| armouredheart | Hello! I am trying to install the "amd64_netinstall.iso" for Beowulf. I made sure to verify integrity and the checksums of the disk and downloaded file. Both are valid and not corrupt. However I get the error "failed to load installer components from cd-rom" and the built-in "check the cd-rom integrity" fails. Any ideas as to diagnosing a solution? | 04:40 |
|---|---|---|
| gnarface | just a few guesses about common pitfalls... | 04:42 |
| gnarface | try not specifying a block size with dd and make sure to run "sync" afterwards | 04:42 |
| gnarface | make sure the iso was not compressed | 04:42 |
| golinux | In a netinstall wouldn;t those files be dl from the net? Maybe disable the cdrum line? | 04:42 |
| golinux | (sorry for the typos . . .) | 04:43 |
| golinux | gnarface will probably have the right answer . . . | 04:43 |
| adhoc | armouredheart: where did you get the files from? | 04:45 |
| armouredheart | I have been using the various torrents on the devuan website. I've gone through two cds so far to rule out bad disks, and three torrents. | 04:48 |
| adhoc | so how are you booting from teh ISO image ? | 04:50 |
| armouredheart | I put the iso on the cd, and then put the cd into the computer I want to install on. | 04:51 |
| adhoc | ok | 04:58 |
| adhoc | what kind of CD device is it? scsi, ide, usb attached ... ? | 04:58 |
| gnarface | someone recently had an issue with a relatively new Dell that still wouldn't boot from a flash drive larger than 2GB, that's the only other thing i can think to try | 05:04 |
| gnarface | could be hardware failure too, but usually at this step it is user error | 05:06 |
| * blockhead scrolls up and re-reads. So, an actual CD is being burned from the iso? no dd-ing to a usb stick. If so, what program was used to burn the image to the CD? | 05:11 | |
| armouredheart | Sorry for delay the cat wanted cuddles. | 05:16 |
| armouredheart | I used the program and settings "cdrecord -v -sao". | 05:16 |
| armouredheart | The cd devices are dvd-rw and have worked fine in the past for burning gentoo and artix discs. | 05:16 |
| armouredheart | The computers in question have built-in drives and are from 2005 if that helps. | 05:18 |
| gnarface | hmm, cdrecord might need -swab ? | 05:19 |
| gnarface | have you used the cdrecord command successfully for boot before? | 05:20 |
| armouredheart | the cdrecord command was used to burn artix and gentoo cds, both of which work. I have gentoo on my desktop machine, artix on laptop, and I am trying to put devuan on this amd opteron (first generation 64-bit x86 winter 2005) to use as a server | 05:23 |
| * adhoc had so much greif with early opterons | 06:10 | |
| rrq | armouredheart: ig the desktop or laptop have cd readers you can byte-compare the media and iso | 06:19 |
| rrq | you have confirmed the iso sha256sum I guess: 1723cbbeb1aee26a54e1370b688e7dc03921402348d2a60086c58c18cd9cf24b | 06:20 |
| armouredheart | yep | 06:21 |
| armouredheart | I just double checked | 06:22 |
| armouredheart | exactly the same as what you posted rrq | 06:22 |
| rrq | ok. and the opteron seems happy to boot those other cd but not devuan... or it boots but then have problems ... using the normal "Install" option? | 06:46 |
| rrq | btw you have byte-compared the media with the iso as well? | 06:53 |
| * rrq scanned the backlog.. ok | 06:56 | |
| rrq | when it says " failed to load installer components" you may use C-A-F2 to possibly figure out why. | 06:57 |
| rrq | it *should* have /dev/cdrom available to load those installer components from | 07:00 |
| armouredheart | I have updated info. | 07:01 |
| armouredheart | I have done some tinkering and the specific error is "file:///cdrom/dists/beowulf/main/binary-amd64/Packages was corrupt" | 07:01 |
| armouredheart | Judging by the perfect checksum, I'd say that the release version itself may be corrupt. | 07:02 |
| rrq | the sha256sum for that file is in file:///cdrom/dists/beowulf/Release: ca7c24292535fdd96f5763bb875cb919b69330483eca05f7d42cda47aaee7b7f | 07:03 |
| rrq | in my copy of the iso the summing agrees | 07:04 |
| armouredheart | I have no idea how to check that one specific file while the cd is running. | 07:04 |
| rrq | C-A-F2 | 07:05 |
| rrq | gives a command shell | 07:05 |
| rrq | which has sha256sum | 07:05 |
| armouredheart | Ctrl Alt F2? | 07:05 |
| rrq | the path should be /dev/cdrom/dists/beowulf/main/binary-amd64/Packages | 07:05 |
| rrq | yes | 07:05 |
| rrq | (I'm lazy typer) | 07:06 |
| rrq | might suggest a shonky cd reader | 07:06 |
| rrq | usb would be "safer" if available | 07:07 |
| armouredheart | Ok yeah the checksum is different | 07:09 |
| rrq | right but you verified the media, which suggests the reader lies | 07:10 |
| rrq | there used to be those "cleaning discs" | 07:11 |
| armouredheart | cleaning disks? | 07:11 |
| armouredheart | So the problem is one of the disk drives? | 07:12 |
| rrq | yeah.. not sure if they were any good because the problem is usually that the laser is aged | 07:12 |
| rrq | the reader h/w yes ... some readers enjoed getting a good kick but other took real offence to manhandling | 07:13 |
| armouredheart | The target computer has all 2005 parts | 07:13 |
| armouredheart | Strange that this never happened with the gentoo and artix disks? | 07:14 |
| rrq | same type of media? | 07:14 |
| armouredheart | yep, same package of disks too | 07:14 |
| armouredheart | from the same box all the same age | 07:15 |
| armouredheart | artix install is most recent this past july | 07:15 |
| rrq | and those other disks work on this machine now? | 07:15 |
| rrq | on that reader I mean | 07:16 |
| armouredheart | Only this specific devuan disk fails. The other disks were fine on this reader. | 07:16 |
| rrq | are they fine now? | 07:16 |
| rrq | sometimes it's just some dust speck on the mirror system | 07:17 |
| armouredheart | I tried two different disks with devuan and both reported this file as corrupt, but I had gentoo partly installed (I gave up because old style bios bootloaders are a pain in the ass) | 07:18 |
| rrq | do you have another reader? how did you verify the media? | 07:18 |
| armouredheart | I have 3 computers. | 07:20 |
| armouredheart | -IBM T500 | 07:20 |
| armouredheart | -Asus 2015 (with some long stupid name) | 07:20 |
| armouredheart | -AMD opteron tower. | 07:20 |
| armouredheart | All these machines have dvd-rw drives | 07:20 |
| rrq | is that file bad on all/some other reader? | 07:20 |
| armouredheart | You know, I didn't even think to check that. Just a moment | 07:21 |
| armouredheart | The checksum passed on my T500 machine. The problem is my opteron's disk reader it seems. | 07:35 |
| armouredheart | This is the first time i've encountered a bad reader | 07:36 |
| rrq | we can form a club ;) | 07:37 |
| armouredheart | Thank you for the patient help! I'll have to think of another way to install this thing | 07:38 |
| rrq | usb? or another reader? or install onto the disk via a sata caddy on the laptop? | 07:43 |
| rrq | netboot might work though that might want to install components as well | 07:44 |
| rrq | given that installer starts, that in itself is a linux with an initrd, though rather minimal ... | 07:46 |
| rrq | it does have busybox v1.30.1 and commands fdisk, mkfs.ext4, nc and tar .. so you could transfer stuff onto a local partition and "fake" the cdrom access | 07:58 |
| rrq | assuming ethernet though (or non-firmware dependent) | 07:59 |
| armouredheart | I'll see if I can get a cheap usb drive at the shop | 08:09 |
| adhoc | the instructions on https://beta.devuan.org/os/keyring | 08:28 |
| adhoc | for using gpg reference a dead key server | 08:28 |
| onefang | armouredheart: Just keep in mind the last person that had your issue, think it was tracked down to "old computer BIOS couldn't handle USB drive bigger than 2GB". | 08:30 |
| adhoc | a boot drive bigger then 2GB at the time should not be an issue though | 08:31 |
| adhoc | there ceratinly were issue with USB thumb drives in larger sizes though | 08:32 |
| armouredheart | onefang: Noted | 08:32 |
| onefang | I'm only saying what I recall. More details will be in the channel archives / scroll back / other peoples memories. | 08:33 |
| adhoc | yeah, fair enough | 08:33 |
| adhoc | there certainly were very few MoBos that had BIOSs that could boot off USBs back then | 08:33 |
| adhoc | 2005 certainly was the peak of IDE PATA CDROMs | 08:34 |
| rrq | adhoc: which non-dead gpg key server should it refer to? | 08:35 |
| adhoc | rrq: i don't know | 08:36 |
| adhoc | rrq: are there ones? or is it a DNS resolution thing ? | 08:36 |
| adhoc | the domain exists | 08:37 |
| adhoc | I get a 403 to http://gnupg.net/ | 08:37 |
| adhoc | and this wierdness; | 08:38 |
| adhoc | $ host -t any keys.gnupg.net | 08:38 |
| adhoc | keys.gnupg.net descriptive text "GnuPG uses an internal mapping for this name, see dirmngr/server.c." | 08:38 |
| adhoc | rrq: the question (for me at least) is if that is the official way to verify gpg keys, and you can't get to them, how do you verify anything ? | 08:40 |
| rrq | constructive advice is preferrable | 08:41 |
| adhoc | ok, cool, so is there another reference to follow? | 08:41 |
| adhoc | how can I help fix that ? | 08:41 |
| adhoc | what can I do to be more constructive/helpful ? | 08:42 |
| adhoc | I have had a number of folks ping me about this recently | 08:43 |
| rrq | yeah it looks like someone with a little bit of interest in the matter needs to do something soon | 08:57 |
| djph | adhoc: in theory, the keyservers all tend to share info; so you _should_ be able to ask pgp.mit.edu or keyserver.ubuntu.com and get the key | 11:22 |
| sadoon_albader[m | Anyone else getting a 503 certificate expired when downloading anything from kernel.org? | 11:38 |
| gnarface | make sure your system clock is right | 11:39 |
| sadoon_albader[m | It is | 11:39 |
| sadoon_albader[m | Tried multiple systems | 11:39 |
| sadoon_albader[m | Including my phone | 11:39 |
| gnarface | oh, you're right | 11:40 |
| gnarface | something is whack at their end | 11:40 |
| gnarface | the home page loads but the tarballs don't | 11:40 |
| sadoon_albader[m | Damnit | 11:40 |
| sadoon_albader[m | Just when I need to build a kernel :) | 11:40 |
| gnarface | you just want to build a custom kernel? probably easier to use the source package anyway | 11:40 |
| gnarface | apt-get source linux-image-amd64 | 11:40 |
| gnarface | or whatever is appropriate for your arch | 11:40 |
| sadoon_albader[m | I need 5.4 | 11:41 |
| sadoon_albader[m | Apt source will give me 5.10 | 11:41 |
| gnarface | 5.10 is later than 5.4... it has to be exactly version 5.4? | 11:41 |
| sadoon_albader[m | Yep :) | 11:41 |
| sadoon_albader[m | There's a bug in 5.10 | 11:41 |
| gnarface | well that's a bummer | 11:41 |
| sadoon_albader[m | When I run a vm with pcie passthrough of a gpu, it crashes | 11:41 |
| gnarface | you sure debian's 5.10 doesn't have the patch? they do include a lot of patches that vanilla upstream doesn't ... | 11:41 |
| sadoon_albader[m | Idk if it's only with my specific configuration | 11:41 |
| gnarface | oh, you've tested already | 11:42 |
| sadoon_albader[m | I tried | 11:42 |
| gnarface | hmm, damn | 11:42 |
| gnarface | maybe poke your head in #kernel and see if they know | 11:42 |
| gnarface | i'm assuming #kernel moved here to libera chat too, i guess i don't know for sure though | 11:42 |
| sadoon_albader[m | I'll try that, thanks :) | 11:42 |
| gnarface | good luck | 11:43 |
| adhoc | djph: that is great, i can not update the page, so folks will still bug me or folks on this channel when the instruction there don't work. | 11:51 |
| djph | adhoc: oh, sorry I misread the conversation :( | 11:51 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!