[00:45] <jrwren> i just had the kswapd0 process using high CPU usage. echo 1 > /proc/sys/vm/drop_caches fixed it. I'm on 4.4.0-22-generic. Anyone else seen this?
[00:59] <nacc> jrwren: you just took away memory pressure
[00:59] <nacc> jrwren: not seen any high kswapd here, but not heavily loaded
[01:42] <tom[]> somehow i screwed up my lo interface during an update. i don't know how. `ip addr show lo` has odd output shown here together with my inyerfaces file https://gist.github.com/anonymous/48985686561070a55e55a00ea503b716
[01:42] <tom[]> what is the correct command to bring lo up?
[07:45] <lordievader> Good morning.
[10:35] <LeMike> hello. I have installed ubuntu-server but now the resolution is very low. so the screen shows the terminal only in the upper left quarter of the screen. is there a way to solve this? I am working on an old fujitsu futro and don't know a thing about it, neither the max possible resolution.
[10:56] <lordievader> Sounds like modesetting is not happening.
[10:56] <lordievader> Then again, why not use ssh?
[11:16] <AtuM> Hello... Can someone tell me what is the best way to put a line "rdma 20049" into /proc/fs/nfsd/portlist ? RedHat made a service called nfs-rdma for this.. does Ubuntu have anything similar?
[11:17] <AtuM> I've put it into rc.local, but it disappears after service restart.. so I need something more persistent.. perhaps a unit file that would add the option after nfs-kernel-server restart
[11:33] <jamespage> ddellav, coreycb: blue balls everywhere
[11:33] <jamespage> :-)
[11:51] <coreycb> jamespage, lol thank you for the early morning giggle
[12:01] <gelbeEnte> Hello! I have some problems with "ecryptfs-mount-private": "Signature not found in user keyring. Perhaps try the interactive 'ecryptfs-mount-private'". My home folder doesn't decrypt. I tried "ecryptfs-rewrap-passphrase" but nothing changed. Is it possible to decrypt?
[12:06] <AtuM> ecryptfs-recover-private
[12:07] <gelbeEnte> Inserted auth tok with sig [0496470557959537] into the user session keyring
[12:07] <gelbeEnte> ERROR: The key required to access this private data is not available
[12:09] <gelbeEnte> :(
[13:17] <ruben23> hi guys any one familiar with ddos deflate..?
[13:37] <cpaelzer> rbasak: there is a dovecot related package called dovecot-antispam
[13:37] <cpaelzer> rbasak: it is not build from the same source, it looks like a sync to me
[13:37] <cpaelzer> rbasak: debian and Ubuntu only have rebuilds from time to time
[13:37] <cpaelzer> rbasak: now that we will "up" dovecot given the merge takes place
[13:38] <cpaelzer> rbasak: we will need to rebuild (or resync) that as well
[13:38] <cpaelzer> rbasak: how to coordinate that best ?
[13:38] <cpaelzer> rbasak: reference http://paste.ubuntu.com/17641820/
[13:46] <rbasak> cpaelzer: good job spotting that. We don't really have a process for dealing with this, but if the dependencies are right dovecot will end up stuck in proposed until antispam is rebuilt - looks like this is the case.
[13:47] <cpaelzer> rbasak: yes it depends
[13:47] <rbasak> cpaelzer: sponsors should be willing to accept whatever you want to do here to point it out. For me, I suggest just pointing out that it needs a no-change rebuild in the merge proposal, and the sponsor can do it at the time. Or if you want the credit for the no change rebuild, supply a debdiff for that too (dch -R is a handy shortcut)
[13:47] <cpaelzer> rbasak: it never breaks, as on a dist-upgrade without a rebuilt dovecot-antispam it gets removed (as it reaslized it is no more compatible)
[13:48] <cpaelzer> rbasak: since it is a no-change-sync instead of a no-change-rebuild I think I'll mention it on the sponsor request
[13:48] <rbasak> cpaelzer: right, so proposed migration should detect that moving dovecot to the release pocket would cause dovecot-antispam to become uninstallable, and thus refuse to do it.
[13:48] <rbasak> cpaelzer: ah right. Yes, sorry. I forgot you mentioned that.
[13:48] <cpaelzer> rbasak: ok, will do so and we will pick up on any issues that will show up along the way
[13:48] <cpaelzer> thanks
[13:49] <rbasak> cpaelzer: I don't think it'll be syncable. It'll need to be a no change rebuild upload.
[13:50] <rbasak> cpaelzer: note the output of "rmadison -u debian dovecot-antispam". Debian has done binNMUs for it, so the source is still actually on 2.0+20150222-1
[13:50] <rbasak> Ubuntu doesn't have binNMUs.
[13:52] <cpaelzer> rbasak: that is the +b4 in what I linked?
[13:53] <rbasak> cpaelzer: that's an indicator - Debian convention (maybe policy?) for binNMU versioning.
[13:53] <rbasak> cpaelzer: also note though that "source" doesn't appear against +b4, only against the plain -1.
[13:54] <rbasak> cpaelzer: you can confirm with chdist. I did "hcdist sid apt-cache show dovecot-antispam" and see that it has "Version: 2.0+20150222-1+b4" but "Source: dovecot-antispam (2.0+20150222-1)".
[13:54] <cpaelzer> rbasak: that is a good one to note, thanks
[13:55] <rbasak> nacc: ^ you might be interested in this edge case too. Shouldn't impact the importer since it doesn't need to care about binaries and Ubuntu doesn't have binNMUs, but an interesting edge case nontheless.
[13:56] <cpaelzer> rbasak: I'll do a debdiff with a rebuild request then - does that has to go to an extra bug ?
[13:56] <cpaelzer> the merge I'd associate to the functional issue that drove me to create it
[13:57] <cpaelzer> I wonder if I can add extra package to the bug and then add the debdiff for the rebuild as well
[13:57] <cpaelzer> rbasak: ^^
[13:59] <rbasak> cpaelzer: if you wish. We don't really *require* bugs for uploads; it's just generally handy to track progress of the sponsorship request over time, a place for review feedback, and an item that can go into the sponsorship queue. I wouldn't bother with an additional bug, but if the sponsor misses it *and* you want it in the sponsorship queue then you may end up creating one anyway. As a sponsor,
[13:59] <rbasak> I'll accept a debdiff for this without a bug :)
[14:38] <coreycb> jamespage, what do you think about installing the theme files into /usr/share/openstack-dashboard/openstack-dashboard-ubuntu-theme/static/ubuntu instead of /usr/share/openstack-dashboard-ubuntu-theme/static/ubuntu?
[14:39] <coreycb> that may fix the SuspiciousFileOperation issue
[14:53] <bitfury> hey all good morning
[14:56] <bitfury> does anyone know if the trusty-security universe repos contain security updates for universe software?
[14:56] <jrwren> what happens when my LVM thinpool fills?
[14:59] <rbasak> bitfury: yes, where security updates for universe have been contributed.
[15:00] <bitfury> got it, thank you rbasak
[15:05] <cpaelzer> this tests wants to be enjoyed I think - autopkgtest.enjoE5
[15:05] <cpaelzer> to the power of 5
[15:16] <tom[]> what comand(s) can i run to apply (changes to) interfaces config files in the same way that system init applies the configs in those files?
[15:21] <nacc> rbasak: yeah, i've seen them before, thanks for the heads-up
[15:22] <nacc> tom[]: you mean /etc/network/interfaces?
[15:22] <nacc> tom[]: you presumably want to reload or restart the networking service
[15:23] <tom[]> probably
[15:23] <rbasak> nacc, tom[]: you don't want to restart networking. Things can explode: http://askubuntu.com/a/376586/7808
[15:24] <nacc> rbasak: thanks for the link
[15:24] <rbasak> The right way is to ifdown the interface(s) being changed, *then* change the interface files, then ifup again. From a screen, and carefully.
[15:24] <tom[]> kreikey
[15:24] <rbasak> Or reboot.
[15:25] <tom[]> ifdown requires the config files to correspond to the current interface state?
[15:25] <rbasak> Yes. It's a bit backwards.
[15:25] <tom[]> i'm reacting to your *s
[15:25] <rbasak> Otherwise, for example, dhclient won't be stopped for an interface switching from DHCP to static, since it won't know to do that later.
[15:26] <rbasak> There are various efforts to make this better. All of which deprecate ifupdown I think. But that won't be coming for a while.
[15:27] <tom[]> i'm trying, among other things, to understand what happened when i screwed up my loopback on a server yesterday
[15:27] <rbasak> It's a pretty challenging problem to solve when you consider complex network setups that include bonds, bridges, vlans, tunnels and so on. The dependency tree is pretty complex.
[15:27] <tom[]> sure. i'm glad i don't deal with any of those complex configs
[15:28] <nacc> rbasak: i suppose, then, it's not entirely true that restarting networking won't work (as i've done it many times). But it may not work depending on how much eni was changed
[15:28] <rbasak> nacc: it also depends on what you mean by "restarting networking" as to whether it'll do what you want or not.
[15:29] <nacc> rbasak: yep, that makes sense
[15:29] <rbasak> So yeah, it might work as expected in certain cases, but certainly not in the general case.
[15:29] <nacc> rbasak: and i'm sure there are dependencies that may not re-trigger properly (or at all)
[15:29] <rbasak> Although I hear the dbus issue is fixed, it's also not unreasonable for running programs to explode when the loopback interface disappears underneath them.
[15:30] <rbasak> So then "restart networking" clearly shouldn't cycle loopback interfaces, and then we're back to the definition problem.
[15:32] <nacc> yep, it makes sense to be more nuanced in what you're trying to control
[15:32] <patdk-lap2> clearly shouldn't?
[15:39] <nacc> rbasak: ok, so re: php-horde-db, shall I file a `requestsync`?
[15:39] <rbasak> nacc: no need, I synced it.
[15:39] <nacc> rbasak: ok, wasn't sure, thanks!
[15:50] <cpaelzer> nacc: about the importer I think I would change the doc once you ack about the following
[15:50] <cpaelzer> nacc: "Go to Launchpad page for USD git project and request a merge review."
[15:50] <cpaelzer> to
[15:51] <cpaelzer> nacc: "Go to the Launchpad page of your just uploaded merge branch and propose for merging into the USD git project"
[15:51] <cpaelzer> nacc: ?
[15:51] <cpaelzer> or is it really reverse than I expected
[15:51] <nacc> cpaelzer: ack, that's correct
[15:52] <nacc> please fix :)
[15:57] <cpaelzer> nacc: while at it I should provide guidance about "target repository" and "Target reference path"
[15:57] <cpaelzer> nacc: is that already in the readme files and I could copy from there ?
[15:57] <cpaelzer> my first assumption of "~usd-import-team/usd-importer/+git/usd-importer" was wrong ... :-/
[15:58] <cpaelzer> nacc: I'll open a quikc HO so you know what I'm asking about
[16:00] <nacc> cpaelzer: sorry, missed the pings, yeah we should provide better instructions for that part
[16:00] <nacc> for target repo, just search by the source package name
[16:01] <nacc> and it should find the corresponding usd-import-team tree
[16:01] <nacc> cpaelzer: and for target path, use 'debian/sid' (presuming that's what you merged to -- use whatever referenced you merged to in the process)
[17:07] <Kallis> Could anyone possibly help me with my openvpn setup, I can connect fine but I cannot access internet when connected or the local shares on the VPN server
[17:24] <magicalChicken> rbasak: I can take a couple more bugs this week probably. I was looking through the list and I saw 1296835, which I can take care of pretty quick
[17:24] <magicalChicken> if there are any other simple-ish ones i can probably do one or two more
[18:30] <rbasak> magicalChicken: thanks! I hadn't prepared any more, but I'll find a couple more for you tomorrow.
[18:30] <rbasak> There's a big enough backlog - I just pick a few in advance of the meeting every week.
[18:32] <magicalChicken> Thanks. I'm getting a bit quicker using the tools so hopefully I can get more bugs patched now
[18:47] <jhobbs> I have an openstack cloud I've deployed with juju charms, and have an internal network and a provider network. By default, only the internal network's interface is being brought up in instances that have both networks attached. What's the best way to get the second interface to come up also?
[19:21] <coreycb> jamespage, ddellav: here's my fix for the mitaka-proposed error - https://git.launchpad.net/~ubuntu-server-dev/ubuntu/+source/horizon/commit/?h=stable/mitaka&id=55e2f7ec3666d66d5cbc7e22d12f90bcd60163fa
[19:21] <coreycb> I just tested it and the theme seems to be fine
[19:48] <akira1980> What would be the way to do this,  I'm wanting to have a private email relay server. Basically pull all the mail from the mail server as pop then have the secondary server as imap and keep all the mail on that one. It's there a better solution?
[20:48] <concerti> is there an equivalent to SAR preinstalled on ubuntu servers?
[20:49] <mgriffin> concerti: you can install sysstat
[20:50] <concerti> yep that's the plan , just making sure there is not something like that already running w/ ubuntu
[20:50] <mgriffin> concerti: look in cron log?
[20:50] <mgriffin> concerti: and ps
[21:42] <RoyK> concerti: just install sysstat and enable it in /etc/default/sysstat
[21:42] <RoyK> concerti: for some reason debian/ubuntu doesn't enable it by default