| _ds_ | https://fosstodon.org/@gabrielesvelto/113002382002193832 is a bit interesting (re. Windows security updates & dual-boot) – from that and local checking, it appears to me that shim-signed in stable is revoked as a consequence. | 14:32 |
|---|---|---|
| _ds_ | It also appears that grub-efi-*-signed in stable aren't revoked. | 14:34 |
| _ds_ | I see a new-enough shim-signed in daedalus-proposed-updates for those who need it… | 15:08 |
| morenonatural | sup, y'all... every time I boot into desktop I have to run `sudo service unbound restart` to get working internet access (yes, I willingly installed unbound to have local resolver I can manage & maintain) ... maybe unbound isn't starting at boot? how can I make sure of that | 15:31 |
| gnarface | morenonatural: pretty sure you should see it running in the output of "ps aux --forest" | 15:32 |
| gnarface | maybe it's getting run before your networking is setup and it fails | 15:33 |
| gnarface | there should be a log somewhere you can check | 15:35 |
| nemo | /etc/network/interfaces can use UUID instead of eth* right? | 16:09 |
| fsmithred | nemo, do you mean mac address? | 16:13 |
| nemo | fsmithred: er. that. yes 😃 sorry. wrong term. | 16:38 |
| nemo | fsmithred: probably been corrupted by networkmanager | 16:39 |
| nemo | which does have UUIDs (probably derived from the MAC if I know them) | 16:39 |
| nemo | $ nmcli connection | grep loopback | 16:39 |
| nemo | lo 1c814e6f-3cff-4d02-b9ea-bb1c6a202b9c loopback lo | 16:39 |
| nemo | actually honestly no idea where it is getting that from | 16:39 |
| nemo | but could be feeding unique stuff like the mac into a UUID mapper | 16:40 |
| nemo | fsmithred: but whatever. mac, uuid, anything that is predictable-ish | 16:40 |
| nemo | fsmithred: getting some annoying behaviour with the VMs | 16:40 |
| fsmithred | sorry I wandered away | 16:45 |
| fsmithred | let me catch up | 16:45 |
| fsmithred | nemo, if you have network-manager installed it will steal your resolv.conf. If you want to edit that file manually, you can use whatever nameserver you want and then make the file immutable so that it can't be edited. | 16:47 |
| fsmithred | chatter +i /etc/resolv.conf | 16:47 |
| fsmithred | then change that to -i if you want to edit it again. | 16:47 |
| fsmithred | oh, I'm talking about the wrong file | 16:48 |
| fsmithred | if you have static ip set, you should disable network-manager | 16:49 |
| nemo | fsmithred: I don't think I'm using network-manager. that's just a habit from the past | 16:49 |
| fsmithred | oh, ok | 16:49 |
| nemo | afaik I'm doing all config in /etc/network/interfaces | 16:49 |
| nemo | the machines are a mix of dynamic and static | 16:49 |
| nemo | and I'm trying to stick to the base devuan as much as possible | 16:49 |
| fsmithred | daedalus or excalibur? | 16:50 |
| nemo | devuan is already enough of a deviation to make life difficult for other admins ☺ | 16:50 |
| nemo | fsmithred: dædalus still.. is excalibur stable? | 16:50 |
| fsmithred | no, it's not even in freeze yet | 16:50 |
| nemo | ok ☺ | 16:50 |
| fsmithred | (trixie) | 16:50 |
| nemo | yeah. that isn't going anywhere near work computers | 16:50 |
| fsmithred | lol, good plan | 16:50 |
| nemo | fsmithred: BTW. did you know that ubuntu charges for basic security patching of packages that were patched in debian stable? | 16:51 |
| fsmithred | if you want to use the new style "predictable" network names, boot with net.ifnames=1 | 16:51 |
| fsmithred | no, I did not know that and it's disgusting. | 16:51 |
| nemo | ooh. gotta look at that one | 16:51 |
| nemo | fsmithred: basically as far as I can follow, ubuntu's release schedule does not follow debian's. in fact they often roll their April stables, just a few months before debian stabilises mid-summer | 16:52 |
| fsmithred | if you're switching them over from systemd, that might be the way to go if there are custom scripts that use interface names | 16:52 |
| nemo | fsmithred: so... you end up with ubuntu stables that are often no longer on debian stable. | 16:52 |
| fsmithred | I think they fork from sid | 16:53 |
| nemo | fsmithred: no custom scripts. the hassle I had was when firing up a new VM image clone, perhaps due to interface name changes, the nic would be some new value, often old nic+1 but not always. then they'd add another nic for ipv6 and further complicate matters. so I was like. what if I just went to the console, got the mac addresses or something, or UUIDs from nmcli if those are in standard use, and used | 16:54 |
| nemo | those in the config. no more ambiguity | 16:54 |
| nemo | fsmithred: anyway. they said anything in their "core" 1000 or so packages would only be "best effort" i.e. - only if debian did all the work | 16:54 |
| nemo | they would not even pull in a debian stable patch to see if it might work on sid for same version... | 16:54 |
| nemo | fsmithred: the upshot is actual severe CVEs not getting patched unless you purchase a security plan to get access to the sekret packages | 16:55 |
| nemo | I found this interesting because folks at work are big ubuntu fans. told them to hurry up and buy the security plan then ☺ | 16:55 |
| nemo | I will say budgets are tight all around, but they can probably afford $500/y - not everywhere can though | 16:56 |
| nemo | I guess this is off-topic for the channel, sorry 😃 | 16:57 |
| nemo | but I am feeling smug about the devuan cluster | 16:57 |
| freaxeh | > #devuan-offtopic exists | 16:57 |
| nemo | s/in there/outside their/ | 16:57 |
| aswjrisp | I am back with my no bootable device issue. | 17:00 |
| aswjrisp | I spent the last two days preparing to deal with the issue. | 17:00 |
| nemo | aswjrisp: oof. I had that yesterday on my gentoo machine. | 17:00 |
| nemo | aswjrisp: because kernel update randomly decided to hop to /dev/sdb for no reason I can fathom given nothing is on /dev/sda ... maybe some UUID changed? took me a while to figure out what was going on since it was all scrolling by so fast. | 17:01 |
| nemo | fortunately was able to edit grub, boot on sdb, then fix the grub conf to use UUIDs instead | 17:01 |
| nemo | similar issue to my NIC problem I guess 😃 | 17:01 |
| aswjrisp | I did some preparation. I have set up a router with ddwrt so I can hopefully move the data off of the nvme m.2 disk with the problem. I was thinking if moving the recoverable data to another disk with scp connecting to another computer by ethernet cable and through the ddwrt router. | 17:03 |
| aswjrisp | I also had to deal with removing a bios password and turning off the secure boot setting on the bios so I could boot to a live image on a usb. | 17:04 |
| nemo | aswjrisp: I've done network copies many many times. definitely works. rsync would be faster. if you don't have a reliably readable filesystem you might want dd | nc | 17:04 |
| nemo | aswjrisp: I once suffered severe hardware damage on the devuan machine I'm on now, and was able to copy 99.99% of the disc over to a new machine by using seek/skip to jump over the handful of bad sectors by trial and error ☺ | 17:05 |
| aswjrisp | I am now able to boot the devuan live minimal amd64 on a usb. It gives me options in a grub menu which are std, access, std-toram, acces-toram, no prob and memory test. I am not familiar with the different options. Is there one that would be better for the data recovery task I am working on? | 17:08 |
| gnarface | aswjrisp: the desktop live image has a GUI. it might be easier to use if you have enough RAM | 17:10 |
| aswjrisp | nemo: In this case I am not sure if the no bootable device is related to a system update. I turned off the computer and when I turned it on again it gave me the no bootable device message. | 17:10 |
| aswjrisp | gnarface: I am fairly comfortable working on the terminal. I think I will save the ram for any programs I might need to install for the live system. | 17:12 |
| gnarface | unfortunately i don't know what all those options mean on the minimal one | 17:13 |
| gnarface | one of the "toram" ones seems a likely candidate | 17:13 |
| nemo | aswjrisp: when you boot off your USB rescue... are you able to mount and find your existing HD? I really do wonder if your issue could be similar to mine. even though devuan does use UUIDs by default when installing I think. could be an old setup? | 17:17 |
| nemo | aswjrisp: I mean. if the existing HD mounts, fscks cleanly, has no reported errors in smart and whatnot | 17:17 |
| nemo | maybe it's just config ☺ | 17:17 |
| nemo | maybe check your lsblk -f values against grub and fstab? | 17:18 |
| aswjrisp | When I set up the bios for the live usb in the bios boot order it did not show the m.2 drive. | 17:19 |
| nemo | ah :/ | 17:19 |
| nemo | that is not a good sign for sure | 17:19 |
| nemo | aswjrisp: do you see it from inside linux in the device list? | 17:19 |
| nemo | I'm guessing no 😉 | 17:19 |
| nemo | if so. good luck recovering it... | 17:20 |
| nemo | aswjrisp: is the HD an SSD? | 17:20 |
| nemo | oh. you said M.2 | 17:20 |
| nemo | so yes.. ☹ | 17:20 |
| nemo | uuugh. those are even more of black boxes recovery-wise. | 17:20 |
| fsmithred | Wasn't there a problem with some nvme disks not being bootable? I don't recall if it was a kernel issue or a motherboard issue. | 17:20 |
| aswjrisp | I found an article about the minimal live. It says the access options are for accessibility, so I will skip those. I am going to go with std-toram, which loads squashfs into ram. | 17:21 |
| fsmithred | solution was to put a /boot partition on non-nvme disk. | 17:21 |
| fsmithred | toram just makes it faster to access stuff in the iso | 17:21 |
| nemo | aswjrisp: is the disc visible to linux at all? if so fsmithred's theory could be the issue | 17:22 |
| nemo | that would explain not seeing it in the bios maybe | 17:22 |
| aswjrisp | nemo: fsmithred: I am going to try to answer your questions. I am going to try the std-toram now. | 17:22 |
| fsmithred | Is there a boot device menu available? Usually some key like F8 or F12 | 17:23 |
| nemo | last time my laptop SSD failed it was, curiously, accessible using my USB adapter for SSDs and I was able to copy data off of it that way. makes me think the failure was in the laptop itself, some marginal power thing, or just the magic of plugging it into something else reseating a marginal connection ☺ | 17:23 |
| nemo | I wonder if there are M.2 USB connectors out there.. | 17:23 |
| nemo | probably. | 17:23 |
| nemo | I mean. yes I had backups, but it's always nice when the disc works ☺ | 17:24 |
| aswjrisp | I am signed into the minimal live std-toram with user devuan. | 17:26 |
| aswjrisp | I will now start on answering the helpful questions I have already recieved. | 17:30 |
| aswjrisp | The m.2 drive in question shows up in the output of lsblk. | 17:32 |
| fsmithred | If your live system is connected to internet you could send text to termbin.com | 17:33 |
| aswjrisp | nemo: It is not the spinning disk type. I think nvme m.2 are ssd. | 17:33 |
| fsmithred | fdisk -l | nc termbin.com 9999 | 17:33 |
| fsmithred | and it'll give you a short link | 17:34 |
| aswjrisp | fsmithred: I like termbin.com. Let me see if I can connect in to the internet. | 17:34 |
| fsmithred | same for lsblk or any other commands | 17:35 |
| fsmithred | or... | 17:35 |
| fsmithred | nc termbin.com 9999 < some-file | 17:35 |
| aswjrisp | On second though I think I will try to connect the computer with the problem drive through the router by ethernet cable to the computer I am working on here. Then I can scp the output files and send them to termbin on this computer. | 17:45 |
| nemo | aswjrisp: yes. I eventually noticed you'd said M.2 :) | 17:46 |
| aswjrisp | I now have networking working. I am going to scp output files over from the computer with the problem m.2 drive to this computer I am using and then termbin them. | 18:06 |
| aswjrisp | fsmithred: thanks for explaining what std-toram is. I could have gon with std then. | 18:15 |
| aswjrisp | fsmithred: yes there was a boot device menu for the bios. It did not show the m.2 drive in question. | 18:16 |
| aswjrisp | nemo: yes there are m.2 use adapters that was what I was initially planning before it was suggested I use a live image. | 18:18 |
| aswjrisp | s/use adapters/usb adapters/ | 18:22 |
| aswjrisp | fsmithred: fdisk --list output https://termbin.com/lj8c | 18:40 |
| aswjrisp | fsmithred: lsblk output https://termbin.com/9okk | 18:42 |
| aswjrisp | I installed smartmontools. | 18:44 |
| aswjrisp | smartclt --all /dev/nvme0n1 output https://termbin.com/az1z | 18:55 |
| nemo | aswjrisp: ah. so you could unplug the M.2 drive from the failing machine, and plug it in by USB to another machine and verify that it reads correctly. If it works there that would help eliminate the drive as source of problem and point more towards the machine you are installing it in | 18:56 |
| nemo | debugging is all about splitting out variables if you can ☺ | 18:56 |
| nemo | might also make doing drive testing easier too, less annoying than in a live image | 18:57 |
| aswjrisp | fsmithred: blkid output https://termbin.com/pghe | 19:19 |
| aswjrisp | fsmithred: I do not have the live iso I used when I installed the system. I downloaded a new iso which is probably different that what I used when I installed devuan on this m.2 drive, as that was a while ago. | 19:23 |
| ErRandir | aswjrisp: Looks like your system would normally boot off /dev/nvme0n1p1. That should have grub on it. With parted check that has the boot flag set. | 19:42 |
| aswjrisp | ErRandir: /dev/nvme0n1p1 does have the boot flag, as shown by parted --list output https://termbin.com/4xiq | 19:50 |
| aswjrisp | nemo: I might try getting an nvme m.2 adapter. But I would like to see how far I can get without having to buy the adapter. | 19:53 |
| aswjrisp | What I would like to accomplish is to get a copy of the data that is in the /home/user/ home directory. | 19:55 |
| ErRandir | good. Can you add it back as a boot device in the bios? | 19:55 |
| ErRandir | The directory you want is probably on /dev/nvme0n1p3. You can probably just mount that. | 19:56 |
| aswjrisp | ErRandir: The bios does not detect this nvme m.2 drive at all. If I go to the boot section and it's subsection for boot order this m.2 drive does not show up at all. | 19:57 |
| ErRandir | There is probably a different submenu where you have to add it. Depends on the bios vendor. I'm not familiar with all of them! :) | 19:59 |
| aswjrisp | Given that the drive shows up in the output of fdisk, lsblk, bldik and parted. Also that smartctl says that the overall-health self-assement test passed. I think I might try decrypting the partition and copying the user's home directory. | 20:02 |
| aswjrisp | Is there a reason why decrypting and copy data would be a problem? | 20:02 |
| aswjrisp | ErRandir: I will try adding the boot partion back as a boot device in the bios after I try to copy the user's home directory. | 20:03 |
| aswjrisp | I was not playing in the bios settings before running into this no bootable device message from the bios. | 20:05 |
| ErRandir | sure try decrypting it first. It's good to have multiple options forward. | 20:15 |
| aswjrisp | There is a Luks header. `cryptsetup luksDump /dev/nvme0n1p3` outputs https://termbin.com/z6em | 20:34 |
| aswjrisp | cryptsetup open let me enter a password exited without errors. lsblk shows the partition is decrypted. | 20:43 |
| fsmithred | mount /dev/mapper/whatever-it's-called some-mountpoint | 20:49 |
| fsmithred | probably /mnt is a good place | 20:50 |
| aswjrisp | I ran `cryptsetup open /dev/nvme0n1p3 root`. | 21:46 |
| aswjrisp | When I try to mount with `mount /dev/mapper/root /mnt/` I get "mount: /mnt: unknow filesystem type 'LVM2_member'". | 21:50 |
| aswjrisp | lsblk now outputs https://termbin.com/1n81 | 21:53 |
| aswjrisp | `mount /dev/alaptop-vp/root /mnt` outputs "18 orphaned inodes deleted", "recovery complete" and "mounted filesystem with ordered data mode. Quota mode: none." | 22:05 |
| aswjrisp | The message "recovery complete" sounds good. | 22:05 |
| aswjrisp | Those messages were from the kernel they were not from the mount command. | 22:09 |
| * aswjrisp using scp to copy home directory. | 22:17 | |
| fsmithred | oh you have to activate the lvm first | 22:19 |
| fsmithred | vgchange -ay something | 22:20 |
| fsmithred | alaptop--vg-root I think | 22:21 |
| fsmithred | then you can mount /dev/mapper/alaptop--vg-root (I think) | 22:23 |
| fsmithred | gotta go, bbl | 22:23 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!