[00:11] <Epx998> i think i finally beat this thing into submission
[06:48] <cpaelzer> good morning
[06:48] <rbasak> o/
[11:42] <Jenshae> Greetings and salutations, denizens and ladies.
[11:44] <Jenshae> joelio: I have set up the machine with the SSD to boot it and using " sudo zfs mountpoint=/home/user pool/user " I have linked it onto the SSD. How do I now confirm how much available space is in that folder? Disk Analyser for example only shows what is used but not what is left as free space.
[13:11] <gunix> anybody used juju charms to deploy openstack?
[13:48] <jamespage> gunix: yes
[13:48] <jamespage> gunix: #openstack-charms is a good place to go ask questions on that topic, but here is ok as well
[13:49] <jamespage> coreycb: I've been working on deps again today - mostly done - just working on the stestr/ostestr bits and pieces to allow the neutron/ovsdbapp uploads
[13:49] <jamespage> coreycb: zunclient builds now - can you upload a backport-less version to the PPA please
[13:49] <coreycb> jamespage: great, yep will do
[13:52] <coreycb> jamespage: i recently added py2 support for python-daiquiri (it's py3 only in debian) but I think I can drop that in favor of moving gnocchi to py3 only.
[13:52] <jamespage> coreycb: agreed
[13:53] <coreycb> jamespage: i also need to fix dep8 tests for daiquiri so i'll work on that in the ci-train ppa. i'll look a gnocchi today too.
[13:56] <joelio> Jenshae: with ZFS it's (kinda) sparsely allocated so you can add more mounts from the pool.. what does df say?
[13:57] <jamespage> coreycb: I'm going to start working on some of the core projects now
[13:57] <jamespage> will use the ss to track
[13:57] <jamespage> coreycb: gnocchi might be a good place to start I guess
[13:58] <coreycb> jamespage: great, will do the same!
[14:19] <jamespage> coreycb: we're working on the same thing here
[14:19] <coreycb> jamespage: are you doing gnocchi?
[14:19] <jamespage> yes
[14:19] <jamespage> 4.1.0
[14:19] <coreycb> jamespage: ok go for it
[14:19] <jamespage> coreycb: ta
[14:20] <jamespage> sorry about that - did not make it super clear that I was going to start on gnocchi
[14:22] <coreycb> jamespage: np, i had a start on it but wasn't working so i'll diff vs what you end up with and learn something
[14:23] <joelio> Jenshae: zpool list
[14:31] <jamespage> coreycb: synced https://launchpad.net/ubuntu/+source/python-tenacity/4.4.0-2 btw - no good reason to keep our diff
[14:32] <coreycb> jamespage: yep works for me
[14:48] <coreycb> jamespage: will that require breaks/replaces to replace python-gnocchi with python3-gnocchi?
[14:49] <jamespage> coreycb: no - the pathing is different
[14:51] <coreycb> jamespage: ok. shall we move all core projects to py3 only where possible for this milestone?
[15:03] <cpaelzer> dosaboy: jamespage: on bug 1466926 I have a ppa and some expertements, but I fail to reproduce so I'm unsure if it is a fix
[15:03] <cpaelzer> dosaboy: jamespage: since you meant you had some context in this issue, if you know how to trigger it please help to get some steps to reproduce that work
[15:08] <Ussat> so, played with netplan a lot lastnight....I can live with it
[15:16] <drab> it's not gonna be pushed down to LTS, is it?
[15:17] <drab> I mean xenial
[15:21] <sdeziel> drab: netplan is available for Xenial (in universe though)
[15:23] <Ussat> its actually very useable once ya get used to it
[16:07] <drab> I'm sure, I just can't imagine to learn something new and update all the instances we have
[16:07] <drab> I just have enoguh work for the next 2 yrs
[16:09] <rbasak> It's office hours in #ubuntu-server. Highlight rbasak for attention. See https://community.ubuntu.com/t/irc-meeting-office-hours/1491 for details.
[16:10] <slashd> rbasak, dpb1 I have one thing since there is no agenda anymore....
[16:10] <rbasak> Sure  :)
[16:11] <slashd> We are working on a MIR for pcp (LP: #1700827).... we got an ACK from security and few comment by MIR approval team... we need to find a owner for pcp/papi bug
[16:11] <slashd> could server team look if they can own them ?
[16:12] <rbasak> That's really a question for the Canonical server team, rather than us wearing an Ubuntu hat.
[16:12] <slashd> rbasak, sure make total sense
[16:12] <rbasak> And for that, it's down to dpb1, who is travelling at hte moment.
[16:12] <slashd> rbasak, I'll ping dpb later then thanks
[16:12] <ahasenack> what's missing, sponsorship?
[16:13] <rbasak> I assumed he needs a team subscription
[16:13] <ahasenack> ah, finding an actual owner
[16:13] <ahasenack> before sponsoring
[16:13] <ahasenack> I see
[16:13] <slashd> rbasak, yep team subscription
[16:13] <rbasak> And this usually comes down to: is dpb1 willing to commit Canonical resources to maintaining the package in main.
[16:13] <slashd> rbasak, ack I'll contact him
[16:13] <slashd> rbasak, thanks
[16:14] <rbasak> You're welcome :)
[16:21] <cpaelzer> slashd: last time I tested that things it broke on about every place possible - so make sure I'm not asked if we want to own
[16:21] <cpaelzer> slashd: was the new dependency I pointed out in th gdoc sorted out - I didn't see an update but must admit I didn't check explicitly
[16:24] <slashd> cpaelzer: dgadomski tested it and he couldn't reproduce what you are experimenting
[16:25] <slashd> dgadomski, you reviewed the pcp gdoc right ? and couldn't reproduce what cpaelzer had experimented ?
[16:26] <dgadomski> slashd: I tried to crash any binary I could find in the package, was there a particular scenario to test in the gdoc?
[16:27] <slashd> dgadomski, I don't think there was an exact reproducer AFAIK
[16:27] <dgadomski> so I did my best when I was working on the autopkgtests
[16:27] <dgadomski> without any segfaults
[16:27] <slashd> and cpaelzer the code that has been ACK by security is the newest upstream one, not what is found in ubuntu atm
[16:28] <slashd> after they made serious improvement
[16:28] <cpaelzer> maybe the new version is better
[16:28] <cpaelzer> I'll take a look if I find some time
[16:29] <cpaelzer> slashd: and about the dependency?
[16:29] <slashd> cpaelzer, the papi dep is under review by MIR approval team atm and security team
[16:30] <cpaelzer> slashd: ok, and no depeer follow on dependencies there?
[16:30] <slashd> dgadomski, ^
[16:31] <dgadomski> no, I don't think there are any non-main dependencies involved
[16:35] <slashd> cpaelzer, so papi is waiting on security team ACK for now, and pcp I'm working on a few stuff asked by MIR team (switch dpkg-source from 3.0 (native) to 3.0 (quilt), ...
[16:36] <slashd> thanks dgadomski
[16:36] <dgadomski> thanks
[16:37] <slashd> cpaelzer, let us know if you need anything from us
[16:56] <xnox> cpaelzer, somehow i was under the impression that we do have /dev/kvm inside our e.g. Openstack Virtual machines.
[16:56] <xnox> cause e.g. we do launch nested kvm, as part of systemd autopkgtest
[16:57] <xnox> smb, is ubuntu kernel patched to somehow speciality default to nested_kvm=1?
[16:58] <smb> xnox, it does not change defaults, so intel no, amd yes. but that was changed by kvm-qemu install
[16:59] <smb> something to add modules parm into /etc/modprobe.d iirc
[17:00] <smb> xnox, on xenial.x86: cat /etc/modprobe.d/qemu-system-x86.conf
[17:00] <smb> options kvm_intel nested=1
[17:01] <xnox> cpaelzer, are you planning to keep that? ^
[17:02] <xnox> cpaelzer, smb - i wonder if we need to sru qemu to specify nested=1 option on s390x too, exted qemu-system-init to load module on s390x; and then sru kernel that makes kvm a module, rather than a built in.
[17:02] <xnox> cpaelzer, smb - or can we somehow change the built-in module option as a kernel config change? (if that at all makes sense, no idea if built-in module options are tweakable)
[17:04] <smb> xnox, that option might be an early one, which would require it on kernel cmdline. generally option values can be changed via sysfs but if its something done at init time there would be no effect
[17:04] <xnox> smb, code patch ? =)
[17:04] <xnox> #ifdef __s390x__
[17:05] <xnox> int nested=1;
[17:05] <xnox> #else
[17:05] <xnox> int nested=0;
[17:05] <xnox> #endif
[17:05] <xnox> lolz
[17:08] <rbasak> (office hours are now over)
[17:14] <Jenshae> joelio: Not familiar with df but need to go now. o7
[19:35] <azerty> hi there
[19:36] <azerty> is it possible to animate the color of shell ?
[20:02] <dpb1> if he were here, I would tell him, kind of
[20:05] <sarnold> dpb1: well now I'm kinda curious :)
[20:05] <sarnold> I've seen those PS1 tricks .. but that's not really what I'd call _animated_
[20:08] <dpb1> heheh
[20:09] <dpb1> I think those are the closest I got!
[20:09] <sarnold> I mean people are doing crazy things with terminals these days https://github.com/p-e-w/ternimal but .. that's not really the shell that's animated there, hehe
[20:10] <dpb1> sarnold: I like the fallout themed one, like an old CRT.
[20:12] <dpb1> "cool-retro-term"  great package name
[20:28] <drab> is there a known nice ubuntu based rescue iso that ppl use?
[20:28] <drab> something that will ship with dd_rescue, some data recovery stuff,and general debugging tools
[20:29] <drab> systemrescuecd seems to not be working very well anymore
[20:36] <drab> https://distrowatch.com/table.php?distribution=ubunturescue
[20:36] <drab> but apparently it's now discontinued
[20:37] <genii> Not ubuntu based, but UBCD is very useful
[20:37] <genii> !ubcd
[20:37] <genii> Hm
[20:37] <genii> drab: http://www.ultimatebootcd.com/
[20:42] <drab> sadly enough here people need GUIs and to an extent that's fair beacuse sometimes we need to debug audio or video peripherals on desktop machines and need GUI for real
[20:42] <drab> I might just have to use an ubuntu livecd and have a script that pulls in the tools I need
[20:42] <drab> to bite the bullet and figure how to customize an iso
[20:43] <drab> I tried a few times and always ended up being a painful experience
[20:46] <genii> I think remastersys might still be a thing for that
[20:47] <genii> Alternately, just install to a USB stick, boot to that and install things, make copies
[20:48] <drab> oh, that's not a bad idea. I wanna pxeboot tho. I'm wondering if I then dd the key to a file if I can boot that
[20:50] <genii> If you go that way, might want to make modifications so GRUB only looks for the local hdd it booted from, and revert to old eth0 wlan0 interface naming scheme
[20:51] <drab> good points
[20:51] <genii> So if you use some scripts the naming can be the same on any machine, plus GRUB won't make entries for a computer that's it's not currently plugged into
[20:51] <drab> maybe I'll give it a go if I get a minute one of thse days, not super urgent
[20:51] <drab> fot now I'll just give ppl a live ubuntu desktop
[20:52] <drab> it'll keep them busy for a while :)
[20:52] <drab> pxe booting liveCD is really neat, wish I knew that years ago
[20:52] <drab> actually tbf I found you can pxe boot squash files
[20:53] <drab> cuztomizing that should be simple as well
[20:53] <drab> in a chroot, but maybe not
[20:56] <genii> And with dnsmasq pxe is way easier than the old days of setting up isc-dhcp-server
[21:18] <drab> the only thing I wish dnsmasq had is a db backend so that I could script it more easily, especially from the inventory system and tied into new device registration
[21:18] <drab> similar to dns with pdns
[21:18] <drab> (if one wanted to use that for dns instead of dnsmasq, which can be useful)
[22:05] <DammitJim> man, I am so confused
[22:05] <DammitJim> I think I need to add more RAM to my ubuntu servers, but it's hard to tell
[22:06] <DammitJim> I"m using free -m and it's showing I'm basically using all my RAM
[22:06] <DammitJim> however, if I run: cat /proc/meminfo, it shows like more than half of the RAM as Inactive
[22:06] <DammitJim> what does that mean?
[22:08] <TJ-> DammitJim: 'free' RAM is used for caches and buffers
[22:08] <TJ-> !ram | DammitJim
[22:10] <DammitJim> so, I am OK by just looking at "used memory" and subtracting that from the total memory
[22:10] <DammitJim> that gives me a good number for utilization, right?
[22:10] <DammitJim> versus looking at "free" memory
[22:44] <sarnold> dammitjim has a knack of leaving just before I get here
[23:49] <drab> didn't freenode allow you to record a msg for registered users?
[23:50] <sarnold> yeah, it probably still does, but it's hard to spot when logging in
[23:50] <drab> fair enough
[23:51] <sarnold> drab: the last time I used it, the guy replied to me weeks later, it took a while to find it, hehe
[23:56]  * drab just typoed one of his favorites
[23:57] <drab> I realize it's lame, but, while telling, it lifts up my humor every time it happens: sudo shitdown -h now
[23:57] <sarnold> heheh