[00:47] does anyone know anything about ajenti? === arlen is now known as arlen_ [02:30] is there some way to tell if I'm actually vulnerable to this? http://people.canonical.com/~ubuntu-security/cve/2014/CVE-2014-0118.html [02:30] tash: The deflate_in_filter function in mod_deflate.c in the mod_deflate module in the Apache HTTP Server before 2.4.10, when request body decompression is enabled, allows remote attackers to cause a denial of service (resource consumption) via crafted request data that decompresses to a much larger size. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2014-0118) [02:31] aside from checking to see if the mod is enabled or not. I definitely see that the mod is enabled, but is there some way to look inside the module code to see if said function is there? [02:32] tash: That CVE lists the package versions that contain the fix [02:34] TJ-, I understand that. But how can I find out the actual function is being used. Is it possible that it is commented out? That's what I'm getting at. [02:34] Would I have to talk to the developers of the site/app to determine if the vuln affects my stuff? [02:39] tash: which release/apache version are you using? [02:39] hey guys, which d-i option will skip asking me the "hostname" in preseed cfg file [02:39] tash: You can see the patches in the debian/patches/ directory of the source diff file published from the security team [02:40] tash: e.g. for Trusty, https://launchpad.net/ubuntu/trusty/+source/apache2/2.4.7-1ubuntu4.1 and the "apache2_2.4.7-1ubuntu4.1.debian.tar.gz" file [02:40] right now the pxe installation stops at the "hostname" setting pharse, waiting my press enter. [02:41] I just want to acceput the hostname which getting from dhcp/dns automatically [02:44] gambol: I'm not sure, almost guessing here, but have you considered "d-i netcfg/get_hostname seen true" [02:45] * gambol tries on TJ- 's tip... [03:25] TJ-, Thanks a lot [03:26] I finally figure it out, but it is not "seen true" issue for me [03:26] it is "priority=critical" in kernel boot [03:26] thanks a lot [03:27] gambol: I did wonder about that :) [05:05] Hi all. Hoping to upgrade LTS-to-LTS, and waiting for the .1 release... do-release-upgrade still tells me nothing is available. .1 was released yesterday I thought? === arrrghhh is now known as arrrghhhAWAY [05:37] arrrghhhAWAY: "a few hours ago" would probably be more accurate.. === brrr is now known as BRRR === BRRR is now known as brrr === unixoid|afk is now known as unixoid [07:37] has anyone had any success suspending a ubuntu VM to disk? (not using the hypervisor's pause feature) === a1berto_ is now known as a1berto === Mogwai is now known as Guest6472 [08:19] seems like disabling KSM in qemu-kvm is not working after all !! :( [08:19] Look like I am going to revert back to 12.04 - this is rubbish! [08:25] zetheroo, yes there is a problem in the 3.13 kernel - its being investigated now; for now the workaround is to disable KSM [08:26] zetheroo, that should be OK - let me find out the details === unixoid is now known as unixoid|afk === unixoid|afk is now known as unixoid [08:28] I disabled KSM yesterday ... it seemed to have solved the networking issue but then today it's starting again ... :P === support is now known as ashleydrees [08:29] jamespage: basically the guests are no longer pingable - 40% - 90% packet loss [08:29] zetheroo, that's inline with what other people have seen [08:29] and then a couple minutes later they are pingable again [08:30] zetheroo, normally this starts to happen after XXX period of time === support is now known as ashleydrees [08:30] right - for us it's happening a day after disabling KSM [08:30] zetheroo, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1346917 [08:30] Launchpad bug 1346917 in linux "Using KSM on NUMA capable machines can cause KVM guest performance and stability issues" [High,Fix committed] [08:31] is there anything I can do to keep things running? - Even if I have to pass a command to restart a service ever few hours - ... [08:32] zetheroo, still pinging folks - people have definately been able to workaround this problem and disable KSM completely [08:32] i need the cloud, how do i install it? === support is now known as ashleydrees [08:33] jamespage: is it enough to disable KSM in qemu-kvm ? [08:33] zetheroo, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1338277/comments/1 [08:33] Launchpad bug 1338277 in linux "Ubuntu 14.04 + QEmu 2.0 + KSM = 1, makes Windows 2008 R2 guests to crash (BSOD) (dup-of: 1346917)" [Undecided,Confirmed] [08:33] Launchpad bug 1346917 in linux "Using KSM on NUMA capable machines can cause KVM guest performance and stability issues" [High,Fix committed] [08:33] Will restarting the qemu-kvm service and libvirt-bin service every so often help? [08:36] jamespage: I have already disabled KSM in the qemu-kvm file ... shutdown all the guests and rebooted the host. But that was yesterday, and it seems to have worked until this morning. [08:36] zetheroo, you probably want todo it at the system level as well [08:37] jamespage: This morning I restarted both the libvirt-bin and qemu-kvm services and things are holding for the last 15 min ... [08:37] jamespage: At "system level"? [08:38] zetheroo, echo 0 | sudo tee /sys/kernel/mm/ksm/run [08:39] output = 0 [08:40] zetheroo, that's good - if cat /sys/kernel/mm/ksm/run returns "0" its disable in the kernel itself [08:40] yes, it returns 0 [08:41] zetheroo, so even if qemu tried to use KSM, the kernel won't do it now I think [08:41] zetheroo, see how that goes; it looks like the kernel team have a fix in the pipeline for the next kernel update which is good [08:41] but I turned KSM off yesterday ... and the networking issues reoccurred today [08:42] how long do you think before the fix update hits the mainstream repos? Days/weeks/months? [08:43] And it's happening again :P [08:44] zetheroo, gah [08:44] zetheroo, arges will be around later (he's US based) he would know the timescale better that I do [08:45] zetheroo, I appreciate that this is less than ideal [08:45] :) only slightly === Lcawte|Away is now known as Lcawte [09:32] rbasak, any thoughts on https://launchpadlibrarian.net/180674038/buildlog_ubuntu-utopic-arm64.mysql-5.6_5.6.19-1~exp1_FAILEDTOBUILD.txt.gz ? [09:33] I'm assuming that's some sort of compiler bug [09:50] zul, hallyn: I raised this - https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1348551 [09:50] Launchpad bug 1348551 in qemu "qemu-kvm upstart configuration from qemu-system-x86 relies on binary from qemu-kvm" [Undecided,New] [09:50] it seem to so obvious that I must be doing something wrong. [09:50] but utopic is the same [09:55] jamespage: I agree - compiler bug. [09:55] I'm not sure who could take that on. doko? Linaro? [09:56] jamespage: shall we hit the rebuild button and see if it reproduces? [09:57] rbasak, already did that [09:57] rbasak, raised https://bugs.launchpad.net/ubuntu/+source/gcc-4.9/+bug/1348560 [09:57] Launchpad bug 1348560 in gcc-4.9 "/build/buildd/mysql-5.6-5.6.19/storage/perfschema/pfs_host.cc:289:1: internal compiler error: in final_scan_insn, at final.c:2897" [Undecided,New] [09:58] Subscribed, thanks! [10:04] 14.04.1 is out right? [10:05] i'm still getting 'No new release found' on my precise servers... [10:07] has anyone had any success suspending a ubuntu VM to disk? (not using the hypervisor's pause feature) [10:32] mdeslaur: uploaded apache2 2.4.10-1ubuntu1 [11:22] rbasak: thanks! :) [12:35] rbasak, It looks remarkably like the nagios-nrpe merge is done \o/ Thank you for all the help [12:42] gnuoy: I ahdn't got to that yet! Looks like dholbach sponsored it for you. Nice one! === psivaa is now known as psivaa-afk [12:44] rbasak, how do I see who sponsored it ? [12:45] gnuoy: https://launchpad.net/ubuntu/+source/nagios-nrpe/2.15-1ubuntu1 [12:46] ah, I was lookking on the bug. thanks [12:46] You can construct that URL for any source package, or click through from https://launchpad.net/ubuntu/+source/nagios-nrpe [13:00] jamespage: You don't say in the bug what went wrong and what you expected; i have no idea what bug 1348551 is asking for [13:00] Launchpad bug 1348551 in qemu "qemu-kvm upstart configuration from qemu-system-x86 relies on binary from qemu-kvm" [Undecided,New] https://launchpad.net/bugs/1348551 [13:01] oh [13:01] hallyn, the check at the start of the pre-start script exits if /usr/bin/kvm does not exists [13:01] so none of the rest of the script executes [13:01] unless you have qemu-kvm installed [13:01] feh, i think this happened with one of hte recent pkg reorgs in debian [13:02] jamespage: any downside to using /dev/kvm? [13:03] oh yeah, won't work with old systems where kvm isn't autoloaded [13:04] but qemu-kvm is not a transitional package [13:10] apw: smb: does having kvm mounted at boot depend on a newer kernel, or a newer udev? [13:11] hallyn, you mean the kvm modules loaded automatically, that is kernel [13:14] smb: are there any support kernels left where that does not happen? [13:14] hallyn, Lucid probably [13:14] smb: i'mthinking for cloud archives, [13:14] do we still support any precise with 3.2 kernel? [13:15] hallyn, Let me check to make sure, but I think Precise already was having the alias setting [13:16] smb: that woudl be great [13:19] hallyn, Hm, no. I remembered that wrong [13:19] jamespage: do we support cloud archive with 3.2 kernel, or does cloud archive force a newer kernel? [13:19] So Precise would be the one which does not auto-load the kvm modules [13:20] smb: ok, so we could have qemu-system-common.postinst modprobe kvm i suppose. that's kind of ugly [13:21] or, we can make the check "if [ ! -c /dev/kvm -a ! -f /usr/bin/kvm ]; then { stop; exit 0; } fi;" [13:22] then we say that in precise you have to have qemu-kvm installed; minimize the problem at least [13:22] jamespage: ^ preference? [13:23] hallyn, 3.2 is supported with the cloud archive [13:24] hallyn, I'm not sure that device presense is the right thing todo here; the check is really to deal with when the package is removed but not purged [13:24] you want the upstart configuration to just no-op in this case [13:24] so checking for a binary provided by qemu-system-x86 would make sense IMHO [13:25] it's not only for qemu-system-x86 [13:25] hallyn, but its provided by that package [13:25] arm, arm64, ppc, [13:26] hm, maybe it should be provided by qemu-kvm [13:26] and made avaialble on the other arches [13:26] hallyn, is qemu-kvm not transitional then? [13:27] jamespage: all right, fine. I think it's ugly, but I'll change it to that. thanks. [13:27] no [13:27] mjt had wanted it to be :) [13:27] So, this means we never have ksm enabled by default on arm64 systems? [13:27] That's a shame [13:28] hallyn, qemu-system-common might be a better place? [13:28] I think so [13:28] script will need updating then to load the right modules for other arches; but that's worth doing. [13:28] hallyn, its depended on by all of the arch specific packages [13:29] what is [13:29] qemu-system-common [13:29] oh, yes. that's the point of it :) [13:29] indeed [13:29] hallyn, so right now the way the charms deploy nova-compute means that qemu-kvm upstart configuration never applies any changes AFAICT [13:30] hallyn, result is that KSM never gets enabled :-) which is why we did not see this issue on serverstack [13:30] this issue == KSM issue [13:31] "you're welcome" (j/k) [13:31] lol [13:31] yeah unfortunately this stuff got moved around quite a bit the last cycle or two, as debian is wanting to change things; i apparently lost track ofthat part [13:36] jamespage: fixed in git://anonscm.debian.org/pkg-qemu/qemu.git #ubuntu-dev [13:38] hallyn, +1 [13:39] looking to see if there are any other fixes i should push in along with it... [13:59] Can anyone recommend an alternative to dovecot? [14:03] I can't manage to set it up properly [14:03] zul, coreycb, matsubara, jamespage, gnuoy, gaughen, kickinz1, beisner-afk, rharper, lutostag-away, smoser, hallyn: ready for another merge sprint? [14:03] yes! [14:03] rbasak: \o [14:03] Everyone else: my team are going to do some virtual sprinting for the next couple of hours to get a bunch of server packages merged and sponsored. [14:04] Ready! [14:04] rbasak, yes [14:04] Feel free to watch, or join in, etc. There are sponsors here to try and help and get packages and other server fixes landed. [14:04] o/ [14:04] http://reqorts.qa.ubuntu.com/reports/ubuntu-server/merges.html is a list of ~ubuntu-server subscribed packages that may need a merge [14:04] https://merges.ubuntu.com/ has the full list [14:04] https://wiki.ubuntu.com/UbuntuDevelopment/Merging is a helpful guide [14:04] Shall we refresh the list of packages that everyone was working on? [14:04] How about a pad, actually? [14:04] I'll take pacemaker unless anyone has that already [14:04] rbasak: sure [14:05] http://pad.ubuntu.com/server-team-merges [14:05] was on keepalive, needed to finish discuss about, and maybe uploading. [14:05] rbasak: I left off with vgabios, confused because it appeared that all of the patches were already applied but quilt didn't think so. [14:06] rharper, my patch says "It looks like this package is maintained in revision control: ... You almost certainly don't want to continue without investigating". You had that last week right ? [14:07] * gnuoy goes to the irc logs [14:07] gnuoy: yes, that's common. [14:07] gnuoy: yeah; though that's not a big deal, just means we need to look at the patches [14:07] ok, sounds good. thanks [14:07] Some packaging is maintained in VCS. Openstack packaging, for example (I think?) [14:08] Most of the Ubuntu desktop packages are as well I think. [14:08] The VCS-* fields in debian/control are supposed to point to the VCS [14:08] jamespage: ok, finally fixed in all places I hope, thanks for raising that. [14:08] But we end up with those fields in packages with an Ubuntu delta, where strictly speaking it isn't the Debian VCS since we have diverged. [14:10] So what I'd like to try today is: [14:10] 1) Everyone pick a package, update the pad so we know what everyone is working on, and work on it. [14:10] 2) When blocked, please ask, and if I'm behind, put a note on the pad, so that I don't forget to get to you. [14:11] 3) I'll answer questions here on IRC, review and sponsor, etc, and work on clearing the "blocked" list. [14:12] 4) Anybody else who can mentor or sponsor, please do the same. [14:12] rbasak: I run grab-merge vgabios; the vgabios-0.7a-5ubuntu1/ already has the patches applied; to confirm this, I unpacked vgabios_0.7a.orig.tar.gz; copied in the debian/patches, then ran quilt push -a; and they all cleanly apply [14:12] Any suggestions, feedback, questions, comments on that process? [14:12] grabbing erlang [14:13] seems good to me [14:13] OK, so I'll try to go in order. kickinz1 first with keepalived. [14:13] ok [14:13] Meanwhile if anybody else can help rharper please do, otherwise I'll get to you asap. [14:14] kickinz1: so where are you with keepalived? [14:14] rbasak: sounds good [14:14] rharper, what do you have? [14:14] jamespage: vgabios [14:14] we end up with an old patch that was not applied any more, for a libnl3 compatibility [14:15] jamespage: there's a trivial conflict in the debian/control file (orig maintainer update) [14:15] rharper, OK [14:15] * beisner looks at asterisk [14:15] after resolving, the package builds fine as-is; but rbasak suggested that I try to apply the patches; but the odd thing was that all of the patches are applied in the -1ubuntu5 dir -- but not .pc dir [14:15] so quilt doesn't think they are [14:16] and a upstart modification that came with edgy release, was asking if those patches still need to be out of sync with debian. [14:16] rharper: ok lemme take a look over ehre [14:16] -end of memory- [14:16] kickinz1: OK, looking and trying to remind myself now [14:17] rharper: if it builds, then that suggests to me that the patches applied OK. [14:17] rbasak: same problem as rharper -- my autofs_5.0.8.orig.tar.bz2 already has debian/patches/* applied [14:17] rharper: I tend to ignore .pc - that's just quilt's internal state. I tend to "quilt pop -a" and rm -Rf .pc if in doubt [14:17] rbasak: they do actually, if I unpack the orig.tgz, and copy over the debian dir, quilt will appliy them all [14:18] rharper, I think there is a bug in the grab-merge tool or the bit serverside that generates one of the tarballs [14:18] dovecot is the only mail server for smtp? [14:18] cipher__: dovecot is primarily IMAP. postfix or exim for SMTP [14:18] * rbasak focuses on kickstartd [14:19] Uh, keepalived. [14:19] rbasak: nvm, my orig.tar is clean; I think I can move forward [14:19] rharper, I just hit the same thing with erlang [14:20] jamespage: yeah -- I tried with the grab-merges.sh versus the one in the dev tools; they do the same thing w.r.t patch appling [14:20] applying [14:20] so, what;s the next step then since it builds OK ? [14:20] rharper, generate a debdiff and give it to me for sponsorship [14:21] * beisner keeps picking doozies. grab merging asterisk sends a scary message [14:21] jamespage: ok, so debdiff between the debian version and the 1ubuntu5 version ? [14:21] using the .dsc files? [14:21] rharper, yep [14:21] ok [14:21] ie. *** WARNING *** It looks like this package is maintained in revision control ... You almost certainly don't want to continue without investigating. [14:22] beisner, don't worry to much - if the Vcs fields are debian ones thats fine [14:22] * jamespage listens to the fan spin up as erlang builds [14:28] kickinz1: AFAICT, the previous ubuntu_libnl3.patch was never applied because it didn't appear in debian/patches/series [14:28] rbasak, Does it take some time to update http://reqorts.qa.ubuntu.com/reports/ubuntu-server/merges.html? For python-flake8 it says the zul uploaded the last version (2.1.0-1ubuntu2) quite recently but when I run grab-merge on it, I get 2.1.0-1ubuntu1 instead [14:28] kickinz1: looking at the actual diff [14:29] matsubara: yeah i did it this morning === daker_ is now known as daker [14:29] matsubara: yes - both the report and grab-merge are processed only periodically. [14:29] zul, rbasak: ah, ok. I'll pick another one then. [14:29] kickinz1: also, my diff doesn't show the build-depends line changing either. [14:30] the only diff I saw was in the upstart. [14:30] kickinz1: I only see the postinst change, and the init script [14:30] don'tknow if it worth keeping a difference. [14:30] ANd update-maintainer. [14:31] And the patch file, but no change to the series file. [14:31] kickinz1: so it looks like other changes were dropped previously, and the changelog message was inaccurate. [14:31] kickinz1: I see no need to keep the patch file around unless it was actually used. Launchpad archives old source packages anyway, so they will always be available to anyone investigating. [14:32] kickinz1: so, it looks to me like the merge before was correct, except for the changelog message and an unneeded patch file. [14:33] kickinz1: do you want to prepare a debdiff against 1.2.13-1 for sponsoring, with only those changes, dropping the unneeded patch file, and with a changelog message that reflects only the changes in that new debdiff only? [14:33] Ok I'll do [14:34] I'll keep you informed if I'm blocked, thanks [14:36] Thanks. [14:36] I think nobody is blocked right now, right? [14:36] I'll work on exim4, then. [14:38] jamespage: yeah I agree that merge-o-matic is buggy [14:38] I see it for exim4 too [14:38] exim4_4.82.1-2ubuntu1.src.tar.gz contains patches already applied, but no .pc directory [14:43] is there a way to bind nfs to a certain IP? [14:45] rharper: hey, [14:45] rbasak, first pkg merge, still taking in the pieces. not necessarily blocked, other than by familiarization [14:45] rharper: so regarding vgabios, I woudl ignore grab-merges [14:45] hallyn: here [14:46] rharper: pull-debian-source and pull-ubuntu-source, and look at the ubuntu changelog - the only ubuntu delta was the last patch [14:46] beisner: sure, no problem. There's a ton of stuff to take in when you first start. [14:46] beisner: let me know if I can help [14:46] rharper: so, look at the contents of that patch, and see that it is all applied in the source tree now [14:46] rharper: if it is, then you can do a simple sync [14:47] hallyn: yeah, it looks like it should be a sync [14:47] so, how does one do the sync ? [14:47] rharper: agreed, so you can run 'syncpackage' [14:47] hm, i'm not sure how to do that in a sponsored way. [14:48] zul: ^ do you know how someone can do syncpackage in a way that someone can sponsor it? [14:48] sounds like I need to find a different package and let a sponser, like jamespage do it ? [14:48] hallyn: yes - you run syncpackage. There's a flag, hold on. [14:48] hallyn: not off the top of my head [14:49] i do see '-s', but that would be for me to use, not him, i assume :) [14:49] hallyn: -s to sponsor. [14:49] hallyn: right. You have to run syncpackage. rharper would technically file a sync bug to explain and request, but you could just do it for him and sponsor with -s. [14:49] I think that's how it's supposed to work. [14:49] surely we don't need a sycn bug [14:49] Right [14:50] I think it's fine to request and explain on IRC, and if you're satisfied, for you to JFDI with -s [14:50] ok - rharper your lp id is raharper right? [14:50] i'll do try that, thx [14:50] hallyn: raharper [14:50] rharper: you can meanwhile go on to a ne wpackage :) ipxe or seabios? :) [14:50] hallyn: hehe [14:51] ipxe could be a doozie [14:51] neither of those are on merges [14:51] someday we need to reconsile the symlinks between debian and ubuntu between ipxe and qemu :( [14:51] ok [14:56] Hi all. I've got a raid 10 through mdadm that just had a hiccup (now in the process of resyncing). This array holds the company files and webserver files. After/because-of the hiccup, the system fully locked up. I can't see any relevent info in syslog. Two questions: first, where can I check to deturman the reason for the full system lockup, and for the array problem? And seccond, can a raid10 mdadm array be used/access [14:57] rharper: so what i did was "syncpackage --force -s raharper vgabios" [14:57] rharper: in dpoing so, i noticed the debian package is orphaned === superboo1 is now known as superboot [14:57] you might want to consider adopting it :) [14:58] you can join us on oftc#debian-qemu and tell mjt how much you'd love to do so === psivaa-afk is now known as psivaa [15:09] hallyn: hehe [15:10] !info bind [15:10] Package bind does not exist in trusty [15:10] !info bind9 [15:10] bind9 (source: bind9): Internet Domain Name Server. In component main, is optional. Version 1:9.9.5.dfsg-3 (trusty), package size 273 kB, installed size 931 kB [15:10] hallyn: don't you have to be a debian/ubuntu dev to adopt a package ? [15:11] rbanffy, so, I took pep8 package instead, did grab-merge.sh and it has a ton of conflicts. I sorted all of them but one in pep8.py which seems that the ubuntu package introduce more code than there's in the debian package. [15:11] rbasak, ^ [15:12] :-) [15:12] matsubara, I just went through a WTF moment ;-) [15:12] rbanffy, sorry, tab completion failure [15:13] matsubara: OK, I'll take a look. [15:14] BTW, I'm told that merge-o-matic is broken with respect to the .src tarball when there is a conflict, with quilt patches applied but no .pc directory [15:14] rharper: no, I maintain netcf, i'm not a dm or dd. [15:14] http://people.canonical.com/~cjwatson/dpkg-quilt-setup is a workaround to fix that up [15:14] rbasak: you just need someone to sponsor every upload [15:14] matsubara, no problem. [15:14] i bet rharper could get mjt to sign on as his sponsor [15:15] hallyn: ok, I'll give it some thought; would be interested in going through the process a few times [15:15] rharper: cool [15:15] rbasak, this is the conflict I'm not sure how to sort out: http://pastebin.ubuntu.com/7856600/ [15:16] matsubara: OK so you have a special case here. [15:16] matsubara: first, look for the reason that Ubuntu diverged. [15:16] This probably applies to all merges. [15:16] In this case, it wasn't because we had some additional patches or anything. [15:16] It's because we needed a new upstream version ahead of Debian. [15:16] https://launchpad.net/ubuntu/+source/pep8/1.5.6-0ubuntu1 is the evidence of that. [15:17] The 0 in -0ubuntu1 is the standard for a version in Ubuntu that is not packaged in Debian. [15:17] matsubara: so that's why grab-merge is giving you strange results. [15:17] matsubara: the goal here is to resync with Debian, so to upload a new package to Utopic that has only the minimal changes that Ubuntu needs. [15:18] matsubara: as Debian now has a more recent upstream version that Ubuntu, we can probably just sync that package. [15:18] matsubara: all we have to look out for is that we're not going to regress anything in Ubuntu - is there anything in the current Utopic package that would be lost if we synced? [15:18] matsubara: if the answer is that there would be no regression, then we can sync. [15:20] matsubara: does that make sense? [15:22] rbasak, It doesn't look like we're going to regress anything by using debian 1.5.7, which is basically what I did sorting out the conflicts pointed out by grab-merge (that is, I kept the debian changes and deleted the ubuntu conflict markers) [15:22] rbasak, so how do I sync it? [15:23] matsubara: right. But we probably don't even need to examine the detail. At a high level, Ubuntu didn't introduce any change diverging from either Debian or upstream, so now that Debian has caught up, we don't need to examine conflicts manually. [15:23] matsubara: so https://wiki.ubuntu.com/SyncRequestProcess is the normal process, but that's just to request a sponsor, and we have sponsors available here. [15:23] rbasak, I see [15:23] Looks like I can't upload pep8, so we need to ask someone who can. hallyn, maybe please? [15:24] Or zul, as he last uploaded pep8? [15:24] sure just gimme the place where i can get it [15:25] zul: it's a sync :) [15:27] rbasak, I'm blocked... [15:27] kickinz1: OK. What's up? [15:27] I don't have a good process for debuild... [15:28] http://paste.ubuntu.com/7856653/ [15:28] sorry in french [15:29] too much dch -i think... (ubuntu4) [15:29] jamespage: swichted to etckeeper; grab-merge looks good, one conflict it couldn't merge, README file, kept the .UBUNTU version of the file; the REPORT says to run: dpkg-genchanges -S -v1.11ubuntu1 -- when I do, it complains that it can't find ../etckeeper_1.12ubuntu1.dsc -- as that's not there; what's the right next step then ? [15:30] kickinz1: looks like your quilt patches aren't applying cleanly. [15:30] kickinz1: before you build the package, "quilt pop -a" and "quilt push -a" should work without errors. [15:30] Also with no fuzz, but that doesn't seem to be the problem here. [15:31] rharper: "debuild -S -nc -S -sa -v1.11ubuntu1" or similar [15:31] ok, right now I've done quilt new upstart.patch, quilt shell , mod, then 'exit', quilt refresh debuild -S, not ok? [15:32] ok auilt pop -a & push -a working... [15:32] I think the problem is that grab-merge gives you an unpacked tree with patches applied, but no .pc directory, so quilt is lost from that point. [15:33] rbasak, I didn't used grab-merge.... [15:33] Oh, OK. [15:33] kickinz1: does "quilt pop -a" and "quilt push -a" work correctly in your source tree? [15:33] just pull-lp-source, and pull-debian-source [15:33] I'm not familiar with quilt shell. [15:33] rbasak: yes seems so, pop -a removes patch, push -a apply it [15:34] kickinz1: oh. [15:34] kickinz1: sorry I didn't notice before. [15:34] We don't use quilt for the debian/ directory. [15:34] rbanffy: ok -- I guess I missed the step where I should have built it before running that genchange [15:34] I use debuild -S -uc -us, not ok? [15:35] ok so I pop the patch [15:35] Instead of using quilt, change what you need in debian/ directly. [15:35] ok [15:35] I'm not sure if that's related to your problem or not. [15:37] so I restart. [15:40] I take the control directly from old ubuntu package, or I merge both of them?, newer entries only? [15:40] rbask ^^ [15:40] kickinz1: the debian/control file? [15:41] kickinz1: you need to take the Debian one, and re-apply the logical changes that still need to remain in the Ubuntu delta. [15:41] no was thinking of changelog, sorry [15:41] Oh [15:41] There's a tool to merge the changelogs [15:41] dpkg-mergechangelogs [15:41] dch ? [15:42] ok [15:42] It does a 3-way merge. [15:42] ok [15:42] Give it the Debian version before the last merge, the latest Debian version, and the latest Ubuntu version. [15:42] Then from the generated changelog, add a new changelog entry on top using dch. [15:43] grab-merge/merge-o-matic does this automatically, but you have to do it by hand if doing the merge by hand. [15:43] ok [15:47] I can't get original debian 1.2.7-1 [15:48] ok got it [15:49] forgot "1:" [15:53] rbasak: I'm trying to generate the debdiffs, the guide asks for one between the debian version and the one I've modified; I've built the package, so I have my updated deb, where do I get the debian version deb to feed to debdiff ? [15:54] rharper: how did you get the other sources? You're looking for .dsc files. [15:54] rharper: "pull-debian-source -d " can be used to download Debian source packages. [15:54] rbasak: right, I have .dsc files ... [15:55] rharper: run debdiff against .dsc files [15:55] bah, my fault; [15:55] that worked [15:55] missed a redirection for the output [15:57] rbasak: I'll send you the debdiff? [15:58] kickinz1: you can, but I have to run to a mysql call now. You can hold on to it until next week, or file a merge request bug if you like. [15:58] kickinz1: https://wiki.ubuntu.com/UbuntuDevelopment/Merging#File_a_merge_bug [15:58] ok np. [15:59] thx [16:00] I need to run now. I hope the session was useful! [16:00] yes, thx again [16:00] hi folks! [16:01] o/ [16:02] i have an issue with isc-dhcp-server on 14.04 lts. It doens't start at boot start. I need to start manualy the service to work. [16:04] do you know this issue? thanks! [16:07] rharper, ../merge-debuild [16:11] someone able to tell me what i'm doing wrong: [16:11] http://paste.ubuntu.com/7857020/ [16:12] i'm sure i'm doing something stupid. [16:12] !ubuntu+1|smoser utopic is supported in the other channel [16:12] smoser utopic is supported in the other channel: Utopic Unicorn is the codename for Ubuntu 14.10 - Support only in #ubuntu+1 [16:13] cfhowlett, not really. this is a development channel too. [16:13] ha! [16:17] jamespage: thanks! [16:17] rbasak, i know you're not here, but see http://paste.ubuntu.com/7857020/ for my failure. [16:18] jamespage: should I file the merge bug now, or maybe share the two debdiffs with you first ? (I'm looking at etckeeper now) [16:18] just ping me the debdiffs [16:18] k [16:19] smoser: sorry i've not yet used adt with lxc. pitti and stgraber are probably your best bet [16:23] jamespage: debian_1.12_to_1.12ubuntu1.debdiff -> http://paste.ubuntu.com/7857095/ ; ubuntu_1.11ubuntu1_to_ubuntu_1.12ubuntu1.debdiff -> http://paste.ubuntu.com/7857097/ [16:25] rharper, so.... [16:25] when merging I normally start with the previous merge changelog and figure out which bits can be dropped, if any [16:26] the changelog should detail what the delta is and why its required, so that the next merger can read that and re-check [16:27] hey folks, i'm trying to remember the name of a linux app that would generator a sequenece of letters for oyu based on some parameters you provide to it [16:27] like all a00 - a99 [16:29] crunch! [16:36] zul, can you take a look at https://bugs.launchpad.net/ubuntu/+source/nova/+bug/1219658 ? [16:36] Launchpad bug 1219658 in nova "Wrong image size using rbd backend for libvirt" [Undecided,In progress] [16:36] I'm nearly eod and can't face it right now [16:36] :-) [16:41] jamespage: can't face it? ;) [16:41] zul, its hot and I need a beer [16:41] jamespage: you need air conditioning [16:42] then you can have a beer and look at the bug [16:42] zul, not worth it for the 2 weeks of hot weather we get a year [16:42] jamespage: its totally worth it for us [16:42] zul, is a sponsorshup request [16:42] jamespage: ack [16:44] jamespage: uploaded that already hasnt been acked by the SRU team [16:45] zul, ah - OK [16:46] jamespage: just put a little more meat on it...go enjoy your beer [16:46] ta [16:57] jamespage: I take it those debdiffs are too big then? === tkeith_ is now known as tsrk [17:22] Does ufw by default limit the number of connections in some way? It looks like it based on this iptables rule: -A ufw-user-limit -m limit --limit 3/min -j LOG --log-prefix "[UFW LIMIT BLOCK] " [17:22] jamespage: where? [17:22] But I'm no iptables expert (hence me using ufw) so I don't really know what's going on [17:23] tsrk: that just limits logging [17:23] RoyK: Oh ok, that makes more sense, thanks! [17:23] tsrk: so that ufw doesn't flood the logs in case someone (or a lot) nmaps you :P [17:24] RoyK: Makes sense, thank you! [17:24] tsrk: it will limit logging (controllable via ufw logging). you can use the 'limit' command in place of 'allow' to have rudimentary limiting (see man ufw for details) [17:26] it'd be nice to get ULOG support in ufw... [17:30] All I've done on a server is "ufw default reject; ufw allow 22; ufw allow 443; ufw enable". There shouldn't be any reason that ufw is restricting some incoming ssh connections is there? I'm intermittently seeing SSH connections be rejected, but it might be a network issue. [17:40] unfortunatelthere is no ULOG support for ipv6 [17:40] and they won't add it cause all effort is on nftables [17:41] tsrk: that shouldn't block ssh. you could try sudo /usr/share/ufw/check-requirements if you are running a non-ubuntu kernel [17:41] if anything fails, then need to update the kernel configuration for it [17:42] jdstrand: I'm running a Linode kernel, so I ran those tests but they all passed [17:42] ok, well, that is good :) [17:43] maybe there is something you need to do with security groups or something to make sure that traffic passes [17:44] but if you disable ufw and can login, that wouldn't be it [17:46] tsrk: never seen ufw drop ssh traffic, really [17:46] if it is, you should see something in /var/log/ufw.log === unixoid is now known as unixoid|afk [18:08] hallyn, https://bugs.launchpad.net/ubuntu/+source/autopkgtest/+bug/1348749 [18:08] Launchpad bug 1348749 in autopkgtest "autopkgtest fails sometimes with adt-virt-lxc" [Undecided,New] [18:15] smoser: any useful logs under /var/log/lxc ? [18:17] hallyn, no. [18:17] i really think i'ms eeing overlayfs race :-( [18:19] smoser: very possible. Can you do it without --ephemeral? [18:19] or is that baked into adt? [18:20] i pass '--ephemeral' [18:23] right, and what happens if you don't [18:58] !info bind9 [18:58] bind9 (source: bind9): Internet Domain Name Server. In component main, is optional. Version 1:9.9.5.dfsg-3 (trusty), package size 273 kB, installed size 931 kB [19:14] Good evening. [19:16] any idea how to mount partition 1 in this thing? Can't figure out the correct losetup -o data... http://paste.ubuntu.com/7858237/ [19:17] RoyK, i might try mount-image-callback [19:18] but if you want to try with '--offset' [19:18] you should : [19:18] sfdisk -uS -l Windows.iso [19:18] sectors. [19:18] then, whatever it puts in that start is *512 [19:18] what is mount-image-callback? [19:19] from cloud-utils. [19:19] mount-image-callback Windows.iso /bin/bash [19:19] err.. with sudo [19:19] then [19:19] echo $MOUNTPOINT [19:20] do whatever you want, exit and it will tear down. [19:20] or: [19:20] mount-image-callback Windows.iso -- tar -C _MOUNTPOINT_ -cvzf contents.tar.gz [19:26] i am facing proble with 2 nic on 12.04 and 14.04.1 when i install two nic with dhcp internet become very slow ,but if i disble 1 card then it work as expect what could be aproblem === edwardly is now known as edwardlz [19:27] samba35: both NICs on the same subnet? === edwardlz is now known as edwardly [19:27] no [19:28] it should be on same sabnet ? [19:28] no [19:28] both using dhcp? [19:28] yes [19:28] pastebin "ip addr list", please, when both NICs are connected [19:29] ok later /may be today its 1 am in india now i am planning to go to bed now [19:29] sorry ,i have line up some download and update with 1 nic [19:30] i will get back to you today only later [19:30] it's rather hard to help you debugging something when you can't give us data ;) [19:30] anyway [19:31] yes true but its night here i have open chassic and connect card then other member @home will should @me [19:31] if both NICs are on dhcp, the dhcp servers may give you default route on both nics, which may mess things up a bit [19:31] samba35: both interfaces are in the same subnet? [19:31] please ..understand ,if you are married i hope you undersrand [19:31] no [19:32] samba35: maybe the wrong interface is the default route? [19:32] anyway - with two default gateways, if that's what you end up with, it'll be a mess [19:33] samba35: If interface A is supposed to route to the internet, but starts before interface B, then A's default route will be replaced by one from B if DHCP is handing out a gateway [19:33] i was trying to setup openvswitch with 1 card and 1 card for standalone system [19:33] if you want something like load balancing or failover, go to http://www.lartc.org/ [19:35] ok ,royk ,tj i will get back to you today later ,if 1 am here in india [19:36] ok [19:36] nite ;) [19:36] sorry ... [19:36] np [19:36] if you are married you will better understand :) [19:37] I'm not, but I have been living with partners, so I know, even though I don't have children [19:37] again sorry RoyK [19:38] again sorry TJ- [19:38] bye === mjohnson151 is now known as mjohnson15 [21:02] Any idea where to find the MD5 sum of the Ubuntu Server 14.04.1 release? Or any Ubuntu 14.04.1 release for that matter? [21:27] Stern: http://releases.ubuntu.com/trusty/ [21:28] Thanks [21:28] Integrity check of USB stick seems to fail. Just want to be sure the download is ok. [21:30] Download is ok. Second USB stick has the same problem. [21:45] I have a windows machine. I want to do dual boot (Ubuntu server 14.04). I messed up the first time. So, I'm wiping out the entire disk. Might as well do things properly now. Should I set up two disks? === quix_mac_ is now known as quix_mac === Beliq_ is now known as Beliq === sarnold_ is now known as sarnold === beisner- is now known as beisner === freeflying__ is now known as freeflying === Ursinha_ is now known as Ursinha === NomadJim_ is now known as NomadJim