[00:03] <michael2> hi. does anyone know where the root cronjobs are stored?
[00:05] <mason> michael2: man 1 crontab
[00:07] <michael2> Ive got some apt updating process running as a crobjob - Im trying to find out where it is stored
[00:09] <mason> The answer lies in the very first paragraph of the man page. :)
[00:09] <mason> ...along with other good stuff.
[00:41] <teward> I knew there was a reason I hated the Perl packages... it holds up a lot of stuff...
[01:01] <xnox> teward, remove perl is my lifetime goal! imho, it is a good ecosystem/language, but in no way it should be this core / this system-low-lever.
[01:01] <xnox> *level
[01:25] <teward> xnox: heh.  (It's hanging up getting the latest NGINX upload into the release.  Again.)
[01:28] <teward> at least this time I can't blame the searx package :P
[01:33] <blackflow> michael2: it's handled by apt-daily.timer, not a cronjob, if you're talking about the default apt updating task
[01:34] <blackflow> michael2: otherwise, packages drop cronjob scripts under /etc/cron.d/  or /cron.{daily,hourly,monthly,weekly}  depending on the task
[01:57] <mason> xnox: It seems safer than Python for system stuff at present!
[01:57] <mason> blackflow: But, I gave him the fishing manual!
[01:58] <mason> Oh, he modified his question.
[02:14] <mwhudson> teward: well there's only 1260 packages in proposed currently by my count...
[04:59] <cpaelzer> thanks teward, sounds good
[04:59] <cpaelzer> and yes the port 80 bug comes up on triage every now and then
[05:00] <cpaelzer> I saw you and sdeziel discussing, it lGTM on first sight - lets hope it works well and causes less issues on its own than it prevents :-)
[05:13] <cpaelzer> jamespage: FYI 1786157 is fixed as expected by the new package that got unblocked by the build dependency now being available
[05:13] <cpaelzer> let me know if there is more to it
[07:46] <jamespage> cpaelzer: ta - I'll work through our blocked packages now
[07:47] <cpaelzer> jamespage: I've seen other people stumbling over qcow images not having set metadata correctly
[07:47] <cpaelzer> jamespage: as FYI when you are taking a look
[07:48] <cpaelzer> without virt-aa-helper can't follow the backing chain
[07:48] <cpaelzer> AFAIK openstack and libvirt will set up correctly since years
[07:48] <cpaelzer> just manually created qcow image files might be a problem
[07:49] <cpaelzer> https://www.redhat.com/archives/libvir-list/2018-August/msg00971.html
[07:50] <cpaelzer> jamespage: I doub't it will be an issue but wanted to know you informed :-)
[08:17] <blackflow> mason: yeah, XY questions are common :)
[08:30] <jamespage> cpaelzer: ta
[09:51] <Tekneek> Is it possible to provide a custom port for dns in /etc/network/interfaces? I want the system to use a name server that is not running on port 53, but have not been unsuccessful. I have tried <IP address>:XXXX and <IP address>#XXXX and the system continues to only try port 53.
[11:14] <tobias-urdin> coreycb: sorry for disturb, do you know anything about this?
[11:14] <tobias-urdin> http://logs.openstack.org/30/593530/1/check/puppet-openstack-integration-5-scenario003-tempest-ubuntu-bionic/df227a5/logs/apache/horizon_ssl_error.txt.gz
[11:14] <tobias-urdin> new openstack-dashboard package promoted just recently?
[11:14] <ahasenack> good morning
[12:22] <tobias-urdin> jamespage: do you know if coreycb is around?
[12:22] <tobias-urdin> http://mirror.ca-ymq-1.vexxhost.openstack.org/ubuntu-cloud-archive bionic-updates/rocky/main
[12:23] <tobias-urdin> as repo pulls in rc1 but according to this https://launchpad.net/ubuntu/+source/horizon rc1 should only be in proposed?
[12:23] <tobias-urdin> broke some horizon stuff with the rc1 packages
[12:48] <coreycb> tobias-urdin: I'll take a look and get back to you
[12:53] <tobias-urdin> cool, thanks
[13:19] <mason> blackflow: heh
[14:55] <ahasenack> cpaelzer: when an upstream tarball installs a systemd service file in make install, how to we declare/include that in debian packages?
[14:55] <ahasenack> cpaelzer: copy it do debian/ as foo.service and let dh_installsystemd handle it? but then there would be two copies of the service file
[14:55] <ahasenack> on somewhere in the upstream checkout, and one in debian/
[15:01] <rbasak> ahasenack: any chance your ndctl/pmdk work will need further updates to Bionic after this SRU?
[15:01] <rbasak> Because we'll be more locked down after the first set goes in.
[15:02] <ahasenack> rbasak: upstream just pinged me about an ndctl update
[15:02] <ahasenack> rbasak: but for cosmic
[15:03] <ahasenack> I think we should proceed, as I didn't get any specific request for this update for bionic, it was specific to cosmic
[15:38] <cpaelzer> ahasenack: no need to copy it out to debian/*
[15:38] <cpaelzer> dh_systemd tools will spot it
[15:38] <cpaelzer> if you are new dh_systemdinstall I think
[15:38] <ahasenack> and add the appropriate postinst bits?
[15:39] <ahasenack> looks like it
[15:39] <cpaelzer> old dh_systemd_start I think
[17:17] <ahasenack> cpaelzer: this krb5 ticket: now that krb5 migrated should I select "finalize" or "abandon"?
[17:27] <sarnold> blackflow: oh! thanks for the desktop netplan correction :) I'll have to look for it on my laptop ;)
[17:29] <sarnold> teward: nice, I like the port80 fix too
[17:29] <teward> sarnold: cpaelzer: indeed, I was trying to figure the best way to solve the issue, it just so happens that roaksoax's request for the separate daemon package made me go "FINE I'll Fix it..." and take a stab at it.
[17:30] <teward> and thanks to several individuals who tested, that fix should be good to go (so long as nobody rips out iproute2 randomly, which would probably destroy their system anyways?)
[17:31] <teward> mwhudson: heh, 1260 packages in proposed.  Of which how many are hung up with the Perl transition/proposed item I wonder :P
[17:31] <teward> well, it's sitting in Proposed now, so I just have to let the queue process itself :)
[17:38] <sarnold> teward: yeah if someone actually goes to the effort of removing iproute2 they're firmly in the "get to keep both pieces" category
[17:39] <teward> sarnold: oh I have a security team question for you if you don't mind me bothering you briefly?
[17:46] <sarnold> teward: sure :)
[17:46] <teward> see PMs
[18:00] <dpb1> oooh, new theme on cosmic, neat
[19:48] <tafa2> can anyone recommend a cli based speedtest
[19:48] <tafa2> that isn't speedtest-cli?
[19:50] <ahasenack> for internet, or local networking?
[19:50] <ahasenack> for an internet speed test, you could always download a large file and measure
[19:51] <dpb1> I'd use this: https://github.com/gesquive/fast-cli
[19:51] <dpb1> :)
[19:51] <dpb1> having never tried it
[19:52] <ahasenack> looks cool
[19:54] <cpaelzer> ahasenack: you want to abandon
[19:55] <ahasenack> cpaelzer: abandon it is, thanks
[19:56] <cpaelzer> ahasenack: If you uploaded with exactly  the version that gets merged (unlikely since you want to spin on ~ppaX) it would auto-abandon I think
[19:56] <tomreyn> tafa2: https://iperf.fr/iperf-servers.php
[19:57] <tafa2> tomreyn amazing thanks!
[20:12] <Epx998> The archive http://us.archive.ubuntu.com/ubuntu/dists/xenial/ will have 16.04.5 right?
[20:15] <sarnold> Epx998: yes
[20:16] <tomreyn> sarnold: these files look old to me, like 16.04.0
[20:16] <sarnold> sigh. yes. thanks tomreyn.
[20:16] <tomreyn> Epx998: what are you looking for exactly?
[20:19] <tomreyn> Epx998: maybe you're looking for http://us.archive.ubuntu.com/ubuntu/dists/xenial-updates/
[20:21] <nacc> Epx998: the url you provided is the release pocket, which basically doesn't change
[20:24] <Epx998> tomreyn: Adding 16.04.5 to a cobbler instance, looking for the correct repo to sync locally from
[20:25] <Epx998> i noticed the timestamps to
[20:25] <Epx998> What URL can I sync locally that will have 16.04.5 ?
[20:25] <Epx998> The xenial-updates?
[20:26] <ahasenack> no, that doesn't contain debs
[20:26] <ahasenack> the debs are always in a pool directory, but they are all bundled together
[20:26] <ahasenack> updates, release, etc
[20:27] <tomreyn> Epx998: cobbler, so you're looking for netboot images?
[20:27] <ahasenack> if you really only want to mirror updates, you need a tool one level above a plain rsync, something like reprepro, or apt-mirror
[20:27] <ahasenack> or mirror the iso directly
[20:27] <sarnold> I strongly recommend the double-rsync approach instead.. I've found it more reliable than my friends have found reprepro or apt-mirror or deb-mirror or whatever
[20:27] <ahasenack> that is a "snapshot" of a subset of 16.04.5
[22:01] <nacc> Epx998-: to be clear, do you understand what 16.04.5 is? :) are you asking for the latest xenial packages? or the 16.04.5 hwe kernel and x stacks in an installer image?
[22:07] <Epx998> nacc: I am working in a lab that uses cobbler, they needed help install ub 16.04.5 so I am looking for the correct url to point to to get .5
[22:10] <Epx998> maybe i grabbed the wrong net boot images
[22:12] <tomreyn> Epx998: refer to the architecture dependant 'installer'  directories at http://us.archive.ubuntu.com/ubuntu/dists/xenial-updates/main/ and their current/images/netboot/ subdirectories.
[22:15] <Epx998> roger that
[22:24] <nacc> Epx998: so they want specifically the 16.04.5 kernel and x?
[22:24] <nacc> Epx998: 16.04.5 is just a point in time if you are not using hwe
[22:26] <Epx998> yeah
[22:38] <nacc> Epx998: right, then you want the correct netboot image as tomreyn mentioned
[22:47] <Epx998> nacc: thats the net image i used, ill have to re visit and make sure
[22:55] <X-Rob> tomreyn: been here for years 8)
[22:55] <tomreyn> :) ok, feel free to ask support questions here, too.
[22:57] <X-Rob> Well, I realised that I didn't take a snapshot of the VM before I started installing stuff on it
[22:57] <X-Rob> so I'm reinstalling, but this time I'll reinstall WITHOUT adding keys
[22:57] <X-Rob> see if that is actually the problem
[22:58] <X-Rob> https://www.irccloud.com/pastebin/HsGAXfv1/shotgun.txt
[22:58] <tomreyn> it's unrelated to installing ssh keys
[22:59] <X-Rob> Ah poot.
[22:59] <tomreyn> (but it may be related to cloud-init, not sure)
[23:00] <X-Rob> Someone did say it was related to cloud-init, in the ticket.  The only other non-standard thing I do is change the mirror to australia
[23:00] <X-Rob> https://usercontent.irccloud-cdn.com/file/Ht2wQsxV/mirrorchange.png
[23:01] <tomreyn> my understanding is that it is generic, affects every installation, whether or not you use defaults.
[23:01] <X-Rob> Oh, wow.
[23:01] <X-Rob> So that EVERY 18.04.1 installation is broken right now?
[23:01] <sarnold> ?
[23:01] <X-Rob> Shouldn't that be like a 'pull the ISO this is really bad' situation?
[23:02] <tomreyn> not broken, just not perfect ;)
[23:02] <X-Rob> Technically correct, the best kind of correct, tomreyn
[23:03] <X-Rob> sarnold: talking about this one: https://bugs.launchpad.net/subiquity/+bug/1783129?comments=all
[23:04] <X-Rob> Broken before cloud-init runs. So nope, I was wrong. https://usercontent.irccloud-cdn.com/file/VaSohmRs/image.png
[23:04] <sarnold> X-Rob: thans
[23:04] <sarnold> thanks
[23:04] <sarnold> I hadn't seent his one before
[23:06] <tomreyn> sarnold: there are more subiquity bugs which are still 'new' some of which are about broken / non booting installations.
[23:06] <tomreyn> (18.04.1)
[23:06] <sarnold> huh, is it itentional for the *cd* image to be included in sources too? that seems really inconvenient for 99% of users..
[23:07] <tomreyn> i mean they're filed against subiquity, i can't say whether those are bugs in subiquity
[23:07] <sarnold> tomreyn: yeah, in the security team bug triage role, it's easy to spot something like a dozen bugs each week folks had with problems in the various installers.. :(
[23:07] <X-Rob> I REALLY like the installer, btw. No idea when it was changed, but it's REALLY good, and I love the accessibility of it. Good colours, high contrast, and text so it can be screenread easily.
[23:08] <tomreyn> sarnold: i think apt handles the missing cdrom gracefully if its not mounted, and IIRC d-i always did it this way
[23:08] <sarnold> tomreyn: oh okay
[23:08] <sarnold> it just gave me flashbacks to suse in 1997 or something where it wouldn't let you download the apache update until you'd installed apache off disc 3 or something like that..
[23:08] <tomreyn> i think the cdrom apt source was disabled back then, though
[23:09] <sarnold> I never used the debian cdrom: source..
[23:09] <sarnold> "we'll use the slowest block device you've got! and the data is old!"
[23:10] <X-Rob> Slowest?  1.44mb Floppy: Hold my beer..
[23:10] <X-Rob> oh wait
[23:10] <X-Rob> that WAS floppy
[23:10] <tomreyn> the cdrom source is definitely handy in airg apped environments.
[23:11] <X-Rob> OK, so curtin just ran
[23:11] <X-Rob> https://usercontent.irccloud-cdn.com/file/H6a8Yf90/curtin.png
[23:11] <X-Rob> Lets see if it's broken now
[23:12] <tomreyn> ...or while the corporate firewall team is on holidays and didnt get around to handle your policy request, yet
[23:13] <X-Rob> Yep, broken everywhere.
[23:13] <sarnold> X-Rob: heh, yeah, I remember doing slackware on 53 floppies.. fun times :)
[23:13] <tomreyn> does vmware also send the webconsole password on this HTTP(S) GET request? :)
[23:13] <X-Rob> sarnold: I used to be a slackware mule, back in the day.  Uploading slackware floppies to BBS's all around Australia.
[23:14] <X-Rob> tomreyn: ... whoops, I posed the uncropped snapshot. Ah well. Nothing secret there.
[23:15] <tomreyn> well, i kow you have set a bookmark on "New Tab" now.
[23:15] <sarnold> X-Rob: ah :D the hero for the other aussies :D
[23:16] <X-Rob> sarnold: I got SO MANY BBS credits, I could download whatever I wanted.
[23:16] <sarnold> hahahaha
[23:16] <sarnold> that's awesome :D
[23:17] <X-Rob> New Tab https://usercontent.irccloud-cdn.com/file/c4n2o6ii/newtab.png
[23:17] <sarnold> hheh
[23:49] <tomreyn> voip phone security is ~ on par with printers, if not even worse.