=== Guest59153 is now known as kklimonda
[08:50] <CIA-7> debian-installer: cjwatson * r947 hardy-proposed/ (9 files in 4 dirs): Move to 2.6.24-31 kernels.
[08:51] <CIA-7> debian-installer: cjwatson * r948 hardy-proposed/debian/changelog: releasing version 20070308ubuntu40.15
[08:54] <CIA-7> debian-installer: cjwatson * r1337 lucid-proposed/ (8 files in 2 dirs): Move to 2.6.32-41 kernels.
[08:58] <CIA-7> debian-installer: cjwatson * r1338 lucid-proposed/ (9 files in 3 dirs): Move natty-* images to 2.6.38-15 kernels.
[08:59] <CIA-7> debian-installer: cjwatson * r1339 lucid-proposed/ (9 files in 3 dirs): Move oneiric-* images to 3.0.0-22 kernels.
[09:02] <CIA-7> debian-installer: cjwatson * r1340 lucid-proposed/debian/changelog: releasing version 20081029ubuntu102.16
[09:06] <CIA-7> debian-installer: cjwatson * r1444 natty-proposed/ (5 files in 2 dirs): Move to 2.6.38-15 kernels.
[09:07] <CIA-7> debian-installer: cjwatson * r1445 natty-proposed/debian/changelog: releasing version 20101020ubuntu29.2
[09:10] <CIA-7> debian-installer: cjwatson * r1552 oneiric-proposed/ (build/config/common debian/changelog): Use udebs from oneiric-security, oneiric-proposed, and oneiric-updates.
[09:22] <CIA-7> debian-installer: cjwatson * r1553 oneiric-proposed/ (5 files in 2 dirs): Move to 3.0.0-22 kernels.
[09:23] <CIA-7> debian-installer: cjwatson * r1554 oneiric-proposed/ (build/config/armel/omap4.cfg debian/changelog): Move armel/omap4 to 3.0.0-1212 kernels.
[09:24] <CIA-7> debian-installer: cjwatson * r1555 oneiric-proposed/debian/changelog: releasing version 20101020ubuntu72.1
[09:38] <CIA-7> debian-installer: cjwatson * r1682 precise-proposed/ (build/config/common debian/changelog): Use udebs from precise-security, precise-proposed, and precise-updates.
[09:41] <CIA-7> debian-installer: cjwatson * r1683 precise-proposed/ (6 files in 2 dirs): Move to 3.2.0-26 kernels.
[09:43] <CIA-7> debian-installer: cjwatson * r1684 precise-proposed/ (build/config/armel/omap4.cfg debian/changelog): Move armel/omap4 and armhf/omap4 to 3.2.0-1415 kernels.
[09:46] <CIA-7> debian-installer: cjwatson * r1685 precise-proposed/ (build/config/armel/armadaxp.cfg debian/changelog): Move armhf/armadaxp to 3.2.0-1604 kernels.
[09:52] <CIA-7> debian-installer: adconrad * r1686 precise-proposed/ (build/config/armel/armadaxp/netboot.cfg debian/changelog): Fix armadaxp netboot uImage/uInitrd install paths (LP: #1010708)
[09:55] <CIA-7> debian-installer: cjwatson * r1687 precise-proposed/debian/changelog: releasing version 20101020ubuntu136.1
[10:11] <xnox> Jul  4 09:47:07 ubuntu plugininstall.py: Exception during installation:
[10:11] <xnox> Jul  4 09:47:07 ubuntu plugininstall.py: SystemError: E:Write error - write (32: Broken pipe),
[10:11] <xnox> with broken pipe repeated into infinity until ubiquity crashes
[10:12] <xnox> is this the fallout from run away packages or something is wrong with the langpacks?
[10:12] <cjwatson> Yes, gema reported that last night but I didn't have time to investigate
[10:12] <cjwatson> My suspicion is an apt/python-apt regression
[10:12] <cjwatson> Fancy looking into it
[10:12] <cjwatson> ?
[10:12] <xnox> ok, I will try
[10:13] <cjwatson> I actually rather doubt it has anything to do with the packages being installed; you'll probably need to attach strace -f at some point shortly before that and then pore through that
[10:13] <xnox> cjwatson: my only "problem" is that I still didn't figure out how to run ubiquity interactively under pdb and step through things
[10:13] <xnox> hmm... ok
[10:13] <xnox> i see
[10:13] <cjwatson> I doubt that would be of any use in this case
[10:13] <xnox> ok
[10:13] <cjwatson> Given that I strongly suspect it's in apt's C++ code
[10:13] <cjwatson> Seeing as it's bubbling up as a SystemError rather than a sensible Pythonic exception
[10:14] <xnox> cjwatson: was gemma's report on this channel or in #ubuntu-testing?
[10:14] <cjwatson> #ubuntu-release
[10:14] <xnox> ok, no bug number?
[10:14] <xnox> this basically prevents me from further LVM work, unless I use old images...
[10:14] <cjwatson> Doesn't look like it
[10:15] <cjwatson> Oh, bug 1020574
[10:15] <ubot2> Launchpad bug 1020574 in ubiquity "installing ubuntu in VM or real hardware" [Critical,Confirmed] https://launchpad.net/bugs/1020574
[10:15]  * cjwatson gives that a marginally more sensible title
[10:16] <infinity> A quick grep suggests it's a broken pipe in the decompression bits.
[10:16] <infinity> apt-pkg/contrib/fileutl.cc
[10:16] <infinity> (Honest, I'm asleep)
[10:17] <xnox> infinity: thanks =))) so you don't sleep walk, you sleep code? =)
[10:18] <infinity> Explains most of the armhf port.
[10:18] <xnox> =)))
[10:18] <xnox> from jenkins build log this looks interesting:
[10:18] <xnox> Jul  3 13:18:54 ubuntu apt-setup: warning: /usr/lib/ubiquity/apt-setup/generators/91security returned error code 1; discarding output
[10:19] <xnox> i'm failing to see how that could have failed?!
[10:21] <cjwatson> DEBCONF_DEBUG=developer should reveal why
[10:21] <cjwatson> (assuming you can reproduce it)
[10:25] <xnox> ok
[10:25] <cjwatson> Or of course set -x
[10:59] <xnox> $ find . -name 'plugininstall.py'
[10:59] <xnox> ./scripts/plugininstall.py
[10:59] <xnox> ./ubiquity/components/plugininstall.py
[10:59] <xnox> interesting naming schemes ;-)
[11:00] <cjwatson> The latter is an integration hook for the former
[11:20] <xnox> cjwatson: wasn't bug 882882 already fixed?
[11:20] <ubot2> Launchpad bug 882882 in ubiquity "installer crash with ValueError: invalid literal for int() with base 10: ''" [Undecided,Confirmed] https://launchpad.net/bugs/882882
[11:20]  * xnox is hitting that now *sigh*
[11:20] <cjwatson> That's a generic symptom with many possible causes.
[11:21] <cjwatson> A bit like saying 'wasn't "Segmentation fault" already fixed?'. :-)
[11:23] <xnox> =))))
[11:23] <xnox> but http://pad.lv/977750 is open
[11:27] <cjwatson> What the one-line description of that exception basically means is "the backend process went away for some undetermined reasson"
[11:27] <cjwatson> *reason
[11:27] <cjwatson> I'm not convinced whatever that is with update-notifier is related
[11:28] <cjwatson> And I definitely don't think it's worth going to any effort whatsoever to find duplicate or related bugs
[11:28] <cjwatson> It will be several times quicker to investigate directly
[11:28] <cjwatson> In the case of update-notifier it's *possible* that it means that the debconf stdin/stdout-based protocol has got out of sync somehow, or file descriptors have got muddled
[14:04] <CIA-7> partman-target: cjwatson * r932 ubuntu/ (13 files in 3 dirs): merge from Debian 78
[14:05] <CIA-7> partman-target: cjwatson * r933 ubuntu/ (debian/changelog finish.d/clear_partitions):
[14:05] <CIA-7> partman-target: finish.d/clear_partitions: Don't remove /var/lib/mysql or
[14:05] <CIA-7> partman-target: /var/lib/mythtv, since they contain user data in the form of databases
[14:05] <CIA-7> partman-target: (LP: #992241).
[14:07] <CIA-7> partman-target: cjwatson * r934 ubuntu/debian/changelog: releasing version 78ubuntu1
[14:11] <CIA-7> partman-target: cjwatson * r932 precise-proposed/ (debian/changelog finish.d/clear_partitions):
[14:11] <CIA-7> partman-target: finish.d/clear_partitions: Don't remove /var/lib/mysql or
[14:11] <CIA-7> partman-target: /var/lib/mythtv, since they contain user data in the form of databases
[14:11] <CIA-7> partman-target: (LP: #992241).
[14:13] <CIA-7> partman-target: cjwatson * r933 precise-proposed/debian/changelog: releasing version 77ubuntu2.1
[18:57] <stgraber> cjwatson: good, got ubiquity to explode with the transparent proxy, now to confirm that it's fixed with the SRU :)
[19:45] <stgraber> cjwatson: commented on bug 998492, I'm not sure whether to consider this verification-done or verification-failed. I believe the main bug was fixed but ubiquity is clearly not dealing too well with broken http
[19:45] <ubot2> Launchpad bug 998492 in ubiquity "Fails to detect package download errors on architectures other than amd64" [High,Fix committed] https://launchpad.net/bugs/998492
[20:57] <jamin> having some trouble automating an install via preseed.  I'd like to encrypt most of the drive using partman-auto/method crypto.  I've written a partitioning recipe that works with the lvm method, but it constantly fails with the crypto method claiming there is no physical volume defined in the volume group, recipe: http://paste.ubuntu.com/1075445/
[21:05] <xnox> jamin: which physical devices are in the volume group?
[21:06] <jamin> /dev/sda is called out in the preseed....
[21:07] <jamin> d-i     partman-auto/disk       string  /dev/sda
[21:07] <jamin> and the recipe works as-is if I set the method to lvm
[21:07] <xnox> ok
[21:08] <jamin> I've tried adding a device section to it, but no change in behavior
[21:09] <xnox> jamin: hmmm with crypt i have only use partman-auto with: d-i partman-auto/method string crypto
[21:10] <xnox> that sets up crypt & lvm on top with separate unencrypted boot
[21:10] <xnox> with d-i partman-auto/choose_recipe select multi
[21:10] <xnox> separate /home, /usr, /var, and /tmp partitions
[21:10] <jamin> which is what this recipe should do too
[21:11] <xnox> not sure about how to do this correctly manually  / using the expert recipe
[21:11] <jamin> with lvm it creates a separate boot and the rest an lvm volume group
[21:11] <xnox> jamin: parman crypto is always crypto+lvm
[21:11] <xnox> partman-auto does not do crypto without lvm on top
[21:12] <xnox> so partman-auto crypto is actually: /boot on ext && crypt -> lvm -> (partitions)
[21:12] <jamin> right, it should be creating the same lvm groups that the lvm method would create but using crypto too
[21:12] <jamin> right
[21:12] <xnox> jamin: yeah so change the method to crypto, and keep lvmok and the rest of things =/
[21:12]  * xnox haven't done this
[21:13] <jamin> that's exactly what I did, recipe is the same in both cases, just changed the method
[21:16] <jamin> I did get it working with a more basic recipe, but don't see the difference that is breaking it
[21:17] <jamin> http://paste.ubuntu.com/1075477/
[21:19] <xnox> this one doesn't actually does lvm
[21:20] <jamin> not directly no, but it does work with method crypto
[21:20] <xnox> sorry, i am no help. will try this some time later.
[21:20] <jamin> it does say $lvmok{ }
[21:21] <jamin> that just gave me an idea
[21:21] <xnox> so, that only optionally allows this partition to be on top of lvm-group
[21:21] <jamin> I'm removing all specific lvm configuration from the desired recipe
[21:21] <xnox> since there is no, format {lvm} anywhere, there lvmok is spuriout
[21:22] <jamin> you would think
[21:22]  * xnox finds recipes confusing beyond partman-auto
[21:22] <xnox> if you get it to work please paste here
[21:22] <jamin> insanely... sadly I don't like any of the automatic partitioning schemes
[21:22] <xnox> other people want this and struggle e.g. http://askubuntu.com/questions/24079/full-disk-encryption-with-preseed
[21:22] <xnox> unless that is you as well ;-)
[21:23] <jamin> nah, not me, but I did use the answer there as my basis
[21:23] <jamin> which is why I started with LVM and got it working
[21:25] <jamin> it's insane but that change worked
[21:26] <xnox> hmm?
[21:26] <xnox> jamin: paste, please =)
[21:26] <jamin> about to... one sec
[21:28] <jamin> http://paste.ubuntu.com/1075493/
[21:28] <jamin> That one works
[21:28] <jamin> it shouldn't
[21:28] <xnox> jamin: $ pastebinit path/to/file.txt
[21:28] <jamin> but it does
[21:28] <jamin> sweet, I didn't know that existed
[21:28] <jamin> was doing it manually
[21:29] <xnox> jamin: or it takes a pipe $ command1 | pastebinit
[21:29] <jamin> that recipe lacks any real lvm definition
[21:29] <xnox> in ubuntu package it defaults to ubuntu paste ;-)
[21:29] <jamin> but it creates one and works
[21:30] <jamin> insane
[21:30] <xnox> crypto? =)
[21:30] <jamin> yes
[21:30] <xnox> huh?!
[21:30] <jamin> as in, it solved my problem...
[21:31] <xnox> can you paste the one that works with crypto? or is it just partman-auto/method crypto?
[21:31] <jamin> prompted for a crypto phrase and is installing right now
[21:31] <xnox> what's the rest of the preseed? (unless it's private)
[21:31] <jamin> http://paste.ubuntu.com/1075493/ <-- works with  partman-auto/method crypto
[21:32] <xnox> jamin: confusing, but awesome =)
[21:32] <xnox> well not confusing, but rather unintuitive
[21:32] <jamin> indeed
[21:32] <jamin> I've been fighting this for the better part of a day