[00:20] <nacc> rbasak: fyi, drbd8 is falsely reported at http://reqorts.qa.ubuntu.com/reports/ubuntu-server/merges.html
[01:51] <adlaistevenson> On a server with an SSD drive (so writes a worry because they count against lifespan), what size should I make the swap if I have 10 GB of RAM? Server is running Postgresql RDBMS.
[01:57] <sarnold> I don't think people are too worried about ssd drives wearing thin these days except for really high sustained write loads; one hopes for swap that it wouldn't be used too much
[01:57] <sarnold> one or two gigs is probably good enough to give the kernel some space to shove data it doesn't need often
[02:08] <ubuntu16t> can an old unsupported ubuntu 12 upgrade to 16 using a cdrom?
[02:09] <ubuntu16t> or does the cdrom require overwrite of the old ubuntu?
[02:11] <sarnold> you should be able to use do-release-upgrade  to upgrade to 14.04 LTS and then apt-get update, apt-get -u dist-upgrade, and then another do-release-upgrade to get to 16.04 LTS
[02:20] <madLyfe> so does 17.04 get moved to LTS eventually? not sure how that works
[02:20] <sarnold> no; the next LTS is almost certainly going to be 18.04 but I don't know if that's set in stone or not
[02:21] <sarnold> the usual pattern is an lts release every two years, in april
[02:21] <madLyfe> ok, ty
[02:22] <sarnold> looks like that pattern has held since 8.04 LTS https://wiki.ubuntu.com/Releases
[02:24] <madLyfe> so what will 18 have that 17 doesnt?
[02:25] <sarnold> 17.10 will be the first ubuntu release with gnome as default desktop environment in many years.. I expect it to be a bit rough. hopefully 18.04 LTS will be a bit smoother.
[02:25] <sarnold> otherwise i expect mostly the same; slight improvements across the board.
[02:28] <madLyfe> why not just do LTS to LTS?
[02:28] <madLyfe> skipping the middle?
[02:28] <sarnold> some people prefer to have newer software on a more frequent basis
[02:28] <sarnold> and it's nice to get bugreports on those newer features before a new LTS
[02:31] <madLyfe> so its like a beta
[02:32] <sarnold> in many ways, yes
[02:32] <madLyfe> does this work for anyone? https://instant.io/
[02:32] <madLyfe> like the actual download. supposed to work but ive never been able to get it to download anything
[02:33] <madLyfe> trying to get the most recent ISO
[02:33] <madLyfe> https://instant.io/#145b85116626651912298f9400805254fb1192ae
[02:33] <sarnold> I can't even resolve it
[02:34] <madLyfe> s
[02:34] <madLyfe> http://i.imgur.com/3OrelbN.png
[02:37] <sarnold> hah, interesting. .io is publishing an nsec3 record for instant.io to prove that it doesn't exist http://dnsviz.net/d/instant.io/dnssec/
[02:39] <madLyfe> not sure what that means
[02:39] <madLyfe> site loads fine for me just doesnt work.
[02:39] <sarnold> i've heard before that .io is terrible at running a domain
[02:40] <JaguarDown> Any idea why internet works (update/upgrade, SSH, VPN, etc) DNS resolution "appears" to resolve names but I can't ping any servers? Wondering because I'm setting up mail and the server can't resolve the recipient's email domain
[02:40] <madLyfe> same with this one: https://btorrent.xyz/
[02:41] <sarnold> JaguarDown: it's common for overzealous firewall admins to block ICMP and thus break ping
[02:42] <JaguarDown> Okay I am the firewall admin but I'm a newbie/intermediate (home server) I shall look
[02:42] <sarnold> JaguarDown: qemu "user" networking also breaks ICMP in this fashion; so a VM running with just the right config could do it :/
[02:42] <JaguarDown> Not on a VM
[02:46] <JaguarDown> Sarnold would be correct, I can now ping. Firewall had been set on high security while I went back and opened ports I needed, turns out I forgot to open up ICMP :-)
[02:46] <JaguarDown> Thank you sarnold
[02:47] <sarnold> JaguarDown: great :)
[03:16] <madLyfe> sarnold: sudo apt install linux-generic-hwe-16.04 or sudo apt install –install-recommends linux-generic-hwe-16.04
[03:17] <sarnold> madLyfe: probably the one with install-recommends
[03:22] <madLyfe> sarnold: https://gist.github.com/986a2c95ab12e029938ff8f6db2949b2
[03:25] <madLyfe> "If you fall into the latter category but want the new HWE stack, then you can install it with a single command (server users should omit the xorg package): sudo apt install –install-recommends linux-generic-hwe-16.04 xserver-xorg-hwe-16.04"
[03:26] <sarnold> madLyfe: I never use 'apt' bare so I don't know what command line options it takes. the manpage doesn't document any. try apt-get install --install-recommends
[03:28] <madLyfe> you put two dashes in front?
[03:28] <sarnold> most decent tools use two dashes for long options and one dash for short options
[03:30] <madLyfe> https://gist.github.com/f1ddd7088fcf129c9efdb9fb014d4695
[03:31] <sarnold> looks good. interesting I wouldn't have expected thermald.
[03:32] <madLyfe> here is the full output: https://gist.github.com/11619c37f334005d1681295aa940d8dd
[03:32] <madLyfe> did i bork something there at the end?
[03:32] <sarnold> eww
[03:33] <madLyfe> eww what?
[03:33] <sarnold> that's really gross. I wonder how that even happened.
[03:33] <sarnold> re-run the command and see what happens.
[03:34] <madLyfe> sudo apt-get install --install-recommends linux-generic-hwe-16.04
[03:34] <madLyfe> that one?
[03:35] <sarnold> yeah
[03:37] <madLyfe> looks better, maybe? https://gist.github.com/ff6bff79437980c648b29057d835d039
[03:38] <sarnold> yes, much better
[03:38] <madLyfe> what not? reboot?
[03:38] <madLyfe> now*
[03:38] <sarnold> yeah I think so
[03:39] <madLyfe> i dont need to be physically there do i? im just SSHing in.
[03:39] <sarnold> if it works out fine you don't :) but if something goes wrong it's always nice to have remote console access to fix things up
[03:40] <madLyfe> remote console access = ?
[03:42] <sarnold> a remove vnc or ipmi or serial access so that you can control grub
[03:42] <madLyfe> ah. ya i was worried about grub. these boards dont have ipmi
[03:46] <madLyfe> looks like its all good: https://gist.github.com/344df2892f195ef40e85d167eece6bf1
[03:46] <sarnold> excellent
[03:48] <madLyfe> ty for the hand holding
[03:58] <madLyfe> sudo reboot
[03:58] <madLyfe> lol whoops
[04:10] <JaguarDown> I tried to set up email before and failed. Trying again now with new tutorial, just successfully sent an email by entering commands locally via telnet connection. I guess the real test will come later, sorry but this is exciting! :-)
[04:11] <sarnold> \o/
[04:12] <sarnold> when you see how simple it is to send email with netcat but how hard it is to configure working mail servers it's enough to make you throw the stupid computer out the window :)
[04:12] <sarnold> "just type what I type! do that!"
[04:15] <adlaistevenson> Or if you ever have a job where you have to know what DKIM is and answer fucking questions about why the e-mail went to spam, or worse, never arrived at all. I hate email.
[04:16] <adlaistevenson> And of course, you can't say you don't know why the recipient server gave a 250 OK and then discarded the message.
[04:16] <sarnold> yeah. I'm glad to let google handle my email :(
[04:16] <sarnold> it got hard.
[04:18] <sarnold> you get to have the same conversation over and over again.. "yes you can't send email to amail list from a gmail account to other people with gmail accounts. yes they know about it. no they don't care. because they can't sell advertising on email that's why."
[04:18] <adlaistevenson> I occasionally get dragged into email guessing games, and did today. There goes days of getting anything useful done.
[04:18] <sarnold> "email guessing games", heh
[04:18] <adlaistevenson> Now I will be spending days applying voodoo magic I don't understand like whitelabeling and IP warm-up and crossing my fingers.
[04:20] <adlaistevenson> And pray I can get back to actually developing software sometime this week (the only thing I am marginally competent at).
[04:20] <sarnold> good luck :)
[04:23] <adlaistevenson> Thanks
[04:57] <JaguarDown> lol well we shall see. I am making an attempt to run my own email just as a hobby for now, with the intent of using it primarily, but I guess we'll find out if it's worth the trouble. Surprisingly it's going a lot smoother now than before.
[05:04] <JaguarDown> Well postfix, dovecot, and mysql works, at least sending email over the internet, receiving it locally, and IMAP. That's enough for tonight. I'll wait to figure out why mysql log time is 4 hours in the future later...
[05:08] <JaguarDown> oh apparently I can also receive mail from the Internet too, nice. Anyway, thanks for the help sarnold, until next time, bye.
[05:35] <cpaelzer> good morning
[05:48] <cpaelzer> jamespage: the openvswitch pkg-config patch is now on the 2.8 branch as well - so on your next sync from git you can drop the custom patch
[06:22] <jamespage> cpaelzer: \o/
[06:22] <jamespage> I'll snapshot again today
[08:48] <cpaelzer> nacc: FYI bug 1709573 kills git ubutnu submit for me - but it is not a git-ubuntu bug, but my system or launchpadlib
[10:32] <rbasak> cpaelzer: for bug 1709573, does your password contain non-ASCII characters?
[11:10] <cpaelzer> rbasak: hmm I don't enter my pw anywhere
[11:11] <cpaelzer> rbasak: it is still fetching launchpad content - you men the LP PW then ?
[11:11] <cpaelzer> all asci chars I think
[11:12] <cpaelzer> in the worst case not on some weird codepages, but not that I'd use cyrillic or any german special like äüö
[11:13] <rbasak> cpaelzer: I wondered that, but if you're not actually entering anything, it seems far less likely.
[11:26] <cpaelzer> rbasak: it is fetching something, then wants to encode but the type is wrong
[11:26] <cpaelzer> I didn't check where it fetches it, but it might be the LP auth token itself
[11:27] <cpaelzer> I've seen the issue on ustriage sometimes, but could get around just by retrying
[11:27] <cpaelzer> on git ubuntu submit it seems persistent
[11:27] <cpaelzer> SO I thought it is about to file a proper bug
[11:29] <rbasak> Beret, dpb1: FYI, I've filed bug 1709603.
[12:44] <Beret> rbasak, sweet, thanks
[12:52] <cpaelzer> jamespage: dpdk 17.05.1-2 uploaded to Debian - if fortune is with us I can sync that later today
[12:52] <cpaelzer> jamespage: tonight/tomorrow you can then check to pass ovs 2.8 along that
[12:52] <cpaelzer> jamespage: I'll keep you updated
[13:29] <cpaelzer> jamespage: beisner: two questions
[13:29] <cpaelzer> a) do we (you) care about UCA on Desktops
[13:29] <cpaelzer> b) is virt-manager part of UCA (I don't think so, but you might have plans)
[13:29] <jamespage> cpaelzer: a) no; b) also no
[13:30] <cpaelzer> perfect - that makes me drop some patches
[13:30] <cpaelzer> just wanted to ensure there is no UCA back to Xenial
[13:30] <cpaelzer> (unitey patches)
[13:30] <cpaelzer> or rather Unity
[14:52] <xpistos> Hey guys. I am writing an automation script to install our software at work and want to make it more aesthetically pleasing. How can I combine the output of echo "the quick" and echo "brown fox" so it appears on one line with a return so it looks like "the quick brown fox"
[14:53] <mdeslaur> xpistos: echo -n
[14:53] <xpistos> so lets say the command is one line, the output is another and the new prompt is on the third
[14:54] <xpistos> mdeslaur: How do I get it to add a space after the first string?
[14:54] <xpistos> mdeslaur: without adding an additional " " string
[14:55] <ogra_> note, not all echos are the same ... (depends if you use /bin/echo or the shell builtin echo)
[14:56] <ogra_> (though i guess -n is supported on most of them ... -e is definitely not )
[14:57] <xpistos> ogra_: The -n works like this echo -n "the quick" && echo -n " " && echo "brown fox"
[14:57] <ogra_> yes ...
[14:57] <ogra_> -n should work mostly evereywhere ... but not all echos have the same options available
[14:59] <ogra_> (the echo built into dash only supports -n ... which the echo built into bash supports -n -e and -E ... and /bin/echo supports -n -e -E too)
[14:59] <ogra_> s/which/while/
[14:59] <TJ-> You could use printf(1), or the shell's built-in printf
[15:00] <ogra_> yeah
[15:00] <TJ-> printf(1) is part of core-utils so should be available (almost) everywhere, except possibly the initrd
[15:08] <nacc> mdeslaur: sorry for being dense, but just to follow-up on php7.0 -- you'll upload to -security and I should be able to request a pocket copy to -updates (well -proposed first) by the SRU team?
[15:12] <mdeslaur> nacc: I take care of it from here. Everything that goes to -security automatically gets copied to -updates
[15:13] <nacc> mdeslaur: ack, just wanted to close the loop on that, thanks!
[15:13] <mdeslaur> nacc: I will directly release them as security updates, no processing by the SRU team or wait necessary
[15:13] <mdeslaur> nacc: thanks!
[15:13] <nacc> mdeslaur: the orig tarball is exactly what is in debian, if you need it
[15:14] <mdeslaur> I downloaded it from upstream directly, I should have checked debian
[15:14] <mdeslaur> anyway, building now here: https://launchpad.net/~ubuntu-security-proposed/+archive/ubuntu/ppa/+packages
[15:14] <nacc> mdeslaur: yeah, i don't think there is any munging -- you should be able to check with uscan and see if the hashes match. I think you're fine.
[15:15] <mdeslaur> for some reason I thought debian was using 7.1
[15:15] <mdeslaur> oh well, next time
[15:16] <nacc> mdeslaur: they have both
[15:16] <nacc> mdeslaur: we have only one in each release :)
[16:28] <xpistos> Hey guys I got another question for you the following command shows the changes I want but doesn't write them to the file : sed -e "/server 3.centos.pool.ntp.org\ iburst/a\\server\ $ntpserverip iburst" < /etc/ntp.conf
[16:28] <xpistos> Any ideas how to get it to write those changes
[16:29] <dpb1> xpistos: well, sed -i modified files in place
[16:30] <dpb1> so like, sed -ie "/server 3.centos.pool.ntp.org\ iburst/a\\server\ $ntpserverip iburst" /etc/ntp.conf
[16:30] <dpb1> (untested)
[16:30] <xpistos> dpb1: LOL. I thougth I did that but I guess not
[16:30] <xpistos> Thank!
[17:15] <coreycb> jamespage: zero tempest smoke failures on pike b3 including the latest ceph
[17:26] <cpaelzer> jamespage: upload to Debian done, if it would build faster I could sync earlier
[17:27] <cpaelzer> jamespage: I'll check back before going to bed if it is reachable for the syncpackage call
[17:57] <nacc> rharper: is LP: #1671951 actually ready to sponsor?
[18:07] <rharper> nacc: it was, but not sure of the status now;  the debdiff likely needs refreshed against systemd in artful;  I'm not sure of xnox or anyone else attempted to re-work the issue upstream;
[18:08] <nacc> rharper: ok, can i unsub sponsors for now?
[18:08] <rharper> sure
[18:08] <nacc> rharper: thanks (i'm piloting)
[18:09] <ahasenack> nacc: wrt git workflow, there is a "no-change rebuild" change, which only changed d/changelog and bumped the version
[18:09] <ahasenack> nacc: http://pastebin.ubuntu.com/25278327/
[18:09] <nacc> ahasenack: yeah, so that'd be an empty commit (for the purpose of documenting it) with the same commit message as the changelog version
[18:09] <nacc> ahasenack: and then a contentful change to only d/changelog
[18:09] <ahasenack> nacc: should I record it as a normal "git commit debian/changelog -m changelog" commit?
[18:09] <ahasenack> hm
[18:10] <nacc> that would only show up in the deconstruct/
[18:10] <nacc> then in the logical you can drop both
[18:10] <ahasenack> ok, so git commit d/changelog -m "<same as d/changelog>"?
[18:10] <ahasenack> ah, empty commit
[18:10] <nacc> ahasenack: your way also works, but is a little less obvious
[18:10]  * ahasenack reads for a change
[18:10] <ahasenack> so empty commit with the d/changelog mesage,
[18:10] <nacc> ahasenack: i like my method, only because you have the same logical separation between what was changed (in this case nothing) and what documents the change (the changelog commit)
[18:11] <ahasenack> then d/changelog commit with -m changelog?
[18:11] <nacc> ahasenack: yeah
[18:11] <nacc> ahasenack: yep
[18:11] <nacc> ahasenack: i like that consistency, but as you noted, it's not really different than just a changelog commit
[18:11] <nacc> ahasenack: as long as the results match and is understandable, i think it's fine either way
[18:11] <ahasenack> but then we lose this bit of info
[18:11] <ahasenack> that it was a rebuild
[18:11] <nacc> ahasenack: right, at least in the git-log
[18:11] <ahasenack> so yours keeps that info
[18:11] <nacc> ahasenack: yeah, it's more self-documenting in my flow
[18:12] <nacc> ahasenack: and it's obvious to a reviewer that it should be dropped between deconstruct and logical
[18:28] <nacc> ahasenack: LP: #1677329, sponsors can be unsubbed, right?
[18:29]  * ahasenack checks
[18:29] <ahasenack> nacc: yes
[18:30] <nacc> ahasenack: thanks
[18:39] <ahasenack> man
[18:39] <ahasenack> bitten again by git's "let's comment empty commits by default during rebase" :/
[18:40] <nacc> ahasenack: are you doing an interact rebase? git rebase --abort
[18:40] <ahasenack> it's gone already
[18:40] <ahasenack> I noticed just now in the logical phase
[18:40] <ahasenack> "where is that empty commit that I was supposed to drop?"
[18:40] <nacc> ahasenack: well if they were empty, you don't need them in the logical phase?
[18:40] <nacc> ahasenack: they aren't logically part of the delta if they are empty
[18:41] <ahasenack> yeah, but it was dropped in the deconstruct one, in a last rebase I did to check I didn't miss anything
[18:41] <nacc> ah
[18:41] <nacc> ahasenack: you may want to alias `git-rebase` to `git-rebase --keep-empty` :)
[18:41] <ahasenack> yeah
[19:05] <ahasenack> cpaelzer: all libvirt tests passed in xenial, and trusty has an "always failed" one for armhf in nova
[20:10] <cpaelzer> ok ahasenack, thanks for the info so we are good on dep8 then
[20:11] <cpaelzer> ahasenack: since verification was also good it is down to staying in proposed a while
[20:36] <ahasenack> funny, interactive rebase can't reword empty commits on its own
[20:45] <ahasenack> nacc: I got this in debian's samba: https://anonscm.debian.org/cgit/pkg-samba/samba.git/commit/?id=e22e6b8bcae9b3fdf059d8878811c82b14585b65
[20:45] <ahasenack> nacc: seems to be debian specific
[20:45] <ahasenack> how do we deal with that? We don't care?
[20:46] <ahasenack> I've seen that the samba panic script has an "if ubuntu do this; if debian do that" check
[20:46] <ahasenack> I don't know what calls those bug-presubj and bug-script scripts, maybe debian's reportbug?
[20:49] <ahasenack> duh, it's what the commit msg says, "reportbug script" :)
[20:50] <ahasenack> but still, do we carry that?
[20:51]  * ahasenack sees that his ubuntu system has reportbug(1) and /usr/share/bugs full of stuff
[20:51] <sarnold> so you can report bugs to debian if you wish? :)
[21:16] <hehehe> hiu sarnold
[21:16] <hehehe> LOL
[21:16] <sarnold> hey hehehe
[21:16] <hehehe> whats the safest way to resize boot partition
[21:17] <hehehe> simply resize it and reboot?
[21:17] <hehehe> its full of old kernels
[21:17] <sarnold> I haven't resized a partition in 15 years
[21:17] <ahasenack> safest is to remove old kernerls
[21:17] <ahasenack> kernels*
[21:20] <hehehe> seems so
[21:24] <ahasenack> try apt autoremove
[21:24] <ahasenack> it should do the right thing. But double check
[21:40] <nacc> ahasenack: right, it seems like we either need to update that script to dtrt in ubuntu's samba
[21:41] <nacc> ahasenack: because we don't want to send ubuntu bugs to debian's bts
[22:10] <rose_> can anyone tell me if I need to reboot after doing apt-get upgrade
[22:11] <rose_> also what packages can I remove if i plan only using this xubuntu as console
[23:29] <nacc> rbasak: around (i hope not)?