| cyphermox | what dir? | 00:12 |
|---|---|---|
| cyphermox | ahoneybun: the commands I gave you will do everything for you; after you've branched lp:ubiquity, you 'cd ubiquity', then 'make -C d-i/ update' and what is required will be downloaded automatically for you. This is necessary because what ubiquity does is make use of parts from d-i, they need to be in sync rather than having to do changes in two places when they affect both desktop and server | 00:18 |
| cyphermox | http://paste.ubuntu.com/23109690/ | 00:18 |
| cyphermox | ^ this is part of a typescript of me running through the steps, it works. | 00:18 |
| ahoneybun | I know the logic, just saying on LP it does not | 00:45 |
| ahoneybun | I know that is works | 00:45 |
| cyphermox | I don't understand what you're saying; ubiquity clearly builds fine here, and in the archive. if it doesn't work for you, best is to share exactly what you do, somewhere like in a pastebin. I'm going to bed now though | 01:38 |
| CarlFK | http://archive.ubuntu.com/ubuntu/dists/trusty/main/installer-amd64/current/images/netboot/ubuntu-installer/amd64/linux and initrd.gz | 04:44 |
| CarlFK | segfault in libresolve-2.19 | 04:44 |
| CarlFK | but not right away. it is able to wget my early script.. hmm.. stand by for more. | 04:45 |
| CarlFK | anna[3859]: DEBUG: retrieving libc6-udeb 2.19-0ubuntu6.9 ... anna[3859]: DEBUG: retrieving libcryptsetup4-udeb 2:1.6.1-1ubuntu1 | 05:23 |
| CarlFK | kernel: [ 32.624724] wget[5138]: segfault at 14 ip 00007f2ac2defd46 sp 00007fff68531710 error 6 in libresolv-2.19.so[7f2ac2de4000+17000] | 05:23 |
| CarlFK | not sure what more or where to report. | 05:25 |
| jalt | Hi, how can I debug ubiquity/success_command post-preseeded-install? I had it working fine and now it doesn't seem to be running anymore, and I'm not sure what changed. I checked the logs at /var/log/installer and I can't seem to find anything related to success_command. Any tips? | 13:54 |
| cyphermox | you could run the installer again with 'debug-ubiquity' added on the kernel command-line as you boot; that will write a bit more stuff to the logs in /var/log/installer/debug. | 14:01 |
| cyphermox | perhaps one of the commands necessary wasn't found? | 14:01 |
| cyphermox | or failed early so the rest of the script didn't run? | 14:01 |
| jalt | that was my initial thought, but now there is only a single command. i'll add debug-ubiquity and see if more info comes out. | 14:03 |
| cyphermox | what's the command? | 14:04 |
| jalt | eject ;) | 14:04 |
| cyphermox | have you looked at syslog too? | 14:04 |
| cyphermox | hrm | 14:04 |
| cyphermox | I suppose eject may have changed a bit? I recall some bug from a while ago | 14:05 |
| jalt | all was working fine yesterday, and the changes i did to the preseed file should have not caused any trouble. | 14:05 |
| jalt | nah, eject was working well for the past two weeks | 14:05 |
| cyphermox | right, then the other changes must have affected it ;) | 14:06 |
| jalt | i probably messed up somewhere, but it's hard to debug without any useful info in /var/log/installer | 14:06 |
| cyphermox | otherwise, if you're doing installs of the devel release, other things in the set of packages we install may have broken | 14:06 |
| cyphermox | jalt: I usually look at syslog there; and much less dm | 14:06 |
| jalt | no, xubuntu 16.04.1 desktop iso | 14:06 |
| cyphermox | syslog typically has quite a lot of the required information | 14:07 |
| jalt | it has no information about the success_command | 14:07 |
| jalt | debug has some info about some of the debconf stuff, but that's about it | 14:08 |
| cyphermox | the debug log would have to mention it; success_command is retrieved from debconf | 14:08 |
| jalt | it never mentioned it, even when it was working | 14:08 |
| cyphermox | can you share the logs? | 14:08 |
| jalt | sure, once the current install finishes | 14:08 |
| jalt | in the meantime, here's the preseed: http://pastebin.com/7TCBtQwR | 14:10 |
| jalt | hmm logs are too big for pastebin | 14:19 |
| jalt | the debug-unity added some stuff that shows that the success_command was read from the preseed file, but nothing about it running or not. | 14:21 |
| cyphermox | jalt: should eject really be in-target? | 14:36 |
| cyphermox | seems to me like it probably doesn't need to be | 14:36 |
| cyphermox | (in-target is a chroot, with /dev probably bound-mounted into it; eject on the "real" system should work fine) | 14:37 |
| jalt | hmm that is a good point. in any case, i restored an older version of my preseed and that one worked | 14:37 |
| jalt | the diff is minimal, most of it comments, so maybe it's some issue with spaces or newlines... i'll investigate a bit further. thank you for the help cyphermox | 14:38 |
| cyphermox | did you change the tasksel line? | 14:38 |
| jalt | nope, that one isn't even read. i keep it there as a label mostly | 14:39 |
| cyphermox | oh, right, you're using a CD with a livefs, obviously | 14:39 |
| jalt | actually a dd'ed usb drive, but same difference. | 14:41 |
| jalt | hmm i might have spot the culprit: a stray space between string and the command though it should not matter to have it preceding the command... ("Put only a single space or tab between type and value: any additional whitespace will be interpreted as belonging to the value.") | 14:50 |
| cyphermox | spaces in commands shouldn't matter | 15:14 |
| jalt | and yet they did and it's now working again, cyphermox. | 15:22 |
| cyphermox | interesting | 15:27 |
| jalt | ¯\_(ツ)_/¯ | 15:27 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!