[02:31] <nacc> mwhudson: sweet: http://autopkgtest.ubuntu.com/packages/celery/artful/amd64
[02:32] <nacc> jamespage: mwhudson: unblocking the packages
[02:32] <mwhudson> nacc: \p/
[02:32] <mwhudson> \o/ rather
[02:33]  * mwhudson is making is very own das keyboard
[02:33] <nacc> mwhudson: i figured it was in profile
[02:33] <nacc> mwhudson: maybe a stretch or something :)
[02:41] <mwhudson> heh
[05:53] <jamespage> nacc: awesome - thanks again for all of your work on this!
[06:17] <lordievader> Good morning
[09:02] <db`> cron job for every 1.5 hours
[09:02] <db`> https://stackoverflow.com/questions/247626/how-can-i-set-cron-to-run-certain-commands-every-one-and-a-half-hours
[09:03] <db`> won't this work?
[09:03] <db`> */90 * * * *
[09:03] <db`> I didn't find it anywhere..
[09:04] <db`> all say about splitting the syntax.
[09:04] <db`> */90 * * * * won't work?
[09:16] <hateball> db`: no, as 0-59 are the allowed values for minutes
[09:16] <db`> okay
[09:16] <frickler> db`: "run in when $minute is a multiple of 90" won't happen very often. what's so bad about having two entries?
[09:17] <db`> Okay. will do it like that
[09:17] <db`> tahnks.
[09:43] <cpaelzer> jamespage: I got the patches I needed for openvswitch and tests look good.
[09:43] <cpaelzer> jamespage: I will likely upload the arm-dpdk+ftbfs-fix later today - let me know if there are any conflicts ahead
[09:43] <cpaelzer> patches are arm-dpdk only, still I checked and normal OVS seems unaffected as it should be
[09:44] <jamespage> cpaelzer: nothing that I am aware of
[09:46] <cpaelzer> thanks jamespage
[09:47] <jamespage> cpaelzer: no thankyou for your work in this area - its appreciated by me and the wider community of users :-)
[09:47] <cpaelzer> you make me smile, but I wonder if you just want to raise my mood
[10:42] <ahasenack> hm, why would someone add a "samba" tag to a bug against samba? Kind of redundant? :)
[11:07] <jamespage> cpaelzer: could you look at https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1697729 when you're next in libvirt land
[12:02] <cpaelzer> jamespage: added to my list
[12:03] <cpaelzer> jamespage: but by german luck (public holdiys) this week is a short one, so needs a bit of time
[12:05] <cpaelzer> but the bug and patch are nice and clear - so I don't expect anything stopping me
[12:09] <cpaelzer> jamespage: btw just saw that the new mininet fails on OVS Test in artful-proposed
[12:09] <cpaelzer> jamespage: seems like it needs an explicit shutdown now
[12:09] <cpaelzer> jamespage: https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-artful/artful/amd64/o/openvswitch/20170531_164045_f27a5@/log.gz
[12:09] <cpaelzer> jamespage: the number of control ports rise each test by one and they fail saying it should be stopped
[12:10] <cpaelzer> jamespage: haven't looked deeper into it
[12:10] <cpaelzer> just saw that checking for my upload in excuses
[15:21] <nacc> jamespage: np!
[15:26] <nacc> dpb1: ping me when you want to review your `git ubuntu build` issue
[15:26] <nacc> *build-source
[15:27] <dpb1> nacc: thx
[15:27] <dpb1> nacc: ya, I need some interactive debugging I think
[15:43] <powersj> ahasenack: samba tests appear to be working as expected now
[15:43] <powersj> https://paste.ubuntu.com/24857332/
[16:25] <teward> powersj is doing my bug triage job for me lol
[16:26] <nacc> teward: yw! :)
[16:26] <powersj> :D
[16:26] <teward> :P
[16:26] <powersj> And I even copied and pasted from your page this time ;)
[16:26] <teward> powersj: thank you again for doing my job :)
[16:26] <teward> powersj: yay!  I appreciate the assistance :)
[16:31] <ahasenack> powersj: the tests swallow stdout somehow? like http://pastebin.ubuntu.com/24857739/
[16:32] <ahasenack> I would expect to see the local workgroup SID somewhere in there
[16:32] <ahasenack> as evidence
[16:32] <powersj> ahasenack: yes it does
[16:32] <nacc> dpb1: note also that `git ubuntu build` is *just* a wrapper around dpkg-buildpackage
[16:32] <nacc> dpb1: it's not often used by end-users
[16:33] <dpb1> I have time now
[16:33] <dpb1> so...
[16:33] <nacc> dpb1: sure
[16:33] <ahasenack> powersj: it's hard to verify that test as it is, it could be a false green as we have seen
[16:33] <dpb1> build-source is what I should have been using
[16:33] <nacc> dpb1: and did (it looks like)
[16:33] <nacc> dpb1: yeah, typically i use build-source followed by an appropriate sbuild
[16:33] <powersj> ahasenack: I would suggest going and looking at the actual tests before going that far
[16:33] <dpb1> hm, right
[16:34] <powersj> I do agree in principle, but knowing how we check for output for the tests I like to think we are ok
[16:34] <dpb1> nacc: so, git ubuntu clone; quilt push -a; git build-source
[16:34] <dpb1> that should get me the latest source package as it would be installed in the dev release
[16:34] <nacc> dpb1: no, you don't want to be in patches-applied state
[16:35] <powersj> Although I haven't thought about keeping all the stdout before, that is one I didn't question
[16:35] <nacc> dpb1: you should be able to just do `git ubuntu clone; git ubuntu build-source`
[16:35] <dpb1> nacc: ah, only when making a modification
[16:35] <dpb1> ok
[16:35] <nacc> dpb1: with a cd in  between, i guess
[16:35] <dpb1> right
[16:35] <nacc> dpb1: alternatively `pull-lp-source <srcpkg>` :)
[16:35] <dpb1> makes sense
[16:35] <dpb1> well
[16:35] <dpb1> I have changes I want to make, but first I want a baseline
[16:35] <nacc> dpb1: those two should produce the same output
[16:36] <dpb1> nacc: fwiw, I'm looking at this bug: https://bugs.launchpad.net/ubuntu/+source/nis/+bug/1658653
[16:37] <ahasenack> powersj: I branched lp:ubuntu-test-cases/server/
[16:37] <nacc> dpb1: your 'baseline' for a source pacakge is obtainable from `pull-lp-source <srcpkg>` as well, is all i meant before
[16:37] <dpb1> nacc: got it
[16:38] <nacc> dpb1: or more correclty, sorry, `pull-lp-source -d <srcpkg>`
[16:38] <ahasenack> powersj: is that the automated version of http://iso.qa.ubuntu.com/qatracker/milestones/351/builds/117343/testcases/1409/results ?
[16:38] <nacc> dpb1: that won't extract the source package, and you can use that as the baseline for generating debdiffs
[16:38] <ahasenack> well, samba-server inside that branch, that is
[16:39] <dpb1> nacc: idk, I like git log, it's helpful to have it right at first
[16:39] <nacc> dpb1: sure that will work too
[16:42] <dpb1> nacc: for the mechanics of actually testing my change.  what do you normally do?  new lxd for artful?
[16:47] <nacc> dpb1: so it happens with artful too (already confirmed)?
[16:48] <nacc> dpb1: yeah, so what i'd do is a) setup a PPA to test (a bit slower) or b) build in sbuild and `lxc file push` to a artful lxd and install there
[16:48] <dpb1> haven't confirmed yet, no.
[16:48] <nacc> dpb1: ok, that's probably step 1 -- you can also just look in d/changelog if something stands out
[16:50] <powersj> ahasenack: the automated ISO tests have a repo where the jenkins jobs are kept and another for the tests
[16:50] <powersj> tests are here https://code.launchpad.net/~ubuntu-server-qa/ubuntu-test-cases/server-tests-raring
[16:51] <powersj> which looks like you found further above (sorry didn't read that far up)
[16:51] <ahasenack> np
[16:51] <ahasenack> they don't include the steps from that manual test page
[16:53] <ahasenack> step 28 seems to be the only one missing
[18:37] <Epx998> Is there a guide on tftp/network booting/installing against the full server iso? so we get all the drivers.
[18:37] <sarnold> Epx998: perhaps investigate what maas is doing?
[18:40] <Epx998> sarnold: maybe one day we'll get hours approved to POC MAAS, but for now we use tftp against the netboot images, we just want an all inclusive image with all the drivers.  Installing against the iso, wondering if there is a how-to or documentation somewhere for it.
[18:42] <nacc> Epx998: rather confused -- you're using the netboot image, which implies you're using a rather minimal env to start (just enough to netboot generally). "All the drivers" in what context do you mean?
[18:45] <Epx998> nacc: my manager has a bad habit of ordering servers for many different manufactors, tired of dealing with driver issues every time we get new hardware.  This time its huawei, video drivers out the box with netboot, no issues if we use the iso.  We have some UB14 servers still, good ol controller issue there.
[18:46] <Epx998> meant also we have some ub12 servers still, that are giving controller issues for the drives.
[20:05] <ahasenack> powersj: hey, how would one change a test like http://iso.qa.ubuntu.com/qatracker/testcases/1409/info? Where is that stored?
[20:05] <ahasenack> I would like to propose a change to step 28, which the automated bits are not even running
[20:05] <powersj> ahasenack: at the top of the page there is a file a bug link, which takes you to the project
[20:06] <ahasenack> aha
[20:06] <ahasenack> thx
[20:06] <nacc> ahasenack: https://code.launchpad.net/~ubuntu-testcase/ubuntu-manual-tests/trunk
[20:06] <nacc> ahasenack: i think?
[20:06] <ahasenack> will check
[20:07] <nacc> ahasenack: http://bazaar.launchpad.net/~ubuntu-testcase/ubuntu-manual-tests/trunk/view/head:/testcases/image/1409_Install%20%28Samba%20server%29
[20:07] <nacc> ahasenack: that's "1409_Install_Samba_server" i think
[20:07] <powersj> that looks right
[20:11] <ahasenack> thanks
[20:12] <ahasenack> powersj: I can add the new step 28 to the automated tests, will look into that later
[20:12] <powersj> sweet
[20:12] <powersj> ahasenack: if there are other bigger changes that would make it easier, please do. I have looked at these tests way too much, and taking a step back to review them would be good.
[20:12] <ahasenack> sure
[20:13] <ahasenack> we can add the simple dep8 tests i added recently too, they are simple, exercise the localhost network only, and can be run right after a default install
[20:13] <ahasenack> https://bugs.launchpad.net/ubuntu-manual-tests/+bug/1697996 ftr, now cooking up a branch
[20:27] <Epx998> Does MAAS have to act as a DHCP server or does it have a built-in tftp service that I can next to?
[20:32] <ahasenack> I think not using the dhcp server that comes with maas will lead to pain. I know it was possible in the 1.8/1.9 versions, not so sure about current 2.2. But even with 1.8/1.9, it wasn't really the recommended way
[20:32] <ahasenack> roaksoax will have a better insight :)
[20:32] <Epx998> all you need with dhcp is to point to a tftp server
[20:33] <Epx998> if maas is running a tftp server internally, makes it easy
[20:33] <ahasenack> tbh, maas nowadays only uses dhcp for that step. The Ip assignment is made by writing to /etc/network/interfaces* directly
[20:34] <ahasenack> commissioning and enlisting steps, I mean. In the actual node deployment you get a "static" IP in e-n-i
[20:34] <ahasenack> but you can still select dhcp if you want
[20:35] <ahasenack> Epx998: I suggest you hop into #maas here on freenode
[20:35] <Epx998> we run isc-dhcp-server on another node, all of our subnets have a dhcp helper to the dhcp serve, that just next's our tftp
[20:35] <Epx998> will do
[20:36] <ahasenack> that channel was active 30min ago
[20:37] <ahasenack> Epx998: did you read https://docs.ubuntu.com/maas/2.1/en/installconfig-network-dhcp ?
[20:38] <Epx998> ahasenack: need to validate maas as a replacement for OS provisioning, we really dont need to replace our dhcp server
[20:39] <ahasenack> sure
[20:39] <Epx998> yeah its running tftpd, i can see port 69 in use - so thast good
[21:27] <hehehe> hey hey
[21:27] <hehehe> :D
[21:56] <keithzg> Hmmm, does anybody know how to fully disable winbind auth in dovecot? I have "auth_use_winbind = no" configued, and `doveconf` confirms that's active, yet I'm constantly seeing pam_winbind being invoked for dovecot:auth in the log.
[21:57] <qman__> not an expert, but it sounds like dovecot is using pam, and pam is using winbind
[21:57] <qman__> so, either configure pam to not use winbind, or configure dovecot to not use pam
[21:58] <keithzg> Unfortunately neither of those options is exactly optimal :(
[21:58] <keithzg> Particularly considering that the server in question also doubles as a fileserver to Windows machines.
[21:59] <qman__> well, that's what VMs/containers are for
[22:00] <keithzg> Yeah, because I'm sure I won't run into *other* issues greater than just spurious log file lines if I complicate things like that ;)
[22:00] <sarnold> keithzg: is there an /etc/pam.d/dovecot?
[22:00] <qman__> separation of duty is important for lots of reasons, and that's one of them
[22:01] <keithzg> sarnold: There is indeed!
[22:01] <sarnold> keithzg: does it use winbind? :)
[22:02] <keithzg> sarnold: Not explicitly, although I should look into if any of the inlcuded options in turn include winbind; just lists common-auth, common-account, and common-session.
[22:03] <keithzg> qman__: Eh, I don't disagree in principle, and it's why most of the internal services are indeed run in VMs. But in this instance, the users being authed also have their home directories on the server, which is in turn where their Maildirs live. And everyone insists on accessing all this at least slightly differently than everyone else...
[22:06] <keithzg> sarnold: Yeah, good catch, all three of those include pam_winbind.so
[22:09] <sarnold> keithzg: now you get to decide if you want to neuter those or just duplicate them directly into dovecot without the line..
[22:09] <ndboost> hey all i am trying to move my mysql datadir off to a block volume on DO
[22:09] <ndboost> and syslog shows these entries
[22:09] <ndboost> Jun 14 22:03:49 app01 kernel: [27231.885231] audit: type=1400 audit(1497477829.157:531): apparmor="DENIED" operation="open" profile="/usr/sbin/mysqld" name="/proc/23000/status" pid=23000 comm="mysqld" requested_mask="r" denied_mask="r" fsuid=112 ouid=112
[22:09] <ndboost> Jun 14 22:03:49 app01 kernel: [27231.885383] audit: type=1400 audit(1497477829.157:532): apparmor="DENIED" operation="open" profile="/usr/sbin/mysqld" name="/sys/devices/system/node/" pid=23000 comm="mysqld" requested_mask="r" denied_mask="r" fsuid=112 ouid=0
[22:09] <ndboost> Jun 14 22:03:50 app01 systemd[1]: Started MySQL Community Server.
[22:09] <ndboost> even though i have an entry in /etc/apparmor.d/tunables/alias pointing to the new path
[22:10] <ndboost> the service comes up fine but ansible still thinks its failing :\
[22:14] <sarnold> ndboost: note that those DENIED messages aren't for data files
[22:14] <sarnold> ndboost: https://bugs.launchpad.net/ubuntu/+source/mysql-5.7/+bug/1658233
[22:14] <ndboost> is it denys for starting mysqld?
[22:14] <ndboost> ahhh
[22:14] <ndboost> "For MySQL the call to proc/pid/status is done as part of a check to ensure no other processes are using the same socket file, so it affects the server's ability to detect an invalid configuration."
[22:14] <ndboost> bingo
[22:15] <dpb1> ahh
[22:15] <ndboost> forgive me for my lack of knowing how to navigate launchpad but is there a workaround?
[22:15] <Aison> ChmEarl, thx for your help, lvm2 works now. Unbelievable that this bug is more than an year old
[22:15] <sarnold> ndboost: here's something to fiddle with https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/1658239
[22:18] <ndboost> thanks sarnold
[22:18] <keithzg> sarnold: Seems to me that merely changing the order in common-auth would solve my problem (so that winbind is last, rather than LDAP being last---LDAP being where the users actually live). I guess I have to give up using pam-auth-update then, though, since I don't see a way with it to specify order of plugins, only which plugins are enabled.
[22:19] <sarnold> keithzg: you're firmly into territory I've never needed to know, but i'm surprised re-orderingw ould fix it
[22:20] <keithzg> sarnold: Yeah, it's not a hard failure, all it is is an annoying set of lines printed in the dovecot logs (in scary red in our colorful systemd world, heh) when winbind fails, but then it moves onto ldap and is placated by successfully authenticating the users.
[22:20] <keithzg> So I'm largely just fixing a cosmetic log issue for myself here, haha
[22:21] <sarnold> keithzg: oh! I get it
[22:21] <sarnold> my world is back together again, thanks :)
[22:21] <keithzg> sarnold: haha, no problem :) And many thanks for being, yet again, The Answerer Of Questions :)
[22:22] <ChmEarl> Aison, cheers
[22:24] <sarnold> :D
[22:56] <Aison> is there some nagios nrpe server 3 backport for xenial?
[22:58] <Aison> the nagios 3 client from zesty is not working correctly with nrpe server from xenial
[22:58] <tarpman> https://packages.ubuntu.com/xenial-backports/nagios-nrpe-server says no
[22:59] <tarpman> Aison: is it https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=849417 ?
[22:59] <nacc> Aison: fwiw, nagios3 is being removed from artful (hopefully)
[22:59] <nacc> Aison: it's already been removed from Debian
[23:01] <Aison> tarpman, yes, it is related to ssl
[23:01] <Aison> nacc, why?
[23:02] <nacc> Aison: it's insecure and inactive upstream (icinga2 is the alternative) and nagios4 is the future (i think)
[23:03] <nacc> Aison: there are unfixed CVEs in nagios3
[23:03] <nacc> LP: #1696252 is the artful bug
[23:03] <Aison> well, nagios4 is an option too ;)  but v4 it is also not in ubuntu (not even zesty)
[23:04] <nacc> Aison: not in debian either :)
[23:22] <hehehe> GET /index.php?route=product/product&manufacturer_id=7&product_id=165 HTTP/1.1" 403 177 "-" "Mozilla/5.0 (compatible; Baiduspider/2.0;
[23:22] <hehehe> baidu spiders are weird
[23:23] <hehehe> he followed aliexpress scraper seems so :D
[23:32] <mwhudson> nacc: thanks for taking care of that kombu / celery mess
[23:32] <tarpman> this channel makes me hungry sometimes