[00:00] <stokachu> I should probably be more accommodating there
[00:00] <jjohnston> KeyError: KeyError('credentials',)
[00:01] <jjohnston> i'm using the api key from my maas user
[00:01] <jjohnston> does the specific maas user need to be in the credentials.yaml?
[00:02] <stokachu> oh maas 2.0 uses username/password right?
[00:02] <jjohnston> yes
[00:02] <stokachu> yea.. i need to add support for that
[00:02] <jjohnston> but presents an api key in the account settings
[00:02] <stokachu> ah ok maybe it does both
[00:03] <jjohnston> well maybe not..  it doesn't like my key no matter
[00:03] <stokachu> yea i think they've changed that
[00:03] <stokachu> unfortunately i've only tested on 1.9
[00:04] <jjohnston> should I be trying to put the user name in there and have a corresponding entry in the credentials.yaml for juju?
[00:05] <stokachu> jjohnston: i think im storing the credentials wrong in the credential.yaml for maas 2.0
[00:06] <stokachu> jjohnston: you could try adding the credentials manually for juju, then running juju bootstrap to have a controller ready
[00:06] <stokachu> conjure-up can make use of that as well
[00:06] <stokachu> i need to get a maas 2.0 up and going
[00:12] <stokachu> jjohnston: https://github.com/ubuntu/conjure-up/issues/38
[00:12] <stokachu> feel free to track that there
[00:13] <jjohnston> k will do, thanks!
[01:05] <gagagugu> sarnold, ?
[04:42] <shewless> stokachu: Is there a way to get the controller stuff on LXD but still use KVM for the compute nodes?
[07:16] <ShaRose> If anyone wants it, I basically edited the existing zfs-on-linux guide to installing 16.04 on zfs root so it's more.. automatic. With environment variables.
[07:17] <ShaRose> https://gist.github.com/ShaRose/7554a2f3ece4d5a119757d2dacab4a27 hopefully it works for everyone, etc etc don't blame me if it blows you stuff up
[07:17] <ShaRose> (also, I should probably rehost the bootstrap script..)
[07:21] <sarnold> ShaRose: please poke rlaager about your updates :)
[07:22] <ShaRose> sarnold dunno if it's 'bulletproof' enough though, mostly what I did was make things more automatic
[07:22] <ShaRose> for example, it probably would only work with a single interface
[08:27] <jelly> what is the optimal way to add 8bit latin1 en_US locale to a xenial 16.04 server?
[08:28] <sarnold> I think "sudo locale-gen en_US ; sudo update-locale" would do it (based on http://askubuntu.com/a/76106/33812 )
[08:35] <jelly> I'll see how that works on a fresh system. previously I messed with /var/lib/locales/supported.d/local but the directory doesn't even exist by default in 16.04 http://paste.ubuntu.com/16675192/
[08:36] <sarnold> I've got a /var/lib/locales/supported.d/ ..
[08:36] <sarnold> that also looks like it shold work
[08:38] <jelly> sarnold: did you install a language pack on the system where it exists?  dpkg -S /var/lib/locales/supported.d  says what?
[08:38] <sarnold> language-pack-en-base: /var/lib/locales/supported.d
[08:38] <jelly> right
[08:40] <jelly> coworker did just the core install, even "locales" was missing
[08:40] <sarnold> hehe
[08:41] <sarnold> I just used the standard server install image, nothing fancy
[08:41] <sarnold> but I htink that is something like three times the size of the "core" install
[08:42] <jelly> honestly I'm not sure why debian and ubuntu don't default to pregenerated locales-all, oh noes, 100MB disk space used
[08:50] <jelly> I could file a bug for Tivoli and tell them to "dammit, backup files even if their names are invalid UTF-8 bytecode"
[09:20] <curmet> Hello
[09:20] <curmet> I accidentally misconfigured my apache ubuntu , then it cant be restarted
[09:20] <curmet> it showed this error when restarted : "Job for apache2.service failed because the control process exited with error code. See "systemctl status apache2.service" and "journalctl -xe" for details."
[09:20] <curmet> Itried to purge and install , the problem persists. any idea?
[09:28] <hateball> curmet: does apache itself spit out any logs or does it actually die before that?
[09:29] <curmet> hateball: idk , it just doesnt start
[09:37] <Vertel> Greetings. I'm setting up a personal 16.04 server and am wondering if there's a solid guide to securing it. I want to avoid being one of those servers on the Internet that gets taken over and used to spam China. :P
[09:39] <Vertel> I do know all the basics; don't open unnecessary ports, only install services I need, public-key SSH required, but I'm not quite sure I'm paranoid enough yet.
[09:40] <ikonia> curmet: look at the log files as I told you in #ubuntu and fix why it's broken
[09:40] <ikonia> randomly re-installing won't help
[10:00] <curmet> hateball, any idea?
[10:16] <hateball> curmet: Yes, I gave you a suggestion and it seems that ikonia has as well
[10:19] <curmet> hateball , logged out for awhile lol,which suggestion?
[10:25] <hateball> curmet: To check for apache logs
[10:28] <curmet> "[Wed May 25 05:15:37.211997 2016] [mpm_event:notice] [pid 6039:tid 3074554496] AH00489: Apache/2.4.12 (Ubuntu) configured -- resuming normal operations
[10:28] <curmet> [Wed May 25 05:15:37.214442 2016] [core:notice] [pid 6039:tid 3074554496] AH00094: Command line: '/usr/sbin/apache2'
[10:28] <curmet> [Wed May 25 05:16:29.664671 2016] [mpm_event:notice] [pid 6039:tid 3074554496] AH00491: caught SIGTERM, shutting down
[10:28] <curmet> "
[10:28] <curmet> hateball , that is
[10:35] <pirx_> hello! i just added a new service in my server, by adding a script to /etc/init.d/ and doing "update-rc.d service defaults"
[10:36] <pirx_> the service starts fine, but none of the outputs (the echo's) in the init-script are printed to screen anymore
[10:36] <pirx_> so it seems to work, but its all very clandesine:)
[10:39] <cpaelzer> pirx_: is that already with systemd , so 15.04 or later I think ?
[10:41] <pirx_> ah, forgot to say, this is on 16.04, first time i am installing stuff on a 16.04
[10:41] <cpaelzer> pirx_: ok that makes sense - the old init scripts are only executed by a systemd wrapper
[10:41] <pirx_> all other servers are still 14.04
[10:42] <cpaelzer> pirx_: it is intentional that they don't litter the screen with messages
[10:42] <cpaelzer> pirx_: you should be able to see it at the end of "service yourservice status"
[10:42] <cpaelzer> pirx_: also in the journal it would show up
[10:42] <pirx_> so i should just trust that it started ok if it didnt print anything? :)
[10:43] <cpaelzer> pirx_: well how do you start stop it - with the "service" wrapper or with systemctl ?
[10:43] <cpaelzer> pirx_: or do you call it directly like /etc/init.d/foo ?
[10:43] <pirx_> but it also seems that my init-script doesnt work 100% with "service --status-all"
[10:44] <pirx_> cpaelzer: with service wrapper
[10:44] <pirx_> --status-all shows [ - ] even though the service is started
[10:44] <cpaelzer> pirx_: in that case you should expect it to work if you don't see - an error would show up and in general "status" is your way to go
[10:45] <cpaelzer> pirx_: for the proper integration into status all and so much more it should (these days) best be a proper systemd service/unit
[10:45] <pirx_> oh crap, just realized that this script lacks a status) argument :)
[10:45] <cpaelzer> I don't have a link handy but there are good guides for such a transition
[10:48] <cpaelzer> hateball: you have often a great bucket of great links to refer to - is there a good entry point for sysv to systemd transition for pirx_?
[10:48] <cpaelzer> I only find upstart to systemd at https://wiki.ubuntu.com/SystemdForUpstartUsers
[10:51] <stokachu> shewless: yes use openstack base  selection (the first option)
[10:55] <hateball> cpaelzer: afraid not, I havent really been dabbling anything with systemd yet myself :o
[10:56] <hateball> currently running 12.04 and 14.04 on all servers, so :p
[10:57] <cpaelzer> hateball: ok, was just my perception that there is always a handy link from you :-)
[10:57] <hateball> google-fu goes a long way :p
[10:58] <hateball> But sadly not read up the subject at hand
[10:59] <cpaelzer> sadly I don't remember where I started, butI quickly checked and sorted out a few
[10:59] <cpaelzer> http://unix.stackexchange.com/questions/233468/how-does-systemd-use-etc-init-d-scripts
[10:59] <cpaelzer> https://fedoramagazine.org/systemd-converting-sysvinit-scripts/
[11:00] <cpaelzer> but especially (really) understanding what this does is important => https://www.freedesktop.org/software/systemd/man/systemd-sysv-generator.html
[11:00] <cpaelzer> pirx_: I hope this gives you a good start
[11:02] <cpaelzer> deadnull_: nice, it is almost like a dance with the underscore :-)
[11:04] <pirx_> cpaelzer: thanks a lot! i'll have to go through this now it seems:)
[11:04] <deadnull_> cpaelzer lol :)
[11:09] <jonah> Hi I wondered if anyone can help at all. I have an ubuntu 14.04 backup server that just uses samba and ftp for some backups. It only has a 10/100 built in lan card as it's pretty old. So I bought a PCI-e Lan card which is gigabit. The problem is it doesn't seem to "just work" haha, it was cheap! So does anyone know how I can detect what the driver is and how to get it up and running so I can benefit from blazing speeds!!!
[11:09] <jonah> Thanks if anyone can help at all
[11:11] <ikonia> jonah: find out what chipset the card is
[11:11] <cpaelzer> jonah: I'd start with lspci - identify which card it is, then once you have the slot  and such rnu lspci again but with -vvv and -s slotspec
[11:11] <cpaelzer> jonah: that will give you a start info as what the card is really seen - with that you can start looking for support
[11:11] <jonah> ikonia: well that's just it, I don't know. maybe it was a bit TOO cheap!
[11:11] <jonah> I tried lspci, but can't see it or don't recognise which entry it could be
[11:12] <jonah> I can see my 10/100 intel one there
[11:12] <ikonia> jonah: so that sounds like it's dead
[11:12] <jonah> but no mention of any sort of ethernet/lan/speeds
[11:12] <ikonia> it may not show a speed, it's just a description of the device
[11:12] <ikonia> if you see no device that could be it, then it's probably a broken card
[11:12] <jonah> ikonia: what do you mean by dead, just broken? The light comes on when you insert an ethernet plug...
[11:12] <ikonia> take it back and buy a quality supported card
[11:12] <cpaelzer> jonah: you might "lspci -vvv | pastebinit" so we can look at it
[11:12] <ikonia> jonah: thats just a link light
[11:13] <cpaelzer> but I agree to ikonia that it is likely just dead
[11:14] <jonah> Hi I've pastebinned it here: http://pastebin.com/phV8wQn4
[11:15] <cpaelzer> jonah: yeah just not showing up likely means dead
[11:16] <cpaelzer> jonah: for the sake of trying to ressurrect you might save the lspci to a file, move it to another slot - do the same, remove it do the same
[11:16] <cpaelzer> jonah: and then compare the lspci output files
[11:16] <cpaelzer> jonah: after all you could also have a disabled pci slot or so
[11:16] <cpaelzer> jonah: also watch dmesg and such
[11:16] <cpaelzer> jonah: or shortcut to "buy a good one"
[11:16] <jonah> cpaelzer: ah yeah that's good idea
[11:17] <jonah> ok yes I'll try another slot just in case and then just buy a decent one! haha thanks for the help
[11:17] <cpaelzer> jonah: I have seen weird boards where enabling the primary slot with x16 disables the other slot and such
[11:17] <cpaelzer> jonah: good luck
[11:21] <jonah> cpaelzer: thanks
[11:22] <jonah> also while I'm in here, is there a way to get the backup server to spindown/sleep or even turn itself off and on as it is needed? It only really backs up at night, so it would be great if there was a way to power it down during the day to save it wearing out and then it just come on when the samba, ssh/ftp backups start...
[11:30] <rbasak> jonah: you can do it but I don't recall the details. pm-suspend on older systems, systemd does something to replace it since Vivid. Most systems can be set to wake on a clock but I don't know the details - either from the BIOS or by adjusting the RTC from the running system. You'll probably need to install some packages as it's not a common server use case.
[11:32] <jonah> rbasak: ok thanks I'll try and have a google around in that case
[11:33] <John[Lisbeth]> btw if whoever helped me ealier is still listening, setterm worked for the htop machine
[11:33] <John[Lisbeth]> tyvm
[11:35] <rbasak> teward: http://i.imgur.com/3aUIbfv.jpg
[11:35] <hateball> jonah: do you have other devices on 24/7, perhaps a router? You could schedule WOL for the server, then have it shut off once done
[11:39] <jonah> hateball: yeah I have the router on all the time and another server which it backs up
[11:57] <hateball> jonah: well then you could look into scheduling WOL
[11:57] <hateball> my router has support for it, that's pretty neat
[12:06] <jonah> hateball: ah ok so the router wakes up the bios wol option? that is cool, how do you then get the server to shut itself down again when it's not doing anything too?
[12:06] <hateball> jonah: just add a shutdown command after your backup command completes successfully
[12:07] <teward> rbasak: LOL
[12:07] <teward> rbasak: that made me smile xD
[12:07] <jonah> hateball: that's cool I'll have a look thanks again
[12:12] <jonah> hateball: ah I just thought of something else. If I have 3 or 4 backups that run such as weekly, nightly and another from a different server etc and I add shutdown commands to them they might shutdown each other's backups half way through...?
[12:13] <teward> jonah: then schedule maintenance windows for shutdowns, etc.
[12:13] <teward> times where the backups will have already completed and such
[12:13] <teward> and then schedule the shutdown/reboot then
[12:15] <jonah> teward: this is where it gets a bit confusing though. say there is an incremental backup that runs, but one night no files have changed so the backup only takes 5 mins but another night loads of new files or changes have been made on the server and the backup takes 2 days. This is why I wondered if there is a way to only shutdown if nothing is going on rather than at a set time
[12:16] <hateball> jonah: do you run both full and incrementals at the same time?
[12:16] <rbasak> jonah: get each backup script to hold a lock. Attempt to grab the lock exclusively at the end, and if successful then shut down.
[12:17] <rbasak> Careful about the race condition though that a shutdown happens just before a new backup is due to start.
[12:17] <jonah> hateball: yeah they can overlap. As the incremental is nightly and just depends, and the full is weekly. but the full takes a couple days to complete and as I say the incremental depends. then to throw even more into the mix is another server doing a midweek backup too!
[12:18] <hateball> sounds like there's not much time for the server to be shutdown at all then?
[12:18] <teward> jonah: I would not be doing shutdowns automatically then if you're stating this
[12:19] <teward> jonah: instead, you need to schedule time in *all* the backup schedules to do a shutdown/restart, by halting automated backups at that time
[12:19] <jonah> haha that's true! I just thought there may be a day here or there I could save power and prelong the life of the old box a bit
[12:20] <teward> jonah: with the backup schedule as you saying running "whenever" you're going to have to halt backups globally to achieve your goal
[12:20] <teward> the only way to achieve that is to do backups by the backup server being a 'fetcher'
[12:20] <teward> or, stop all backups
[12:20] <rbasak> For spinning disks you may actually be shortening their life by spinning them down.
[12:20] <teward> ^ that too
[12:21] <teward> rbasak: BTW, can you do a preliminary review of a debdiff for the nginx merge?  At least, to make sure I didn't miss anything blaringly obvious.  I am behind on getting the merge done thanks to network/hardware/firewall replacement
[12:21] <teward> with no backup for this one :
[12:21] <teward> :/ *
[12:22] <teward> rbasak: https://launchpadlibrarian.net/259562731/preliminary-yakkety-merge.v5.debdiff if you have time, if not then don't worry about it
[12:22] <teward> it'll be underoing testing starting Monday in my VMs.
[12:22] <jonah> well they're pretty old disks hard disks so maybe they'll live longing just leaving them on all the time?
[13:44] <jatin30> I am building LAMP from https://help.ubuntu.com/community/ApacheMySQLPHP and I am stuck at point 3 in virtual hosts [19:11] <jatin30> Change the DocumentRoot to point to the new location. For example, /home/user/public_html/ [19:11] <jatin30> how to do this?
[13:47] <Pici> jatin30: what do you mean?  Are you looking to setup multiple virtual hosts?
[13:48] <josepht> jatin30: in your .conf file from point 2 edit DocumentRoot to point to where you will be serving http from
[13:48] <Pici> personally I usually leave it pointing to somewhere in /var/www/
[13:49] <jatin30> josepht: changed "DocumentRoot /var/www/html " to "	DocumentRoot /home/user/public_html/"
[13:50] <jatin30> I am new here , so please excuse me.
[13:50] <jatin30> josepht, will it work fine what I did
[13:51] <josepht> jatin30: no worries, It will work only if /home/user/public_html is a real directory
[13:51] <josepht> jatin30: most likely your username is not "user" :)
[13:52] <jatin30> Ok! i cant find public_html in the path. there is a public tho
[13:52] <josepht> jatin30: I'm with Pici though, I'd just leave it as /var/www/html, make sure the directory exists and use that everywhere you see /home/user/public_html in the wiki
[13:52] <jatin30> in point 3 it was given to change it to /home/user/public_html
[13:53] <jatin30> I cant find var only
[13:55] <josepht> jatin30: it assumes you will create the directory if it doesn't exist
[13:55] <josepht> sudo mkdir -p /var/www/html
[13:55] <jatin30> ok Josepht, can I direct message you?
[13:56] <josepht> jatin30: if you need to, it's fine here as well in case someone else finds it useful
[14:40] <huwjr> heya, trying to run php7 apache module, is it normal that there is no php7.conf or php7.load ? I can create these, but just wondering if I am missing something blatant
[14:48] <sdeziel> huwjr: the files are named php7.0.conf and php7.0.load
[14:48] <sdeziel> http://packages.ubuntu.com/xenial/amd64/libapache2-mod-php7.0/filelist
[14:49] <huwjr> nope :/ weren’t create..
[14:49] <huwjr> +d*
[14:49] <sdeziel> is this the package you (tried to) installed?
[14:49] <huwjr> yep..
[14:50] <huwjr> *double checks*
[14:50] <huwjr> yep. that’s it :/
[14:51] <huwjr> all the other files are there
[14:51] <huwjr> but no load or conf
[14:51] <huwjr> how bizarre
[14:51] <sdeziel> huwjr: stupid question but are you checking under mods-available/ ?
[14:52] <sdeziel> dunno if those are auto enabled during the installation
[14:52] <huwjr> i’ve checked both
[14:53] <huwjr> the install was done with an ansible playbook, but there were no errors and the .so and everything in /usr/lib /usr/share is instsalled
[14:53] <huwjr> i’m properly confused
[14:54] <sdeziel> huwjr: what's the dpkg status of that package?
[14:54] <coreycb> jamespage, nova 1:2014.1.5-0ubuntu1.5~cloud0 is ready to promote to icehouse-proposed when you have a chance
[14:55] <huwjr> libapache2-mod-php7.0              7.0.4-7ubuntu2                      amd64        server-side, HTML-embedded scripting language (Apache 2 module
[14:56] <sdeziel> huwjr: the first field is missing
[14:57] <huwjr> ii
[14:57] <huwjr> soz
[14:57] <sdeziel> huwjr: OK, looks good. Now I'm confused too ;)
[14:57] <huwjr> yeah...
[14:57] <coreycb> jamespage, also qemu 1:2.2+dfsg-5expubuntu9.7~cloud3 is ready to promote to kilo-updates
[14:57] <huwjr> i don’t know if perhaps this is related BUT
[14:58] <huwjr> the OS was installed with a preseed, and it automagically put /home on a separate partition… despite having specified atomic
[14:58] <huwjr> i reinstalled 3 times with the SAME preseed, and every time it has not done that again.
[14:58] <huwjr> which is batsh**
[14:59] <sdeziel> huwjr: I doubt that's related. I'd try a manual (re)installation of the package and see what's going on
[14:59] <huwjr> mm
[14:59] <huwjr> ta
[14:59] <huwjr> glad i’m not barking up the wrong tree anyway
[15:03] <huwjr> sdeziel: would you do a --reinstall
[15:03] <huwjr> or remove, then install
[15:03] <sdeziel> huwjr: I think both would do the same
[15:06] <huwjr> apt-get --purge remove libapache2-mod-php7.0 says it’s going to install php7.0-fpm :/
[15:06] <huwjr> i repeat INSTALL
[15:06] <huwjr> why?
[15:08] <sdeziel> huwjr: I suspect some other package wants a php provider to exist so when you remove the apache implementation it wants to switch to the FPM one to keep the dependency
[15:08] <huwjr> ah i see
[15:08] <huwjr> fair enough
[15:22] <huwjr> reinstall gets: dpkg: error processing package libapache2-mod-php7.0 (--configure): subprocess installed post-installation script returned error exit status 1: Errors were encountered while processing: libapache2-mod-php7.0: E: Sub-process /usr/bin/dpkg returned an error code (1)
[15:25] <sdeziel> huwjr: you can try to add "set -x" to /var/lib/dpkg/info/libapache2-mod-php7.0.postinst and redo the configuration
[15:30] <huwjr> as in above set -e or instead of
[15:30] <huwjr> and thanks *)
[15:30] <nacc> huwjr: in addition to
[15:30] <nacc> so above/below
[15:31] <sdeziel> thanks nacc :)
[15:31] <huwjr> thanks ..
[15:32] <nacc> sdeziel: i believe you're right on the reason php7.0-fpm was installed ... (an alternative solution would be to remove the reverse-deps, or to pick a different provider manually)
[15:32] <nacc> huwjr: would also be curious what `dpkg -L` says for that package, if you haven't uninstalled it yet
[15:33] <huwjr> E: Internal Error, No file name for libapache2-mod-php7.0:amd64
[15:33] <nacc> huwjr: also, stock 16.04? not using any PPA?
[15:33] <huwjr> no, completely stock
[15:33] <nacc> huwjr: ok, just checking
[15:33] <nacc> huwjr: what printed that 'Internal Error'? adding the set -x ?
[15:33] <huwjr> yeah
[15:34] <sdeziel> huwjr: Xenial received a PHP update (yesterday) so you should have 7.0.4-7ubuntu2.1 available for install  (you currently have 7.0.4-7ubuntu2)
[15:34] <huwjr> http://paste.ubuntu.com/16684570/
[15:35] <sdeziel> doubt that it would help with the problem at hand but your version is known vulnerable...
[15:35] <huwjr> i read the notes, didn’t seem to effect
[15:35] <huwjr> k
[15:35] <nacc> huwjr: the other thing to try, not saying this will work, is removing 'libapache2-mod-php', rather than the version specific one
[15:35] <nacc> huwjr: intersting, so dpkg clearly things it the mod .conf and .load are in the file
[15:36] <nacc> *the pacakge
[15:36] <huwjr> hehe
[15:37] <nacc> *thinks
[15:37] <nacc> gah
[15:37] <nacc> :)
[15:37] <huwjr> i’ve exhausted a lot of my knowledge, but this is indeed curious
[15:37] <nacc> huwjr: ok, still reading scrollback, what's the use-case?
[15:37] <nacc> huwjr: as in, how did this happen? :)
[15:37] <huwjr> well everything installed fine (AFAIK), dpkg state ii
[15:38] <huwjr> host was deployed with ansible, so it’s possible something was missed without stderr
[15:38] <huwjr> assuming an error was on stdout instead of stderr
[15:38] <nacc> huwjr: any chance you could pastebin /var/log/dpkg.log ?
[15:38] <huwjr> and i only noticed when php wasn’t running on a website i just set up, and couldn’t find the .load or .conf
[15:39] <rbasak> huwjr: if done with ansible, does that mean that you should be able to reproduce? It might be easier to verify if you can reproduce then reduce it to a minimal manual case.
[15:39] <huwjr> 2016-05-12 17:29:44 status half-installed php-common:all 1:35ubuntu6
[15:40] <huwjr> built this a while ago, but had some issues with partitions hense only looking deeper today
[15:41] <nacc> huwjr: ah, that might be the root-cause? like you mentioned, i mean
[15:41] <nacc> rbasak: good point; huwjr if you wanted to share the playbook, maybe it's something obvious
[15:42] <coreycb> jamespage: for the ironic newton failure, the tests pass locally but /etc/hosts isn't found when it's run in ci for some reason. and it only fails on yakkety.  not sure why..
[15:42] <huwjr> you mean the possibility of stderr to stdout?
[15:42] <huwjr> would be a lot of abstracting from the playbook, but will keep it in mind if i don’t get anywhere soon!
[15:44] <nacc> huwjr: or something isn't doing proper error checking. The dpkg.log should help figure out what initially 'broke' things
[15:46] <huwjr> you want the full log, or just php stuff
[15:52] <nacc> huwjr: i guess if you know what to look for, you're welcome to parse through it
[15:52] <nacc> i realize now the log might be quite large :)
[15:53] <nacc> huwjr: i would mostly be interested in the php stuff, and especially any errors reported
[15:53] <huwjr> http://paste.ubuntu.com/16685152/
[15:54] <nacc> huwjr: ok yeah the 'half-installed' bits are normal
[15:54] <nacc> iirc, it goes into that state during installation itself
[15:54] <nacc> and then transitions to 'installed'
[15:54] <huwjr> yeah noted that now :)
[15:55] <sdeziel> 2016-05-13 00:27:09 status triggers-pending libapache2-mod-php7.0:amd64 7.0.4-7ubuntu2
[15:55] <sdeziel> I am not sure what that means ^
[15:56] <sdeziel> maybe because of the new curl module to enable?
[15:56] <nacc> i think that's ok, because it's followed eventually with 'trigproc trigproc libapache2-mod-php7.0:amd64 7.0.4-7ubuntu2 <none>'
[15:56] <nacc> but i'm not sure, tbh
[15:56] <nacc> err, one less trigrpoc
[15:57] <nacc> sdeziel: yeah, so i think that's the way dependencies forcing reloads work, it sets off a trigger, dpkg tracks what triggers still need to fire, and then processes the triggers
[15:57] <sdeziel> nacc: sounds logical
[15:58] <nacc> the weird part is the end of that log, though
[15:58] <huwjr> the end?
[15:58] <huwjr> the end is me madly trying to get it reinstalled - probs :p
[15:58] <nacc> trigproc -> half-configure -> installed -> upgrade (to the same version?) -> half-configured -> half-installed
[15:58] <huwjr> oh.
[15:58] <nacc> and it seems to still be in that half-configured state by the last line
[15:59] <nacc> huwjr: so it might be your manual intervention, not sure
[15:59] <huwjr> yeah and dpkg was state: ii to start with
[15:59] <sdeziel> could the upgrade be a --reinstall?
[15:59] <nacc> huwjr: is this reproducible
[15:59] <huwjr> only AFTER reinstall did it fail with iF
[15:59] <nacc> sdeziel: ah it could be
[15:59] <huwjr> i did -- reinstall then it entered iF
[15:59] <huwjr> but before was fine :)
[16:00] <nacc> let me spin up a container just to see
[16:00] <nacc> huwjr: sorry, maybe got missed; is this reproducible in multiple ansible deployments?
[16:00] <huwjr> i’ve not retried… will do now :)
[16:00] <huwjr> i suspect it will all go nicely *hehe*
[16:01] <nacc> huwjr: yeah, i'm tempted to say glitch in the matrix, because if this was more commonly happening i think i'd be seeing a lot mor bugs
[16:02] <huwjr> oh yeah definitely
[16:02] <huwjr> the separate partition on /home also spookyk
[16:02] <nacc> i do get a 'Warning: Could not load Apache 2.4 maintainer script helper.' during the --reinstall, but it succeeded in my container
[16:02] <nacc> this is with 2.1, fwiw
[16:02] <nacc> 7.0.4-7ubuntu2.1
[16:04] <huwjr> ok so
[16:08] <huwjr> sorry - got distracted
[16:09] <nacc> huwjr: np
[16:09] <huwjr> so did an apt update to see what would happen
[16:09] <huwjr> http://paste.ubuntu.com/16685572/
[16:09] <huwjr> that at least seem more promising, no?
[16:09] <nacc> huwjr: so if you do `apt-get -f install` right now, what happens (feel free to pastebin)?
[16:10] <nacc> i think right now, dpkg is unhappy that libapache2-mod-php7.0 hasn't finished configuring
[16:10] <nacc> huwjr: you might also try `apt-get -f install libapache2-mod-php7.0`
[16:10] <huwjr> http://paste.ubuntu.com/16685685/
[16:11] <nacc> huwjr: ok, try just that last one (so we can remove php7.0 from the noise)-- i assume it will error out the same, but nt sure
[16:11] <huwjr> do i need to clear the .postinst ?
[16:12] <huwjr> http://paste.ubuntu.com/16685729/
[16:12] <nacc> huwjr: you shouldn't ever have to manually muck with those files, afaik -- oh but you altered it, you mean? um, yeah, i guess reset it, but it shouldn' tmatter (the error codes are all the same, it's justmore verbose with 'set -x')
[16:12] <huwjr> i backed out my set x
[16:12] <nacc> huwjr: are you able to purge out libapache2-mod-php7.0?
[16:12] <huwjr> before this lot
[16:13] <sdeziel> nacc: I'm the one who suggested huwjr to add "set -x" to the postinst
[16:13] <nacc> sdeziel: ack, i recall -- did we get that output?
[16:13] <sdeziel> no
[16:14] <nacc> :)
[16:14] <huwjr> yeah the output was … E: Internal Error, No file name for libapache2-mod-php7.0:amd64
[16:15] <huwjr> anywho, purged - but it required me to /install/ FPM which was never installed, i did so
[16:15] <huwjr> then reinstalled libapache2-mod……. and it worked
[16:15] <huwjr> so it seems there’s a depend on fpm?
[16:16] <huwjr> or, it was just in a mess, and i could only purge by installing fpm (there’s probably a way, I’m dumb!)
[16:16] <huwjr> the latter i guess?
[16:16] <nacc> huwjr: right that's what sdeziel suggested earlier and i agreed (in theory at elast), that there's something else (probably php7.0 itself) requireing a php provider to be installed
[16:16] <nacc> either -fpm or apache or cgi
[16:16] <sdeziel> "No file name for libapache2-mod-php7.0" could this be because of the 2.1 update?
[16:17] <nacc> and since you removed apache, it needed to install the next int he alternatives list
[16:17] <nacc> sdeziel: it *could* be, but shouldn't be, i don't think :)
[16:17] <nacc> huwjr: that was the only output?
[16:19] <huwjr> soz
[16:19] <huwjr> many many things on
[16:19] <huwjr> the only output of what sorry?:)
[16:19] <huwjr> with set x? the only output was above..
[16:24] <nacc> huwjr: yeah, 'set -x'
[16:25] <nacc> huwjr: from running that postinst script manually basically
[16:25] <nacc> huwjr: strange
[16:25] <nacc> huwjr: i'm still not 100% sure how you got into your original state, but it seems like you are out of it now?
[16:25] <nacc> huwjr: also, did you reproduce it in a fresh run of the playbook?
[16:26]  * huwjr scrolls up
[16:26] <huwjr> yeah that was the only errror
[16:26] <huwjr> 1 not fully installed or removed
[16:26] <nacc> ok
[16:26] <huwjr> well this
[16:26] <huwjr> 0 to upgrade, 0 to newly install, 1 reinstalled, 0 to remove and 19 not to upgrade.
[16:26] <huwjr> 1 not fully installed or removed.
[16:26] <huwjr> After this operation, 0 B of additional disk space will be used.
[16:26] <huwjr> :)
[16:26] <huwjr> but that’s kind of obvious..
[16:27] <nacc> so the only other thing i could think of is if somehow the archive got corrupted, but i would think the various checks would have noticed
[16:27] <nacc> or a file didn't get fully downloaded or something
[16:28] <huwjr> out of the water now, but going to have to rebuild this - or another vm - and just check the process with my plays
[16:28] <huwjr> not the end of the world
[16:28] <nacc> huwjr: ack, let us know if it happens again, as i'd like to reproduce it if so
[16:28] <huwjr> sure! well i appreciate the help and suggestions
[16:35] <bc2946088> I'm running juju-ceph and it's up and working fine, however, I have a failing SMART status on one of the drives.  Is there a prefered method of replacement?  Should I just let it ride and replace it when ceph picks up the warning, or is it better to replace the drive in ceph before failure?  It seems logical to replace the drive before, but just curious is ceph automagically understands whats going on better when the drive fails and is repla
[16:35] <bc2946088> ced
[16:44] <John[Lisbeth]`> I need to make apt auto-approve any package changes
[16:44] <John[Lisbeth]`> apt-get install foo --auto-approve
[16:44] <nacc> John[Lisbeth]`: you mean say "yes"? try the -y flag
[16:44] <John[Lisbeth]`> ty
[16:45] <sdeziel> it's the first time I hear about "approve"
[16:45] <sdeziel> oh, nvm
[16:47] <John[Lisbeth]`> pseudocode
[16:49] <sdeziel> yeah, I somehow interpret this as a feature similar to apt-mark
[16:50] <nacc> sdeziel: yeah, i realize that is a reasonable interpretation now too :)
[18:52] <jonah> Hi does anyone know how to add this rule to iptables on ubuntu: http://ubuntuforums.org/showthread.php?t=2117823&p=12548787#post12548787
[18:52] <jonah> Herman right at the bottom of this thread suggests this great little iptables rule
[18:52] <jonah> which I'd like to use
[18:53] <jonah> but the way it has a # Comment above the line suggests you don't just run that command but insert it into a text file somewhere?
[18:53] <jonah> or do you literally just type  iptables -I INPUT -p TCP -m state --state NEW -m limit --limit 30/minute --limit-burst 5 -j ACCEPT
[18:53] <jonah> and that adds it in correctly? or am I supposed to somehow use ufw?
[19:50] <jjrabbit443> hello
[19:50] <jjrabbit443> can someone confirm this command will unblock port 443 on my firewall?
[19:50] <jjrabbit443> "sudo iptables -A INPUT -p tcp --dport 443 -j ACCEPT"
[19:51] <jjrabbit443> is that correct syntax?
[19:52] <patdk-wk> hmm, cannot guarrentee it will work
[19:52] <patdk-wk> syntax is correct
[19:52] <patdk-wk> but that rule highly depends on *what else* is in your firewall
[19:52] <patdk-wk> and it won't stick between reboots
[20:01] <jjrabbit443> patdk-wk: how do i make it stick?
[20:01] <jjrabbit443> do i open firewall config file?
[20:02] <sarnold> how do you manage your firewall now?
[20:03] <jjrabbit443> command line i guess
[20:08] <jjrabbit443> is there a paramater i can add to that command to ameke it permanent?
[20:12] <sdeziel> jjrabbit443: you can use the iptables-persistent package to save/load the rules for you
[20:13] <jjrabbit443> for anyone wondering you can execute "iptables-save"
[20:13] <jjrabbit443> sdeziel: thanks i found the command
[20:13] <sdeziel> jjrabbit443: despite the name iptables-save will only output the currently loaded rules to the console. It won't persist in any way
[20:21] <patdk-wk> the question is, what are you using now, ufw?
[20:21] <patdk-wk> raw
[20:22] <patdk-wk> cause by default nothing persists, except if you installed ufw
[20:22] <patdk-wk> or used the iptables-persistent for raw rules
[20:22] <patdk-wk> don't think firewalld is in ubuntu yet, but not sure
[20:42] <jjrabbit443> sdeziel: you are right that didn't work
[20:43] <jjrabbit443> ufw seems much simpler
[20:43] <sdeziel> indeed
[20:43] <jjrabbit443> and rules added using that are automatically saved correct?
[20:43] <sdeziel> probably why u stands for uncomplicated ;)
[20:44] <sdeziel> jjrabbit443: I would believe so but I don't remember
[20:44] <jdstrand> yes they are
[20:44] <jjrabbit443> psh and here i was wasting my time with iptables
[20:44] <jjrabbit443> thanks guys
[20:48] <jjrabbit443> jdstrand: confirmed
[20:48] <jjrabbit443> beauteefull
[20:48] <jdstrand> cool
[21:28] <arooni> my ubuntu 14.04 server has rebooted several times without my blessing.  how can i track down why?  i've already investigated /var/log/dmesg /var/log/syslog /var/log/kern.log and didnt see anything notable
[21:36] <John[Lisbeth]> can you do dual monitors with just tty windows?
[22:16] <keithzg> stokachu: Windows has Windows Defender included, but only in non-server variants (and not on XP, or at least not in any worthwhile fashion). Sever 2016 will finally ship with Windows Defender, but we have some Windows Server boxes at work for testing because our clients keep running our desktop software on Windows Server for some reason, and I need an antivirus program I can administer from Linux but runs on Windows Server.
[22:16]  * keithzg is back on IRC after nearly 24 hours idle, heh
[22:22] <sarnold> arooni: maybe configure a serial console? perhaps your kernel paniced in interrupt context and thus didn't flush logs to disk, and reboots..