[01:32] <xyz> hi, is mcelog not available on 18.04 anymore?
[01:35] <sdeziel> xyz: see https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=889741
[01:40] <xyz> thanks so I need to use rasdaemon now?
[02:28] <sarnold> xyz: thanks for the pointer :)
[02:34] <xyz> no problem sarnold let me know if you figure out how to use that
[02:34] <sarnold> xyz: on my bionic laptop I did apt-get install rasdaemon ; systemctl start rasdaemon ; ras-mc-ctl --summary
[02:35] <sarnold> no ecc on this thing so ras-mc-ctl --status didn't work (at least I think that's why it didn't work..)
[02:36] <xyz> hmm I see, for me --status just says "ras-mc-ctl: drivers are loaded."
[02:36] <xyz> so I guess it's working
[02:37] <sarnold> it was a bit anti-climactic.. but I guess that's best, it's unfun to see machine check exceptions :)
[04:53] <cpaelzer> good morning
[06:03] <lordievader> Good morning
[11:08] <hwpplayer1> could you please suggest me an article for ubuntu 18.04
[11:08] <hwpplayer1> i read official article
[11:58] <Chryzo> Good morning, i have my DNS entries in the ethernet configuration: /etc/network/interfaces with dns-nameservers  but they are not getting picked up by resolv.conf. any idea why ?
[12:30] <hwpplayer1> could you please suggest me an article for ubuntu 18.04 i read official article
[12:33] <tomreyn> hwpplayer1: what kind of article are you looking for?
[12:48] <hwpplayer1> tomreyn : something tells more detailed about new missions related cloud computing storage artificial intelligence especially machine learning with the view of canonical roadmap and critisitive
[12:57] <tomreyn> hwpplayer1: i only know about https://www.markshuttleworth.com/archives/1518 regarding canonical's strategy.
[13:03] <xnox> tomreyn, hwpplayer1 - i'd recommend https://blog.ubuntu.com/ specifically, e.g. https://blog.ubuntu.com/2018/04/26/ubuntu-18-04-lts-optimised-for-security-multi-cloud-containers-ai & https://blog.ubuntu.com/2018/04/27/whats-new-in-ubuntu-18-04-and-openstack-queens
[13:04] <xnox> Chryzo, which ubuntu release is this on?
[13:04] <xnox> Chryzo, and by not picked up... do you see them in $ systemd-resolve --status ?
[14:17] <sdeziel> teward: I wrongly marked LP: #1578344 as fix release and can't undo this to triaged, sorry about that and could you please fix it for me?
[14:20] <Buoy172> I don't know if this is the right channel to ask but...
[14:21] <Buoy172> I have this webserver that I can login to
[14:21] <Buoy172> and I can do 'ls' and see "bin boot data dev" ... etc
[14:21] <Buoy172> where's the website located at?
[14:21] <Buoy172> where's index.php?
[14:22] <sdeziel> Buoy172: usually somewhere under /var/www
[14:23] <Chryzo> xnox, ubuntu 16 and 14 lts. I forcefully modified the header of the file for now. As for: systemd-resolve --status, my system does not recognized --status
[14:24] <xnox> Chryzo, typically you would not modify the header, you should modify files in /etc/resolvconf/resolv.conf.d and run $ resolvconf -u, to regenerate /etc/resolv.conf
[14:25] <xnox> Chryzo, you mean 16.04 LTS and 14.04 LTS, i presume, right?
[14:25] <Chryzo> by header i meant i modified /etc/resolvconf/resolv.conf.d/head :)
[14:25] <Chryzo> xnox, yep
[14:25] <xnox> ah =) well done, that's support / the right things to do, if things fail to propagate =)
[14:25] <xnox> ah =) well done, that's supported / the right things to do, if things fail to propagate =)
[14:26] <Chryzo> had to push some ansible playbooks so had to fix it that way for now. When I have time i ll need to try and investigate why it happens though
[14:26] <Buoy172> How do I search for "index.html" location from the command line?
[14:26] <Buoy172> btw, there's no www folder in /var
[14:28] <Chryzo> Buoy172,  find / -name index.html
[15:10] <teward> sdeziel: yep, fixed.  Sorry, i'm battling Sophos being stupid at the workplace so sorry for slow response
[15:10] <sdeziel> teward: thanks
[15:11] <teward> yep.  I've got a fix prepped for Cosmic.
[15:11] <teward> so whenever I get up off my lazy butt I'll upload it.
[15:17] <sdeziel> awesome
[15:18] <dr4c4n> hi there, does anyone have a simple apparmor profile writing example?
[15:18] <teward> apparmor profiles are by their nature 'not simple'
[15:19] <sdeziel> SELinux isn't simple, Apparmor is much more accessible
[15:19] <teward> sdeziel: creating the profile from scrach IIRC isn't easy either though
[15:19] <sdeziel> dr4c4n: I have a small collection of profiles I use: https://github.com/simondeziel/aa-profiles/
[15:20] <dr4c4n> sdeziel: thanks ! I'm looking for an easy example, for understanding
[15:22] <sdeziel> dr4c4n: if you want, pick a daemon/binary you know from that list https://github.com/simondeziel/aa-profiles/tree/master/16.04 and I'll be happy to answer your questions re its Apparmor profile
[15:24] <dr4c4n> I have a project idea, (sounds crazy) to write a conversion between an apparmor profile to an selinux policy
[15:25] <sdeziel> teward: yeah starting from scratch can be tedious which is why I always copy from an existing profile ;) I'm lazy too
[15:25] <sdeziel> dr4c4n: oh, that sounds pretty involved to me
[15:25] <teward> depends on which level of "lazy" though
[15:25] <teward> sdeziel: and for what they *sound* like they're doing "tedious' is an understatement
[15:25] <teward> BTW, thanks for your ongoing help with nginx bug triage
[15:26] <teward> it's nice to be able to ignore most of the bugs because others help triage them :P
[15:26] <teward> and by "ignore" I mean "acknowledge but not do anything with"
[15:26] <teward> :P
[15:26] <sdeziel> teward: that's my way to say thanks for the SRUs you are doing :)
[15:26] <teward> heh, well, that's just my job :)
[15:26] <teward> all part of maintaining the software :)
[15:26] <dr4c4n> so there are somethings that selinux can't do, for instance confine NFS mounts specifically that apparmor can do, and selinux can enforce RBAC and sensitivity levels, that apparmor can't.
[15:27] <teward> *waits for his cosmic sbuild to finish*
[15:29] <sdeziel> dr4c4n: not sure what you mean by confine NFS mounts. Got an Apparmor example of what you have in mind?
[15:32] <Neo4> what is version php have ubuntu by default? There written 7.0 and when I install I get 7.2
[15:32] <Neo4> and then when install php5.6 apache can't run
[15:33] <Neo4> what is current version php in ubuntu?
[15:33] <sdeziel> Neo4: the supported version differs from one Ubuntu release to the other
[15:33] <Neo4> sdeziel: I use 16.04
[15:33] <sdeziel> then it would be 7.0.30
[15:33] <Neo4> there apt-cache show php
[15:33] <Neo4> no, it 7.2?
[15:33] <Neo4> I now remove all and check
[15:34] <sdeziel> If you want 7.2 you need 18.04
[15:34] <Neo4> sdeziel: no, I want 5.6
[15:35] <Neo4> many wordpress plugins doesn't' support even php 7.0
[15:35] <sdeziel> Neo4: that version was never officially supported on Ubuntu so you have to resort to PPA
[15:35] <sdeziel> or even better find better maintained plugins
[15:36] <Neo4> sdeziel: yes, I load from ppa and before it works, not doesn't
[15:36] <Neo4> sdeziel: becaue now version 7.2
[15:36] <Neo4> sdeziel: how to check which version?
[15:37] <Neo4> at first remove php:
[15:37] <Neo4> sudo apt-get purge php*
[15:37] <Neo4> then install
[15:37] <Neo4> sudo apt-get install php and check?
[15:38] <sdeziel> Neo4: I don't know/use PPAs providing PHP
[15:38] <teward> there's one.
[15:38] <sdeziel> Ondrey Sury?
[15:38] <teward> one *decent* one by one of the well-known people for their PHP contribs
[15:38] <teward> yep
[15:38] <Neo4> sdeziel: see https://paste.ubuntu.com/p/c3MwdQCtqh/
[15:39] <Neo4> 7.2
[15:39] <teward> that's the only way to get the backported PHP versions to Xenial, unless you're me and are a power user backporting it in order to apply certain patches that PPAs and upstreams said no to for local envs :P
[15:39] <Neo4> how to check ubuntu version?
[15:39] <sdeziel> Neo4: you installed the generic (non versioned) package so you got the latest from the PPA
[15:41] <Neo4> sdeziel: why not from official repository?
[15:41] <Neo4> sdeziel: see https://paste.ubuntu.com/p/qDBq32wTxJ/
[15:41] <Neo4> version and description php
[15:41] <Neo4> it shoudl be 7.0
[15:41] <Neo4> sdeziel: that php7.2 break all my shell script
[15:41] <sdeziel> Neo4: then don't install it
[15:42] <Neo4> sdeziel: no it work well when I install at first from official repository php 7.0 and then php 5.0 from not official
[15:42] <Neo4> it doesn't works because I don't install php7.0
[15:43] <Neo4> sdeziel: can I somehow off that not official repository?
[15:44] <sdeziel> Neo4: add-apt-repository --remove
[15:44] <Neo4> sdeziel: yes, :)
[15:47] <Neo4> sdeziel: I'm doing shell script that installing applications one by one
[15:49] <Neo4> sdeziel: see https://ibb.co/i2Vevd
[15:50] <Neo4> yes, now 7.0 load, nice
[15:50] <Neo4> sdeziel: listen, should I off repository after installation php 5.6?
[15:50] <Neo4> is php equal in both repository?
[15:50] <Neo4> i mean 7.0
[15:51] <sdeziel> Neo4: you need to keep the PPA if you want (security) updates which Ondrej seems to provide regularly
[15:51] <teward> PHP is not equal in both the repos.  You cannot cohabitate the Repos PHP 7.0 with PPA 7.2
[15:51] <Neo4> it should be equal in both and doesn't matter from where you load?
[15:51] <teward> sdeziel: he also needs to install php 7.0 from Ondrej's repo as well, if co-usage is needed
[15:51] <teward> they fight each other otherwise
[15:51] <sdeziel> Neo4: if all you need is 7.0 and 5.6, I'd use 7.0 from Ubuntu and 5.6 from Ondrej's PPA
[15:51] <Neo4> sdeziel: but I mean when I include ppa all libs for php7.0 load from this php, I load php7.0 from ubuntu and libs not
[15:52] <sdeziel> oh right
[15:52] <teward> sdeziel: he'll have to remove the PPA to avoid PPA clobbering of php7
[15:52] <sdeziel> Ondrej's version have the "+" in their version
[15:52] <teward> assuming he's using the only PHP PPA usable.
[15:52] <teward> (the others are deprecated)
[15:52] <Neo4> teward: no, I remove before install php, in script , it's temporarely
[15:53] <Neo4> then script after install php5.6 add thsi ppa and it is left
[15:53] <Neo4> it was install by php , 7.2 version if we repeat run shell script
[15:53] <teward> sdeziel: i should also point out PHP 5.x is dead, so I don't know if there's any security updates any more.
[15:53] <teward> not without lots of backport work
[15:53] <teward> ... i need lunch, back in a while.
[15:53] <sdeziel> teward: upstream is still putting out 5.6 versions
[15:54] <sdeziel> dunno for how long though
[15:54] <sdeziel> but yeah, 5.x is close to dead
[15:54] <Neo4> teward: no, 5.6 isn't dead, I tried use php +7 and impossible with wordpress, blugins don't works
[15:54] <Neo4> php5.6 will live long
[15:54] <teward> sdeziel: ah, well, it's probably 'dead' soon enough.  either way.  *lunchtime*
[15:55] <sdeziel> blugins == plugins for blogs? :P
[15:55] <sdeziel> or plugins full of bugs?
[15:56] <Neo4> I would use php7 if it worked
[15:56] <Neo4> php7 has more possibilities
[15:59] <teward> it sounds to me like you need upgraded wordpress *and* upgraded plugins, because most of the plugins that worked in PHP 5.x work in updated Wordpress on PHP7, if you install the PHP7-compatible Wordpress
[15:59] <teward> (evidenced by the ten or so Wordpress deployments I maintain on PHP 7, just saying)
[16:00] <sdeziel> IIRC, WP upstream now supports and recommends 7.2
[16:03] <Neo4> sdeziel: ?
[16:04] <Neo4> teward: CMS supports but plugins not
[16:05] <Neo4> sdeziel: what do you use for install applications on server?
[16:05] <Neo4> sdeziel: do you want to test my script?
[16:05] <sdeziel> Neo4: no thanks, I use Puppet
[16:05] <Neo4> sdeziel: I little read about puppet in one book
[16:06] <Neo4> sdeziel: can you create virtual hosts in puppet?
[16:07] <sdeziel> yes
[16:07] <Neo4> something like select menu "add virtual host" and then follow tips input name and everything should be done
[16:07] <Neo4> sdeziel: really?
[16:07] <Neo4> sdeziel: what you can't do in puppet that I can do in shell?
[16:10] <Neo4> about repository, when we install ppa, ubuntu give priority packages from outside repository or which last in list?
[16:11] <Neo4> ok, it doesn't mater, it was by the way
[16:11] <Neo4> sdeziel: I will use "grandfather method" shell script for deploy :)
[16:50] <sruli> does a network card need to support nic bonding? i set it up yesterday on a server works fine, now i am trying on a old server but not working
[17:09] <sdeziel> sruli: no, this is NIC agnostic AFAIK
[17:09] <sdeziel> sruli: some bonding modes require cooperation from the switch ports though so maybe that's what different between the old/new servers
[17:11] <sruli> sdeziel: bonding in 802.3ad, switch supports it, i have 1 server working fine with 4 bonded nics, dont know how to troubleshoot this, ubuntu 18.04
[17:13] <sdeziel> 802.3ad support is a per-port thing so I'd double check this
[17:13] <sdeziel> on the switch that is
[17:14] <sruli> sdeziel: its configured correctly on the switch, even tried putting it in the ports that worked for the other server
[17:18] <sdeziel> sruli: I have never really had to debug a bond but maybe this tool would hint you in the right direction: http://folk.uio.no/trondham/software/check_linux_bonding.html
[17:22] <sruli> thanks
[17:26] <rbasak> nacc: FYI, git-ubuntu snap builds are currently failing. Nothing changed in our repo. kyrofa is looking in to it but he's out this week.
[17:29] <nacc> rbasak: yeah i saw the email
[17:29] <nacc> looking
[17:29] <rbasak> nacc: no rush!
[17:29] <rbasak> nacc: my MP is separate to that.
[17:30] <sruli> sdeziel: the tool output is "warning... number of AD ports (1) does not equal number of slaves (2)" however i have green lights on both nics and both ports on switch
[17:30] <rbasak> I'm not sure I'll want to land MPs without the snap build succeeding. But we could still make progress and get them to approved state on the assumption that we'll land once CI is passing again.
[17:30] <rbasak> (given that we're confident that the changes won't impact the snap build)
[17:31] <nacc> rbasak: i'm fine with you taking your cleanup branch (for the tmpdir changes)
[17:31] <nacc> rbasak: that one would be nice to land once CI works, since it does affect the bastion
[17:31] <rbasak> nacc: shall I file a superseding MP then? Would you like to review it, or was that your review? :)
[17:31] <sdeziel> sruli: looks like one of the NIC was not able to setup LACP with the switch
[17:31] <rbasak> nacc: or would you prefer ahasenack to review?
[17:31] <nacc> rbasak: file and i can review (ahasenack can too)
[17:31] <rbasak> OK
[17:32] <sruli> sdeziel: will test each separately for regular dhcp and see where i get
[17:34] <nacc> rbasak: looks like a patch changed
[17:34] <nacc> rbasak: in snapcraft.yaml:          - -lib/python3.6/site-packages/lazr.restfulclient*
[17:34] <rbasak> https://code.launchpad.net/~racb/usd-importer/+git/usd-importer/+merge/345617
[17:34] <nacc> rbasak: in the build log:          - -lib/python3.6/site-packages/lazr.restfulclient*
[17:34] <nacc> bah
[17:35] <nacc> in the build log: lib/python3.6/site-packages/lazr/restfulclient/__init__.py
[17:35] <nacc> lazr/restfulclient vs. lazr.restfulclient
[17:35] <nacc> *possibly* we were getting away with a regex match before?
[17:35] <rbasak> It sounds like you're onto something, but I'm not sure I follow.
[17:36] <nacc> rbasak: our yaml does not stage (in the git-ubuntu part) the lazr.restfulclient package
[17:36] <nacc> rbasak: because we know we get it from the launchpadlib part
[17:36] <nacc> rbasak: perhaps snapcraft changed how it parses those paths, so, and allows for globbing only, and not arbitrary regex
[17:36] <nacc> a simple test would be to change the yaml to be -lib/python3.6/site-packages/lazr*
[17:36] <nacc> and see if cleanbuild works
[17:37] <rbasak> I'll try that, thanks.
[17:38] <nacc> np
[17:38] <nacc> ah there is also lazr/uri
[17:38] <nacc> so you need to be on more path strict, sorry
[17:39] <nacc> try: -lib/python3.6/site-packages/lazr/restfulclient*
[17:39] <nacc> rbasak: --^
[17:39] <rbasak> As the only line that includes lazr?
[17:39] <nacc> line 401 of the yaml
[17:40] <rbasak> Running (as the only line)
[17:40] <rbasak> Dinner's here. I'm not sure I'll be back later today.
[17:40] <nacc> http://paste.ubuntu.com/p/77DP922V9z/
[17:40] <nacc> rbasak: np
[17:40] <rbasak> Thanks. That's exactly what I'd guessed.
[17:41] <sruli> sdeziel: independently each nic works fine, i am really stuck
[17:44] <sdeziel> sruli: I'd check if both NICs negotiate the same duplex/speed
[17:44] <sdeziel> beyond that, you'll have to wait for others to help as I don't know LACP all that well
[17:45] <sruli> sdeziel: just checked in the bios, option to enable embedded nic1 & nic2 or enable nic1/gb i guess they migh not be same speed, waiting for it too boot now will be able to check
[17:46] <sdeziel> sruli: check with ethtool on each NICs
[17:47] <sruli> sdeziel: thanks, will report in 2 min
[17:59] <sruli> sdeziel: both nics reporting exact same
[18:05] <sdeziel> sruli: oh well
[18:25] <sruli> sdeziel: just tried bonding with 2 usb to eth, same issue, maybe some dependency missing? its a fresh install
[18:46] <Chryzo> Been looking into clustering for HA without needing to control resources (just running a load balancer on the cluster members) which software would you recommend (Corosync, Heartbeat, other ?)
[18:50] <Chryzo> i just need to move an IP between them
[18:51] <dpb1> corosync+pacemaker is the usual solution to control a vip
[18:54] <sdeziel> keepalived works well for that too
[19:00] <Chryzo> just installed keepalived, super easy to configure and seems to work amazingly. I pretty much needed to just move a floating IP between 2 servers. Thanks for the help
[19:01] <Chryzo> some days you just gotta love linux :)
[19:37] <sruli> 18.04 headless server, how do i renew dhcp lease? if i disconnect the eth cable i cannt reconnect it without a reboot
[19:47] <tomreyn> sruli: does "systemctl restart networking" help?
[19:52] <sruli> tomreyn: no, dhclient takes forever and does not renew, checking tcpdump form another tty shows its trying to get ip from 0.0.0.0 but its definitely on the network as the dump shows all the traffic from other ip's on the network
[19:53] <Blueking> hello
[19:54] <genii> sruli: For clarification: Are you trying to renew the IP from the client, or are you trying to re-issue a number to the client  from the server?
[19:55] <sruli> genii: in respect of the ip the server is the client
[19:56] <Blueking> I have a question...  not sure where I can get info about this. But question are about plex server to get benifit from hardware accelerated feature from gpu (igp)  but if one got  motherboard that hasn't got hdmi or displayport output, then question are if  xeon cpu with igp on server mobo, would igp be disabled ?
[19:56] <Blueking> or would plex sever get access to igp ?
[20:05] <sruli> i guess will reintall 16.04 and see if i have the same issues
[22:42] <rbasak> nacc: snapcraft.yaml change> looks like it worked. Thanks!
[22:42] <rbasak> nacc: so what exactly changed?
[22:42] <rbasak> (to break it)
[23:29] <nacc> rbasak: i'd need to talk to kyrofa about it
[23:29] <nacc> rbasak: my initial guess is that before, snapcraft took the paths as regex, so we had a '.' that allowed for /
[23:29] <nacc> rbasak: but something (presumably in snapcraft) changed and made it a literal '.'
[23:30] <nacc> rbasak: alternatively, lazr.restfulclient's installed path change, but I find that harder to believe
[23:38] <rbasak> nacc: that makes sense, thanks. I'm not sure what to put in the commit message to fix it, but at least we have what looks like a fix then.
[23:39] <rbasak> nacc: kyrofa is out this week, but hopefully we can land a fix on Monday then?
[23:39] <nacc> rbasak: yeah, i mean it's not 'wrong' to make the change even now
[23:40] <nacc> as if the path 'happened' to work before, it's better to use the right path anyways :)