| al1r4d | hi rrq | 03:57 |
|---|---|---|
| al1r4d | thank you for your attention | 03:57 |
| rrq | hi | 03:58 |
| al1r4d | but now i'm able to migrate from debian to devuan | 03:58 |
| al1r4d | :) | 03:58 |
| rrq | cheers | 03:58 |
| al1r4d | yeah | 03:58 |
| systemdlete | Getting error on apt update on one (daedalus) client of my apt-cacher. But other (daedalus and chimaera) clients not having a problem atm. I've tried several times, ran apt-cacher-ng maintenance, restarted cacher. | 04:08 |
| systemdlete | error is about an outdated key | 04:08 |
| systemdlete | (I can fetch the exact message if you want) | 04:08 |
| al1r4d | maybe you can post it via pastebin | 04:08 |
| systemdlete | it happens to be backports | 04:08 |
| systemdlete | allr4d, sure. give me a moment | 04:09 |
| gnarface | systemdlete: key might actually be outdated, but are you sure the clock is right on that one? | 04:09 |
| systemdlete | https://dpaste.com/D2N8QJSUS | 04:10 |
| systemdlete | gnarface, hmmm. let me check (thanks!) | 04:10 |
| darwin | are there any OpenVZ virtual private server (VPS) companies that run it on SysVInit so you can run Devuan, Slackware, Gentoo, and so on, on their systems, or if they use SysVInit they may as well be running Xen or KVM for VPS? | 04:11 |
| darwin | in converted four of my VPS from Ubuntu to Debian to Devuan, but since they were running OpenVZ in systemd, I couldn't even reboot or poweroff except hard power off | 04:12 |
| gnarface | darwin: i found one a few years back that was using Xen and i was able to upgrade to Devuan and maintain full functionality, but it was in Russia... | 04:13 |
| darwin | i don't really care what country it's in, but with the way things are going between the West (where I live) and Russia I don't know if I'd be breaking some law signing up for that or who I'd be supporting (ethically) | 04:13 |
| gnarface | darwin: i don't think it's illegal for private citizens to rent VPSes in Russia yet, but the legality wouldn't be my primary concern | 04:14 |
| al1r4d | let me try to add backports and see what will happened, systemdlete | 04:15 |
| systemdlete | good one, gnarface. It is off 37s | 04:15 |
| gnarface | systemdlete: is that enough to matter? i suppose if it's 37s past the 5 minute window maybe? | 04:15 |
| systemdlete | I don't know, but I see another problem: after shutting down ntpsec and running ntpd -gq, it says no server found. | 04:16 |
| rrq | darwin: have you tried VPS on vultr.com? seems to be on kvm (maybe?) | 04:16 |
| systemdlete | pretty sure the others are finding my server | 04:16 |
| gnarface | systemdlete: hmm, disturbing... DNS outage somehow on just that one? | 04:16 |
| darwin | i did in past but disliked some things about them | 04:17 |
| systemdlete | the ntp, dns, and other network servcies all served off same VM | 04:17 |
| systemdlete | let me try setting time on this VM... | 04:17 |
| darwin | out of the three medium-sized main ones I read about in past, DigitalOcean, Linode, Vultr, only Linode allowed the most freedom of speech | 04:17 |
| al1r4d | systemdlete, oh yeah, i have same problem too | 04:17 |
| darwin | smaller ones like RamNode/TinyKVM allowed more but don't have many data centres | 04:17 |
| darwin | oops, RamHost | 04:18 |
| gnarface | darwin: i don't remember the name of the one i was using exactly, so i am having trouble figuring out if it's even still in business, but i think i mentioned it here, so it might be in the channel logs if they go back that far | 04:18 |
| systemdlete | hmmm. time reset (ntpd -gq) works here. So it has to be something else. | 04:18 |
| systemdlete | looks like that VM is all knoodled. I'm restarting it. | 04:20 |
| al1r4d | lel | 04:22 |
| al1r4d | ok | 04:22 |
| al1r4d | so solved? | 04:22 |
| darwin | can I install kompozer? | 04:28 |
| darwin | it was a 32-bit KDE HTML/CSS editor | 04:28 |
| darwin | but it's still the best one | 04:28 |
| gnarface | darwin: is it possible the name has changed? | 04:29 |
| darwin | i doubt it | 04:29 |
| systemdlete | al1r4d, gnarface: turns out it was my ntpsec config--somehow on that system, I had forgotten to fix the minsane parameter. Now it works correctly. Luckily I remembered while looking at the config file just now. | 04:29 |
| systemdlete | all is good again | 04:29 |
| systemdlete | (well, not "all" but you know what I mean) | 04:29 |
| systemdlete | but checking the time was a good reminder, thank you gnarface. I'll tell rubber ducky here to suggest that next time. | 04:30 |
| systemdlete | well, the time is set correctly now. But I'm still getting hell from apt update. | 04:32 |
| systemdlete | same as before | 04:32 |
| systemdlete | is there a clean way to tell apt to clear its cache (I've already done "apt clean") | 04:32 |
| systemdlete | I'm thinking maybe it has cobwebs | 04:33 |
| systemdlete | al1r4d, are you still having the problem also? | 04:33 |
| gnarface | darwin: near as i can tell, it was pulled before Jessie, but debian's package search seems down so i can't verify when. the name suggests it's part of the KDE suite though, maybe they've simply replaced it with something different? | 04:33 |
| darwin | it never was replaced | 04:34 |
| gnarface | darwin: bummer. seems like your best option is to try to find a copy of the source and build it. | 04:34 |
| gnarface | might still be in the debian archives somewhere... | 04:35 |
| systemdlete | gnarface, apt update is still being mean to me... | 04:35 |
| darwin | apt always says: Error: GDBus.Error:org.freedesktop.DBus.Error.Spawn.ExecFailed: Failed to execute program org.freedesktop.PackageKit: Permission denied | 04:35 |
| darwin | of course, I deleted /root/.dbus because am not using X as root | 04:35 |
| gnarface | systemdlete: i can only recall once having to delete files manually that "apt clean" wouldn't clear, but i forget exactly where they were | 04:36 |
| al1r4d | systemdlete: Failed to fetch https://deb.devuan.org/merged/dists/daedalus-backports/InRelease Certificate verification failed: The certificate is NOT trusted. The name in the certificate does not match the expected. Could not handshake: Error in the certificate verification. | 04:36 |
| al1r4d | hmm | 04:36 |
| gnarface | al1r4d: that might just be because you're using https in your sources.list | 04:37 |
| systemdlete | al1r4d, I am using apt-cacher-ng here. So the parameters will be different | 04:38 |
| gnarface | al1r4d: some of the mirrors in the round-robin aren't properly configured to host https for devuan domains because they're already hosting https for debian domains. if you want to use https in your sources.list you have to manually pick one of the mirrors that's ok with it. | 04:38 |
| gnarface | (this is a limitation of https on shared IPs) | 04:39 |
| * systemdlete longs for the old days before... bleh | 04:39 | |
| rrq | well it's also a quesiton of "sharing the certificate | 04:39 |
| gnarface | well, right, but they could set it up with multiple ips but that costs more money | 04:40 |
| rrq | virtual servers share IP; the cert is tied to the domain only | 04:41 |
| systemdlete | gnarface, in my case, all of the clients are pulling files from the cacher, and most are doing so without incident. Just one client. | 04:42 |
| darwin | after KDE discontinued Kompozer I think they made Quanta which claims to be a web development IDE or somethng but can't even make a simple HTML nor CSS file. It's some sort of thing for making fancy/garbage that load other types of pages and run code on your PC | 04:43 |
| darwin | it's like 'IDEs' that don't like you make a plain .C or Makefile, just make everything into a 'project' | 04:45 |
| darwin | "don't 'let' you" | 04:45 |
| darwin | the good thing about Kompozer is you could see the webpage you're editing as you edit it, and also do it visually, also while doing the code in a separate pane. I don't think Quanta does that. Mozilla SeaMonkey might have something that does it | 04:46 |
| systemdlete | I tried to reset the key using the gpg --recv-key bit, but it fails to update gpg | 04:54 |
| systemdlete | (I'm nearly clueless with apt) | 04:54 |
| systemdlete | I just cold-booted the system and time.gov says my clock is only off by .020 seconds, which is about the same as my other systems | 04:55 |
| gnarface | systemdlete: i only remember some sort of rare race condition and a invalid https url were associated, and i had to manually delete the file because there was no contingency in the code for that happening. maybe a bunk mirror was involved, not sure. this was before devuan. | 04:55 |
| systemdlete | okaaaaay | 04:56 |
| systemdlete | a race condition | 04:56 |
| systemdlete | very nice | 04:56 |
| gnarface | i don't think it was just something in /var/cache/apt, but i'm not sure | 04:56 |
| gnarface | yea it was like "if you interrupt the query in this state halfway through, it leaves some internal apt configuration in an recoverable state" | 04:56 |
| rwp | Due to repeating problems with the mirrors I had to stop using apt-cacher-ng for Devuan. Too many times where it would have problems and everything would work okay if I bypassed it. | 04:57 |
| gnarface | systemdlete: *unrecoverable state | 04:57 |
| * systemdlete is just about at the point rwp is with apt-cacher-ng, believe me | 04:57 | |
| rwp | I have this belief that the reason is inconsistent mirror pulling from the master which is done with rsync in Devuan while in Debian they insist upon using a dedicated sync script to do things in a particular set of phased steps to ensure correctness. | 04:58 |
| rwp | And then additionally the Devuan mirrors are all independent operators with different TTL time-to-live cache times. apt-cacher-ng is not at fault there, all of the problems I have chased down seemed like apt-cacher-ng was operating correctly but the file had a live TTL but the mirror had updated under it with rsync already. | 04:59 |
| rwp | I still use apt-cacher-ng for my Debian systems that are still Debian. apt-cacher-ng continues to work great there. | 05:00 |
| systemdlete | I just tried undoing apt-cacher on that client that has trouble. I reset the http to https in sources.list and disabled the apt-cacher call in the config | 05:01 |
| rwp | For Devuan I would apt-get purge apt-cacher-ng, purge the cached files, then install it again and everything would work for that event again. | 05:01 |
| systemdlete | re-ran apt update, this time totally avoiding the cacher. Even worse errros | 05:01 |
| systemdlete | errors | 05:01 |
| rwp | I suggest using http rather than https. There is no need for https for the sources.list entries. | 05:02 |
| systemdlete | ok, I'll try that too | 05:02 |
| systemdlete | (I already did that, but I will try again) | 05:02 |
| rwp | The data is checked with GPG signatures which are the gold standard. | 05:02 |
| onefang | rwp: I have a TODO item to look into the Debian ways of syncing mirrors. | 05:02 |
| gnarface | systemdlete: this time the issue really may be upstream... i can't get packages.debian.org to return a search result either right now | 05:02 |
| rwp | onefang, Life and time is what keeps everything from happening all at once. But when you get to that point please ping me and I will try to help out with it. :-) | 05:03 |
| systemdlete | finally, finally! came back clean | 05:03 |
| systemdlete | (with cacher client disabled) | 05:03 |
| rwp | https depends upon SNI set up, depends upon clocks being correct, and so forth. And it is completely unnecessary with the standard GPG signature verification. | 05:04 |
| onefang | So all good now, and I can go back to my game? | 05:04 |
| * rwp is going to get up from the keyboard and go watch a movie :-) | 05:04 | |
| systemdlete | all good... well, kind of... not really | 05:04 |
| onefang | Have fun. | 05:04 |
| systemdlete | cacher is not working. On one GD system! | 05:05 |
| systemdlete | gnarface, just saw your last post. Whoa! really? | 05:06 |
| systemdlete | So this is going around now? | 05:06 |
| systemdlete | now to try upgrade... | 05:06 |
| systemdlete | works. No packages to upgrade acctually, but no errors. | 05:08 |
| systemdlete | EVEN WITH THE apt-cacher TURNED BACK ON! | 05:08 |
| systemdlete | gnarface, is it possible that the cacher is not handing back the keys (or is that done independently of the cacher)? | 05:09 |
| gnarface | systemdlete: not sure... i think what's likely is if they update the keys the cacher will still be holding the old keys until a cache refresh, but that doesn't explain why you're seeing the problem in only one VM... maybe disk-to-ram sync issue? | 05:10 |
| gnarface | er, ram-to-disk, rather | 05:11 |
| systemdlete | hmmm. on the client VM, the cacher server VM, or the host machine? | 05:11 |
| systemdlete | client? | 05:12 |
| systemdlete | it's a virtual disk... | 05:12 |
| systemdlete | there are several VMs running on the host, and none had this except the one client | 05:13 |
| systemdlete | so if there is a ram-to-disk problem, I'd expect it to be more widespread. Esp since I had rebooted the client VM before this (somehow) resolved itself. | 05:13 |
| systemdlete | a rboot is not a guarantee that such a problem would be corrected, but it seems highly likely that it would. Esp. with a cold reboot | 05:14 |
| systemdlete | not saying you are wrong of course | 05:14 |
| systemdlete | Just seems like a long-shot | 05:14 |
| darwin | how can I prevent deletion of /tmp/.directory ? | 05:17 |
| darwin | someone said don't use a directory that is temporary, so I guess that's someone who never needs a temporary place to put things, doesn't use scratch paper, never makes a mistake or so on. They shouldn't have told me what to do | 05:18 |
| systemdlete | not only do I make mistakes, I have a whole machine dedicated to doing just that! | 05:20 |
| gnarface | systemdlete: i meant on the VM... like maybe the sync interval is way low and the software caches the new key but the next thing that reads it is still reading the old key off disk? i dunno, grasping at straws here... | 05:20 |
| gnarface | systemdlete: anything different about what's running in that VM that could cause ... I/O delay? i dunno | 05:21 |
| systemdlete | hmmm. Well, the server VM is definitely busy enough for something like that to happen. But it wasn't the victim. | 05:21 |
| systemdlete | In fact, this client happens to be (probably) the least loaded VM in my network. | 05:21 |
| gnarface | darwin: you're supposed to use /var/tmp for that historically, though i remember recent conversations about Debian wanting to delete that one on reboot too now... | 05:22 |
| systemdlete | Well, before the reboot, there were some time delay messages in the VM's logs, but they were not persistent or many at any one time. Just occassional. But even that being the case, I haven't seen these errors since the reboot, where it was still not working, and then--magically--working again. | 05:23 |
| systemdlete | OK, to be fair: I did have a firefox running, with several tabs open. | 05:25 |
| systemdlete | so that could have been the smoking gun I suppose. | 05:25 |
| systemdlete | I've had several instances where open firefoxes and/or thunderbirds tied up a machine pretty badly. | 05:26 |
| systemdlete | These days, I shut them down if I am not going to work on them for a while. | 05:26 |
| darwin | that's why I want 32x8GHz | 05:27 |
| darwin | soon that will be slow for those even with 256GB RAM | 05:27 |
| systemdlete | I wouldn't advertise you are looking for that. There are firefox devs who are hungry for more CPU to waste on YOUR systems. | 05:28 |
| darwin | i see | 05:28 |
| systemdlete | You know all those nice folks who bring you popups and other crap you don't want to see all over the place. | 05:29 |
| systemdlete | That's how we help pay their salaries I suppose. | 05:29 |
| onefang | I have 256 GB of RAM, and 64 cores. Sometimes it's slow. lol | 05:29 |
| systemdlete | So the more ad space they can sell, the richer they get, while bringing you the convenience and security of trackers and other miscellany. | 05:30 |
| * systemdlete thinks onefang is a brat | 05:30 | |
| systemdlete | but, onefang, how many of them are secretly doing cryptomining and playing ads you are not watching? | 05:31 |
| onefang | With all the protections I have in place, very few if any. Though that's why it's slow sometimes. lol | 05:33 |
| systemdlete | see, you can't win, even when you ARE winning | 05:33 |
| systemdlete | It's like craps. The more money you wager on pass line odds bets, the less money you lose when your winning. | 05:34 |
| systemdlete | there is always a tax. A gotcha. A gimmick. | 05:34 |
| systemdlete | Something you didn't realize you signed up for. | 05:34 |
| systemdlete | but this is sooooo OT | 05:34 |
| * systemdlete finally shuts his clam | 05:34 | |
| mathew | Possibly an easy question, but i cannot find a good answer: How long is beowulf supported? I m thinking about migration of my old buster server to beowulf. because buster goes lts now. | 09:48 |
| gnarface | mathew: slow channel, stick around though and someone will answer | 09:57 |
| gnarface | i don't actually know the answer myself | 09:57 |
| al1r4d | mathew, idk, try it | 10:12 |
| fsmithred | mathew, beowulf is buster, so migrating won't help. We use debian kernels. | 10:44 |
| mathew | ok, i did a migration ones from bullseye to daedalus on my main machine, it worked fine. what i dont know is: is beowulf also long-time supported? so do i have a advantage to migrate now or should i just upgrade the machine to bullseye and maybe migrate later? | 11:10 |
| gnarface | some clarifications: | 11:12 |
| gnarface | - you're not supposed to skip releases during upgrades, even with devuan | 11:12 |
| gnarface | - nothing will stop you from continuing to use beowulf, i still do on some things... | 11:12 |
| gnarface | - if you want timely security updates you should stick with current stable | 11:13 |
| gnarface | - always keep a backup of the old system, this obviates a hell of a lot of the other problems | 11:13 |
| systemdlete | mathew, not too long ago, I upgraded a beowulf system to chimaera. never got to upgrade to daedalus though. | 11:13 |
| systemdlete | and take gnarface's advice: don't skip releases on upgrades. Results will be undefined. | 11:13 |
| mathew | so in devuan everything under stable should get securty update? | 11:13 |
| gnarface | yes, it's basically 99% debian packages, served by http redirect | 11:14 |
| gnarface | it'll get all the same security updates as debian | 11:14 |
| gnarface | likewise, with beowulf | 11:14 |
| mathew | ok this is good to know. of course i have a backup of the server. the cpu is old so i dont know if the knew kernel is good, but i will see | 11:15 |
| gnarface | for the x86 stuff, that includes the kernels, for the arm stuff, it does not | 11:15 |
| gnarface | ah, also worth noting that in most cases you can get away with running the previous release's kernel if you have to | 11:17 |
| gnarface | most software isn't that picky | 11:18 |
| gnarface | there's always exceptions and sometimes they're dramatic, so be careful | 11:18 |
| freepick | Hello how are you? I just installed devuan and it was successful, but when modifying the partitions an error appeared that the installation pendrive had been written, I did an encrypted installation. I just ignored it, but I wanted to let you know and know if it's normal or something has gone wrong. | 13:06 |
| gnarface | too bad freepick didn't stick around... | 14:36 |
| free | i am | 14:36 |
| gnarface | free: oh, you're freepick, aren't you? | 14:36 |
| gnarface | it's normal if you told it to do that | 14:36 |
| freepick | V | 14:36 |
| freepick | yes, my nickname changed | 14:37 |
| freepick | Ok, could it be that it is done on purpose so that the pendraive is deactivated and not booted from it? | 14:39 |
| gnarface | i don't know for sure but i doubt that. if you picked encrypted install and auto-partitioning it might have simply encrypted and partitioned every attached disk, don't really use that feature myself | 14:44 |
| gnarface | stick around and maybe you can get a second opinion | 14:44 |
| fsmithred | what's in the installer log? | 14:44 |
| fsmithred | /var/log/installer/syslog | 14:44 |
| fsmithred | maybe the actual error message. I've never seen that one. | 14:45 |
| Xenguy | mathew, I recently upgraded from Beowulf to Chimaera because the former is scheduled to reach end of LTS as of the end of this month... | 15:03 |
| Xenguy | So you can migrate to Beowulf, but I believe you would want to quickly upgrade from Beowulf to at least Chimaera before the end of the month... | 15:04 |
| Xenguy | Be aware that wicd is no longer supported in Chimaera, if you were using that software | 15:05 |
| Xenguy | Chimaera will continue to receive security updates for another couple of years or so. | 15:06 |
| freepick | I have gone to carry out the installation process again in expert mode. After choosing encrypted lvm and selecting the disk, a banner appears informing you that the partition table of the selected disk and the partition table of the installation media have been modified. But at no time did I choose to modify the installation media. And after that | 15:07 |
| freepick | the error appears that the installation media partition was written but they could not notify the kernel, at that moment I chose to ignore it and it installed without problems. | 15:07 |
| gnarface | freepick: seems weird, that's the daedalus installer? which iso exactly? (paste exact url you used to download it) | 15:08 |
| freepick | i use torrent webtorrent:https://files.devuan.org/devuan_daedalus.torrent devuan_daedalus_5.0.1_amd64_netinstall.iso | 15:10 |
| gnarface | hmm, did the checksum match? | 15:10 |
| freepick | yes | 15:10 |
| freepick | dont check gpg, but checksum yes | 15:11 |
| freepick | sorry about the gpg I also checked it | 15:11 |
| gnarface | freepick: upload /var/log/installer/syslog to paste.debian.net so fsmithred can see it | 15:11 |
| freepick | How do I return the syslog output to text? | 15:12 |
| gnarface | it should already be text | 15:13 |
| gnarface | and paste.debian.net will let you upload it | 15:13 |
| gnarface | you don't have to copy&paste or anything | 15:13 |
| gnarface | is that file not a text file for you??? | 15:14 |
| freepick | yes, i have the file, now i upload. I was looking for the error but I didn't find it, maybe I don't know what I should look at haha | 15:15 |
| gnarface | well i don't know for sure either, let's see what fsmithred says. a screenshot of what you see on the screen during the installer phase wouldn't hurt either | 15:16 |
| gnarface | nobody has ever complained about this before, so it's either a very new bug or you've done something weird | 15:17 |
| freepick | Sorry for being a newbie but I don't know how to upload the syslog :v but I have a screenshot haha | 15:17 |
| gnarface | no worries, everyone starts somewhere. statistically speaking, it's almost certain to be a case of user error, but it's important to check things | 15:18 |
| gnarface | first of all, type this, make sure you can see it: ls -l /var/log/installer/syslog | 15:18 |
| gnarface | as your own user, the one running the web browser | 15:18 |
| gnarface | does that show you a file, or give you an error? | 15:19 |
| freepick | i can see ls -l /var/log/installer/syslog | 15:20 |
| gnarface | ok, on paste.debian.net, do you see the button at the bottom of the right hand box that says "Browse..." ? | 15:21 |
| freepick | choose file see | 15:22 |
| gnarface | yea, so, you'll start out in your home directory | 15:23 |
| gnarface | you have to go UP two directories to see /var | 15:23 |
| gnarface | how this panel looks varies depending on the desktop you chose | 15:24 |
| gnarface | if you don't see an up arrow that's not the cdrom eject button, look for a left arrow instead | 15:24 |
| gnarface | if you don't see either, try just typing into the space over the middle of the file picker and a text prompt i shit you not may magically appear (this is the gtk one) | 15:25 |
| freepick | I already found it - can I also upload screenshots in that same place? | 15:25 |
| gnarface | sorry, no, not as far as i know. i don't have a good location for screenshots off the top of my head. | 15:25 |
| gnarface | someone else around here probably has an idea though | 15:26 |
| Xenguy | From the command-line: curl -F'file=@yourfile.png' https://0x0.st | 15:26 |
| gnarface | there we go | 15:26 |
| gnarface | (you may have to install curl first) | 15:28 |
| freepick | ok, i'm on it | 15:30 |
| freepick | send it 2 pictures | 15:33 |
| gnarface | did it give you links? | 15:33 |
| freepick | "I sent two photos | 15:33 |
| freepick | https://0x0.st/XAMQ.jpg | 15:34 |
| freepick | https://0x0.st/XAM1.jpg | 15:34 |
| gnarface | fsmithred: ^ | 15:35 |
| fsmithred | checking... | 15:35 |
| gnarface | did you get a paste.debian.net link too? | 15:35 |
| gnarface | freepick: ^ ? | 15:35 |
| freepick | The page does not work, now I try again | 15:36 |
| fsmithred | I can't read it (Spanish?) but it looks like the verification screen in the partitioner. | 15:36 |
| fsmithred | and it looks like partitions on sdd and sdc were chosen | 15:37 |
| gnarface | freepick: think maybe you might have selected them accidentally? | 15:38 |
| fsmithred | and the next screenshot is the error message | 15:39 |
| freepick | I have only chosen a single disk, if another partition was chosen it was not me. I have several disks on the PC, I can remove all of them and only leave the installation media and a single disk to see what happens | 15:39 |
| rrq | freepick: how large is the pendrive? | 15:42 |
| rrq | with the installer | 15:42 |
| freepick | pendrive 64gb | 15:42 |
| freepick | in paste.debian I cannot send, or maybe it has already been sent. When I hit send it says that the site cannot be accessed and it exits | 15:43 |
| rrq | and you selected encrypted lvm option? | 15:43 |
| freepick | yes | 15:43 |
| rrq | the partitioner is supposed to avoid the installer disk, but maybe it gets confused seeing so much free space | 15:44 |
| fsmithred | I don't suppose you saved a shot of the screen before these two. | 15:45 |
| joerg | https://0x0.st/XAM1.jpg looks pretty standard for repartitioning [english too] | 15:46 |
| fsmithred | except that /dev/sdd is the installation media | 15:47 |
| joerg | :nod: | 15:47 |
| joerg | though, no idea, is it? | 15:47 |
| fsmithred | based on what OP said, yeah. | 15:48 |
| fsmithred | but still an assumption | 15:48 |
| fsmithred | I'd like to know what's on sdd1. Is that a /boot partition? | 15:48 |
| joerg | wouldn't the installer mount own volume as /dev/sda and partition any *additional* volumes? | 15:49 |
| joerg | no idea what that particular installer does | 15:49 |
| rrq | sdd1 is the efi partition | 15:49 |
| joerg | OUCH! | 15:49 |
| freepick | If you want I can take screenshots of the entire process and send them | 15:49 |
| fsmithred | No, I think it uses the motherboard's idea of which disk is first | 15:49 |
| rrq | no wait... sdd1 would be the cdrom as a whole and sdd2 would be the efi partition | 15:50 |
| joerg | findmnt might help | 15:50 |
| fsmithred | how did you prepare the usb pendrive? | 15:50 |
| joerg | needs a shell though | 15:51 |
| rrq | but the disk sdd should not be included | 15:51 |
| freepick | pendrive with dd if=devuan.iso of=/dev/sdX | 15:51 |
| rrq | freepick: after selecting "do encrypted lvm" I get a dialog for choosing disk | 15:52 |
| freepick | I choose full disk and configure lvm encryption | 15:53 |
| rrq | and the installer is one of the options | 15:53 |
| rrq | ok .. | 15:54 |
| rrq | is sdc a pendrive too? | 15:54 |
| freepick | sdc should be the disk where I want to install | 15:55 |
| rrq | I'm setting up the test case... are there sda and sdb disks too? | 15:58 |
| freepick | yes | 16:01 |
| freepick1 | 3 disk and instalation medio | 16:01 |
| freepick1 | install pidgin and start the pc because the session is closed from the browser and messages are deleted ha | 16:02 |
| rrq | ok.. and flow is: 1) pick "whole disk with encrypted lvm" 2) select target disk ... then encrypting setup? | 16:02 |
| joerg | http://reisenweber.net/irclogs/libera/_devuan/ | 16:03 |
| freepick1 | yes, all files in a single partition, recommended for newbies | 16:03 |
| joerg | http://reisenweber.net/irclogs/libera/_devuan/latest.log.html | 16:04 |
| rrq | how large is sdc? | 16:04 |
| freepick1 | 240gb ssd sata | 16:04 |
| joerg | can you open up a (remote) shell? | 16:05 |
| joerg | findmnt is very informative | 16:06 |
| joerg | df -h as well | 16:07 |
| rrq | ok my guess is htat the partitioner discover that sdd is 64g but its partitioning says like 600M (or something) so it decides to "correct that partition table" just by-the-by ... | 16:07 |
| rrq | no particular harm in that apart from it destroying the sha256sum for the image | 16:09 |
| joerg | and throwing a warning (error?) | 16:10 |
| freepick1 | joerg This is my setup right now, even the installation media is still connected haha | 16:10 |
| freepick1 | NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS | 16:10 |
| freepick1 | sda 8:0 0 223,6G 0 disk | 16:10 |
| freepick1 | ├─sda1 8:1 0 512M 0 part /boot/efi | 16:10 |
| freepick1 | ├─sda2 8:2 0 488M 0 part /boot | 16:10 |
| freepick1 | └─sda3 8:3 0 222,6G 0 part | 16:10 |
| freepick1 | └─sda3_crypt 254:0 0 222,6G 0 crypt | 16:10 |
| freepick1 | ├─devuan--vg-root 254:1 0 221,6G 0 lvm / | 16:10 |
| freepick1 | └─devuan--vg-swap_1 254:2 0 980M 0 lvm [SWAP] | 16:10 |
| freepick1 | sdb 8:16 0 223,6G 0 disk | 16:10 |
| freepick1 | ├─sdb1 8:17 0 512M 0 part | 16:10 |
| freepick1 | ├─sdb2 8:18 0 488M 0 part | 16:11 |
| freepick1 | └─sdb3 8:19 0 222,6G 0 part | 16:11 |
| freepick1 | sdc 8:32 0 953,9G 0 disk | 16:11 |
| freepick1 | └─sdc1 8:33 0 953,9G 0 part | 16:11 |
| freepick1 | sdd 8:48 1 57,6G 0 disk | 16:11 |
| freepick1 | ├─sdd1 8:49 1 456M 0 part | 16:11 |
| freepick1 | └─sdd2 8:50 1 21,5M 0 part | 16:11 |
| joerg | done? | 16:11 |
| joerg | please use pastebin for >3 lines | 16:11 |
| joerg | thanks | 16:11 |
| freepick1 | sorry, I'm a newbie | 16:12 |
| freepick1 | in irc | 16:12 |
| joerg | np :-) | 16:12 |
| joerg | lickily no network bot got angry and kicked you ;-) | 16:13 |
| rrq | freepick1: can you run fdisk -l /dev/sdd ? .. does that complain? | 16:14 |
| joerg | cat foo | nc termbin.com 9999' | 16:14 |
| freepick1 | how to put this in pastebin? | 16:15 |
| joerg | ^^^ | 16:15 |
| joerg | here fdisk -l /dev/sdd | nc termbin.com 9999 | 16:15 |
| joerg | or | 16:16 |
| joerg | here fdisk -l /dev/sdd 2>&1| nc termbin.com 9999 | 16:16 |
| freepick1 | https://termbin.com/lad1 | 16:16 |
| rrq | ok. and an installation attempt now, with that pendrive; does that still raise the error? | 16:20 |
| joerg | sdd2 is pretty "funny" | 16:21 |
| freepick1 | I just tried the installation to take the screenshots that I shared above | 16:21 |
| freepick1 | It is the same installation media that I used to install devuan on the PC I am writing from. | 16:23 |
| rrq | ok. (I just wondered if it was that previously sdd pretended to be 456M and it now has got "corrected" ... but that would have happened already before that screenshot run) | 16:25 |
| freepick1 | no idea | 16:26 |
| freepick1 | I have the syslog/installer that they asked me for today, I don't know if they want it | 16:26 |
| rrq | you got the link for that? | 16:27 |
| freepick1 | https://termbin.com/0fyk | 16:27 |
| rrq | thanks | 16:27 |
| freepick1 | (: | 16:32 |
| rrq | that log is from the installation, and not the one for the subsequent sdc + sdd dialog | 16:33 |
| freepick1 | The sdc + sdd dialog box above is a screenshot of a test I did to take the screenshot, but it is the same scenario that I had when installing the log installer that I have provided | 16:36 |
| rrq | the latter one would be available ate the time when the error dialog is show; you shift to a shell with Alt-F2 and see the file /var/log/syslog | 16:36 |
| freepick1 | Or perhaps, would I have to get the installer log before exiting the installation process? | 16:36 |
| freepick1 | aaaaaaaaaaaa | 16:37 |
| rrq | yes, use Alt-F2 to get a shell | 16:37 |
| rrq | then the log is /var/log/syslog | 16:37 |
| rrq | the shell has a fair few (busybox) commands | 16:37 |
| freepick1 | ok, i can use pastebin in this momento? | 16:38 |
| rrq | like "mount" "cp" and "nc" | 16:38 |
| rrq | nc termbin.com 9999 < /var/log/syslog | 16:38 |
| freepick1 | ok, i'll try | 16:40 |
| rrq | maybe even: nc -w1 termbin.com 9999 < /var/log/syslog | 16:42 |
| rrq | to ensure that nc doesn't terminate too quckly | 16:42 |
| * joerg gave up on redirections etc with nc. alias paste='nc termbin.com 9999' and usual commandline actually is literally `cat|paste` then paste a text from copybuffer into terminal (shift+ctrl+V, or middle-mous-button) then press enter ctrl+d to signal EndOfText and get the link reply from nc/termin | 16:48 | |
| freepick | hi https://termbin.com/9w3s | 16:50 |
| joerg | -w default is no timeout | 16:51 |
| freepick | Maybe I did it wrong, when I saw the error I pressed alt+f2 and took the pastebin | 16:53 |
| rrq | freepick: the log is incomplete at the end; either due to debconf log buffering or due to nc clipping off at the end (exiting due to EOF before finish transmitting) | 16:54 |
| rrq | perhas you can step forward a bit an then copy the log ? | 16:55 |
| joerg | Jun 22 14:41:56 kernel: [ 15.731016] sdd: sdd1 sdd2 | 16:55 |
| freepick | ok, return in a moment | 16:56 |
| freepick | Is there any way to prevent this chat from being deleted? | 16:56 |
| joerg | copy it to somewhere different name | 16:56 |
| joerg | cd; cp /var/log/syslog . | 16:56 |
| rrq | the chat is logged at http://reisenweber.net/irclogs/libera/_devuan/ I think | 16:57 |
| joerg | ooh the chat, sorry | 16:57 |
| joerg | [22 Jun 2024 16:04:07] <joerg> http://reisenweber.net/irclogs/libera/_devuan/latest.log.html | 16:58 |
| freepick | :p I will be right back | 16:58 |
| * Xenguy notes that 'paste' is a pre-existing command coreutils ... | 17:04 | |
| Xenguy | *from coreutils | 17:04 |
| Xenguy | termbin.com looks handy | 17:05 |
| freepick | https://termbin.com/0r9q | 17:10 |
| freepick | now come to the end | 17:10 |
| rrq | ok | 17:11 |
| freepick | I can't find the error in that pastebin | 17:16 |
| fsmithred | partman-lvm: Physical volume "/dev/mapper/sdb3_crypt" successfully created | 17:17 |
| fsmithred | ^^^ Is that really what you wanted? | 17:17 |
| fsmithred | I thought you were working on /dev/sdc | 17:17 |
| freepick | never work on sdc since I use that 1tb disk for other things | 17:20 |
| freepick | in the pastebin above I am using a disk for testing, which is sda | 17:21 |
| fsmithred | ok, the first screenshot showed that sdc and sdd were chosen | 17:21 |
| freepick | At this point my mind has already turned off haha | 17:23 |
| fsmithred | and the last log shows you're working on /dev/sdb | 17:25 |
| freepick | order captures | 17:26 |
| freepick | 1 https://0x0.st/XAuO.jpg | 17:26 |
| freepick | 2 https://0x0.st/XAuV.jpg | 17:26 |
| freepick | 3 https://0x0.st/XAuW.jpg | 17:26 |
| freepick | 4 https://0x0.st/XAu4.jpg | 17:26 |
| freepick | 5 https://0x0.st/XAuJ.jpg | 17:26 |
| freepick | these screenshots work in sda as a test | 17:28 |
| joerg | Xenguy: thanks for pointing me to coreutols paste :-) | 17:29 |
| rrq | freepick: and tho previous /var/log/syslog belongs to those? | 17:30 |
| freepick | yes rrq | 17:31 |
| rrq | it's clearly wrong that sdd partition table change is in the list | 17:44 |
| joerg | llvm bug? | 17:44 |
| joerg | lvm even | 17:46 |
| rrq | well, its partman-lvm or partman-crypt or partman-something | 17:47 |
| rrq | mmm this is uefi .. maybe it's with that | 17:48 |
| rrq | freepick: possible to rerun with pendrive wrte-locked ? | 17:49 |
| rrq | write-locked | 17:49 |
| freepick | how do that? | 17:49 |
| rrq | some "pendrives" have write-lock.. some physical thing; maybe I'm old :) | 17:51 |
| joerg | or -o remount,ro ? | 17:51 |
| rrq | yes that might be an option... to enter shell already at network setup, and remount /cdrom as read-only | 17:52 |
| freepick | My pendrive does not have any physical button or switch. | 17:53 |
| rrq | btw, freepick you do need take the "load components" step even without selecting any of the additional ones | 17:53 |
| rrq | ok. | 17:53 |
| freepick | Then I go to expert install, and when do I put read-only mode? | 17:54 |
| rrq | yes, if you are up to it, then after having mounted the cdrom | 17:54 |
| freepick | Can't you edit it before booting and putting ro somewhere? | 17:55 |
| rrq | no.. after "detect and mount installation media"... then | 17:56 |
| rrq | mount -oremount,ro /cdrom | 17:56 |
| rrq | or it might need the full: mount -t iso9660 -o remount,ro /dev/sdd1 /cdrom | 17:57 |
| freepick | ok, after "detect and mount installation media" do I press alt+f2 and type one of these commands "mount -oremount,ro /cdrom or mount -t iso9660 -o remount,ro /dev/sdd1 /cdrom"? | 17:58 |
| rrq | yes | 17:58 |
| rrq | short form should work | 17:59 |
| rrq | then Alt-F1 and continue | 18:00 |
| freepick | ok | 18:01 |
| rrq | mmm maybe even: chmod a-w /dev/sdd | 18:03 |
| joerg | chmod a-w /dev/sdd* | 18:05 |
| joerg | ;-) | 18:05 |
| joerg | chmod -v a-w /dev/sdd* | 18:06 |
| rrq | mmm busybox might need: chmod 0400 /dev/sdd* | 18:10 |
| rrq | ah, yes, partman is buggy with UEFI | 18:13 |
| rrq | freepick: I found the test case... which is simply a UEFI boot | 18:14 |
| freepick | @rrq not work :( | 18:17 |
| rrq | ok. no worry. it's simply due to the UEFI boot... and it's correct to ignore the error dialog | 18:18 |
| freepick | life was easier without uefi, secure boot, systemd, and all those similar things haha | 18:19 |
| joerg | indeed | 18:21 |
| rrq | freepick: thanks for reporting and testing | 18:25 |
| freepick | v | 18:26 |
| freepick | You're welcome, happy to help, thanks to you for the support too. | 18:26 |
| joerg | freepick: feel free to stay here in this channel | 18:27 |
| joerg | 201 users :-) | 18:27 |
| freepick | Yes, I will stay here, and since I touched on the topic of secure boot, I have an Asus motherboard and in the secure boot section you can choose "other os", very good from Asus | 18:29 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
