.com/p/5CCxVbRx89/
[09:43] <zetheroo> I'm grabbing the data just to be safe ...
[09:44] <zetheroo> and then 'fsck /dev/md0' ?
[09:44] <RoyK> looks good
[09:44] <RoyK> just umount before fsck
[09:44] <zetheroo> ok
[09:45] <RoyK> it won't let you fsck a mounted filesystem unless you --force it, which is generally an extremely bad idea ;)
[09:58] <zetheroo> ok .. unmounted and going to run fsck on it
[09:59] <zetheroo> https://paste.ubuntu.com/p/bTTwJFcCsq/
[10:19] <zetheroo> where to now? :)
[10:25] <zetheroo> RoyK: are you still there?
[11:59] <zetheroo> Btw ... I ended up trying Boot Repair on it from an Ubuntu Live Session .. and it's fixed :)
[13:37] <tieinv>  /part
[14:18] <jink> Is it safe to do-release-upgrade to 20.04.1 LTS yet? :D
[14:29] <lotuspsychje> !20.04
[14:29] <ubot3> Ubuntu 20.04 (Focal Fossa) is the 32nd release of Ubuntu and is the current !LTS release. Download at https://ubuntu.com/download - Release notes at https://wiki.ubuntu.com/FocalFossa/ReleaseNotes
[15:01] <andol> Someone should maybe update ubot3 to not refer to Ubuntu 20.04 as an !LTS release?
[15:01] <jink> !LTS
[16:49] <juliank> !20.10
[16:49] <ubot3> Ubuntu 20.10 (Groovy Gorilla) is the 33rd release of Ubuntu and is the current regular release. Download at https://ubuntu.com/download - Release notes at https://wiki.ubuntu.com/GroovyGorilla/ReleaseNotes
[16:49] <juliank> andol: But ... 20.04 is an lts release
[17:02] <mybalzitch> dang, 33 releases of ubuntu already
[17:13] <andol> juliank: Yes, and hence not a !LTS release.
[17:13] <juliank> andol: ! is like a hashtag
[17:13] <juliank> it's like saying it's the current #LTS release
[17:13] <mybalzitch> is systemd-resolve supposed to completely ignore my secondary nameserver entry?
[17:13] <juliank> !LTS
[17:14] <juliank> arguably ubot3 does not implement the LTS command right now
[17:16] <genii> !lts
[17:16] <genii> Hm
[17:17] <andol> juliank: Yeah, but given the meaning a ! prefix generally has I'd say the result just ends up being somewhat confusing.
[17:33] <mybalzitch> lol, now netplan isn't respecting the routes I've defined. so glad we left ifupdown that acutally worked for this.
[17:37] <mybalzitch> goddamn bugs in 18.04
[18:53] <teward> mybalzitch: if you don't have a question then please don't rant.
[18:53] <teward> as for netplan not respecting the routes you've defined that USUALLY means your routes are wrong
[18:54] <mybalzitch> teward: its a confirmed bug in 18.04 of netplan not applying changes properly, and I'll rant all I like thanks very much
[18:54] <teward> no, you won't, not here you won't anyways.  This is a support channel, not a rant channel.
[18:54] <teward> you can take your rants elsewhere.
[18:54] <mybalzitch> okay
[22:35] <DeeJayTwo> hi
[22:35] <DeeJayTwo> I need to restart a service each 5 minutes
[22:35] <DeeJayTwo> I made a script with service service_name restart
[22:35] <DeeJayTwo> and set to restart each 5 minutes
[22:35] <DeeJayTwo> I know it runs fine cause another line in the script is running
[22:35] <DeeJayTwo> but the service command doesn't seem to work
[22:36] <DeeJayTwo> any idea?
[22:39] <tds> DeeJayTwo: are you running this script from cron or something similar?
[22:50] <oerheks> use a systemd unit, like this? https://unix.stackexchange.com/a/265619 where RestartSec=300s ofcourse