=== JanC_ is now known as JanC [04:39] good morning [05:23] Good morning [06:45] 223122211221122231 [06:46] Oops [07:05] andol: too much entropy spilling over? [07:05] well it is actually not that good being down to three chars :-) [08:11] coreycb: Hi. Is possible to change in Xenial repository package python-openstackclient from version 2.3.0 to version 2.4.0? There are few issues in version 2.3.0 e.g. openstack network list shows other network UUIDs as neutron net-list [08:11] coreycb: pls let me know if its possible to upgrade [08:32] danpawlik: it is always a nightmare with the client. I ended up running my client in a docker container [08:32] danpawlik: https://zioproto.ninux.org/2016/11/16/docker-to-run-the-openstack-client/ [08:33] this way I do oscli [08:33] and I enter in a env with the right client [08:33] also 2.x is very old [08:40] zioproto: exactly. I do almost the same, but I ust directly python virtualenv [08:41] zioproto: but here comes other problem: someone just install openstack client availalbe in ubuntu repo and just want to use all Openstack services [08:42] that is really a problem. [08:42] and when he have a problem he bother you [08:42] the simplest way is to tell him install package min 2.4.0 and then comes next question: where can I find it xD [12:10] danpawlik: can they use openstackclient on xenial from the cloud archive? [12:11] coreycb: I can, you can, but... customers which are buying servers in my company will use openstackclient available in repo [12:12] we will make a notification on the page that user should use openstackclient >=2.4.0 [12:12] but... maybe it will be better to "upgrade" the package? [12:14] danpawlik: we can't do that unfortunately [12:14] danpawlik: if there are particular patches to backport from 2.4.0, that might be an option [12:15] coreycb: I try to find the patch for it [12:17] cpaelzer: https://code.launchpad.net/~paelzer/ubuntu-seeds/refactor-platform-virt-artful/+merge/327076 looks a bit broken - conflicts. I think maybe you targetted ubuntu.artful instead of platform.artful? [12:36] would anyone be so kind as to try and NTP query us-wa.ntp.dark-net.io please? I think it's up, but I can't tell due to crappy internet. [12:42] teward: It responds to (icmp) ping but not to NTP queries [12:42] I am trying to disable IPv6 DAD in the qrouter network namespaces on the openstack network node. When I edit something in /proc/sys/net/ipv6/conf/all/ is that global for all interfaces in any network namespace ? [12:54] coreycb: https://github.com/openstack/python-openstackclient/commit/8c6b5a087a4b0dca16faffdbcb5fc9f2d424ddf8 [12:54] coreycb: or this one https://github.com/openstack/python-openstackclient/commit/c9cfd569fe2f7c9f499843d77ebbc096b333733b [12:55] coreycb: I just check git log --pretty=oneline 2.3.0..2.4.0 [12:57] coreycb, zioproto: so now is possible to create new package with applied patch :) ? [13:04] coreycb: you mean with a huge patch with all differences between 2.3.0 and 2.4.0 ??? [13:04] but does not make any sense :) [13:08] zioproto: new package for 2.3.0 [13:08] but with patch for neutron region [13:10] danpawlik: that seems reasonable, i'll get that backported [13:10] zioproto: :) [13:11] danpawlik: keep an eye on bug 1570491 to follow status [13:11] bug 1570491 in python-openstackclient "Network client (neutron) initialized without specifying region" [Medium,Fix released] https://launchpad.net/bugs/1570491 [13:13] coreycb: ack. Thx [14:13] andol: crap sounds like it's down then. thanks for checking. i'll stab it further [14:14] andol: can you run an `ntpdate` against the server, it *looks* like it's working now at least from work here. After a quick restart of the thing of course. [14:21] is it advised to use fail2ban to protect ssh? [14:21] DK2: probably wouldn't hurt. [14:31] coreycb: I receive an email from launchpad. nice, big thanks ! [14:40] is dep3changelog authoritative regarding the structure of the DEP3 header? [14:40] i.e., if it fails, the header is incorrect and must be fixed? [14:41] ahasenack: I presume your DEP3 headers are created for patches via `quilt header --dep3 -e` or similar? (Because that's how I get DEP3 compliant headers on my patch files) [14:42] no, there was a dep3 header already, but not compliant [14:42] ah [14:42] should I quilt --dep3 it and complete what's missing then? [14:42] (E: No Context :P) [14:42] *yawns and chugs coffee* [14:43] always a good idea :) [16:10] hi all! not sure if this belongs in windows or ubuntu. I am trying to get a script going that would be able to reset AD password. Im having a hard time getting a ldaps connection, can anyone help? [16:11] nomoney4me: you sure that your AD environment supports ldaps? [16:11] I have read up in ldaps, and it looks like as long as port 636 is open, then ldaps is enabled for AD, is this not correct? [16:14] my ubuntu box can do ldapsearch on the AD just fine. However, in order to do a changetype:modify, it looks like I need to be on ldaps (Or something secure) [16:15] but when I change port to 636, there are some ssl things that I don't understand and not sure how to configure it. [16:45] Any known issues with the installer having a dhcp issue with 10g nics? [16:46] I get a pxe dhcp IP addr, but in the installer - the address is not aquired. [17:25] This is so strange. [17:29] Epx998: can you check dmesg and see if the nic was properly detected? [17:38] ahasenack: it had to have, I was watching my dhcp servers log and saw the link hit right before autoconfig went [17:39] Epx998: wasn't that the bios pxe booting and its dhcp request? [17:39] eth4: NIC Link us Up 10Gbps, Flow Control: RX/TX [17:39] you should see two dhcp requests: from the bios, when pxe booting, and later by the installer [17:39] ahasenack: That was doing the debian-installer [17:39] ok [17:39] ahasenack: just before autoconfig [17:40] i'll run it again to dbl check though [17:41] and tail /var/log/syslog if you can (on the installing system) [17:41] if this fails ill disable the onboard nics incase autoconfig is tripping on eth0 nic being unplugged, though im setting eth4 specificallu in my netboot append options [17:42] dhcp offer from pxe, at the menu - going forward [17:43] ok autoconfig with 10g worked, its failing on eth0 [17:43] which i unplugged, even though i statically set eth4, it still attempts on eth0 [17:44] bloody annoying this is ;P [17:45] if this works, only 350 servers to disable onboard nics for [17:48] would it be easier to move to xenial? [17:48] it feels like the predictable-names nic thingies there might make this more approachable [17:48] and save some poor intern from hours of tedium [17:49] sarnold: Google does not support Xenial builds yet for android. [17:49] for our mobile stuff on android, we build on what google certifies, ub12 and ub14 - soon ub12 will be out and we'll be on ub14. [17:50] Epx998: oh :( pity. I had hoped this would be some in-house software that needed updating ;( [17:50] Epx998: you can't run your build stuff in a chroot or something? [17:51] tarpman: its not how our build system is set up, though we are starting to run smaller builders in docker containers on ub14 [17:51] ok well with the onboards disabled, autoconfig still failed [17:51] ill have to troubleshoot a bit more, meeting now :D [17:52] meeting: "why the hell aren't the servers done yet" [17:52] :) [17:52] don't die, Epx998 [18:27] i think.. eth4 gets renamed to eth1, is that possible? since im setting eth4 in my netboot menu, maybe this causes dhcp autoconfig to fail? [18:28] renames are logged [18:28] search for eth[0-9] perhaps in the logs [18:30] So frustrated. Installed ubuntu 16.04.2 server 64 bit this time :D. Did all my updates and snap install conjure-up --classic and completed. Then went to restart. After reboot I could not login. Luckily I already set a root password. It just kept telling me mine is incorrect. So I log in as root and try to change it and it (passwd jbrazier) it asks twice but no matter what the passwords don't match. So I went to shad [18:31] !!! i got it [18:31] Epx998: I am only a bot, please don't think I'm intelligent :) [18:31] VICTORY [18:31] braziercustoms: irc has line length limits. you were cut off at "went to shad" [18:31] Epx998: yeah?? :D [18:31] yeah, i didnt account for the interface rename with the disabled 4 onboard nics [18:32] BUT i have to add compiling the during the install, since it probably wont be present at a reboot [18:32] compiling the ixgbe driver [18:32] SArnold lol ok [18:33] So I went to shadow and removed password hash. Now I can login but still can't fix my password. Not to mention after I reboot only a few containers started and when I info --showlog on a running container sometimes I get the info sometimes I get connection refused, is lxd running? [18:34] isnt root disabled by default? [18:34] braziercustoms: you got like 4 different issues [18:34] braziercustoms: lets just start with one first [18:34] I'm frustrated is 1 of them this my 8th reinstall [18:35] yes that is frustrating, but good news is you are probably doing it wrong [18:35] braziercustoms: i did probably 40 trouble-shooting my 10gb nic / ixgbe driver issue :D [18:35] iirc you had to reinstall the first time because you put a 32bit os down on the machine [18:36] other than that there shouldnt be any reason to keep reinstalling unless you just have that much time on your hands [18:36] Stokachu I know.. I know.. I'm sorry. [18:36] It only takes 8 min [18:36] so.. lets pick one of the first issues and go through that [18:36] preferably most important to least [18:36] It takes longer to reboot [18:36] Ok [18:39] Stokachu Are you going to determine the order or am I supposed to ask? [18:39] balls in your court [18:40] Ok I guess password issue is pretty important [18:40] ok, so explain what you did and how you tried to resolve it [18:40] and please stop deleting files [18:41] I didn't delete any files. I opened shadow and removed my password hash so I could get back in [18:41] And I backed it up beforehand [18:42] ok, so you can login but the problem is you can't change your password now? [18:42] Yes [18:42] are you trying to change passwd for root or jbrazier [18:43] The passwd command complains that my passwords do not match for jbrazier. I'm scared to try root it still lets me log in as root [18:43] I know they do [18:44] braziercustoms: I bet it's something to do with your modification of /etc/shadow [18:45] No it happened before I did that..but I just tried again and now it's working.. [18:45] you had caps lock on didnt you [18:45] i do it all the time [18:45] :), anyway, next problem please [18:45] No I swear.. me and two other people tried putting in passwords I even tried one two three four [18:46] i'm on a role [18:46] roll* [18:46] 1234 [18:46] 1 is not a roll [18:46] ugh and there it goess [18:46] that's just a 1 [18:46] Lol [18:46] dpb1: in bball thats all you need [18:46] hot hand. [18:46] hah [18:49] Ok let's move along. What's with the containers and the conjure-up.lxd? [18:49] braziercustoms: is the root fs read-only by any chance? [18:50] braziercustoms: does conjure-up.lxc list show all your containers running with IPs? [18:50] Ahasenack no [18:50] 1 [18:51] But this time I ran it I got connection refused [18:51] braziercustoms: do a `watch conjure-up.lxc list` [18:52] i wonder if the daemon is just restarting [18:52] ive seen that happen sometimes with snap upgrades [18:53] 1st time refused 2nd showed a few running [18:53] :( [18:53] Refused [18:55] Ok wait. I did fresh install I o my added ssh but I left the tools option in the install menu enabled.. Would this cause a prob? [18:55] Install I only added [18:57] Curious stokachu each different time it does run it's different.. some will have ip some wont.. [19:00] braziercustoms: yea thats some crazy ish with snap and lxd restarting [19:00] braziercustoms: can you try `sudo snap refresh conjure-up --edge` [19:00] i got some additional fixes in there [19:02] Yeah stokachu. Do I need to stop anything? [19:02] nah this will be a good test [19:02] Ok [19:03] stokachu: conjure-up ships a version of lxd in it, right? [19:03] yea [19:04] It's refreshed [19:05] does conjure-up.lxc list show you anything? [19:06] I put it back on watch.. So far 4X and no refuse [19:06] Looks like they are starting. [19:07] Right now everytime it checks again it says error on one container and each time it moves down the list and the one that said error before is running [19:07] I'm exited already [19:07] ok thats good news [19:07] have to say, d-i installer has a serious thing for eth0/1 [19:08] if your link isnt eth0/1 it gets really upset on netcfg auto [19:08] Btw if I had had caps lock on earlier when I tried to change my password it would have accepted it cuz both passwords would have been in caps but it wouldn't accept anything even 1234 [19:08] braziercustoms: i would just blame dpb1 [19:08] he likes it [19:08] wrong [19:08] I think there are quality standards for user-set passwords [19:08] I mean, I do like it [19:08] hah [19:09] stokachu: oh very handy to know thanks :D [19:09] sarnold: lol [19:09] They are all RUNNING. reboot see if they come back? [19:09] braziercustoms: sure we're high rolling right now [19:09] let it fly [19:10] Lol [19:11] What did --edge do? Use a beta? [19:11] uses what we have in our development branch [19:11] snaps have different channels; some are intended to be stable, some betas, and some right from source control with no promises of any sort. of course each snap is free to use them as they see fit. [19:11] we have a few more things to do before we promote it [19:12] yea so --edge is essentially our master branch on github [19:12] but certain features like that lxd one will make it into a properly candidate->stable channel soon [19:14] https://bugs.launchpad.net/ubuntu/+source/netcfg/+bug/713385 <- my exact issue, been reported since 2011, last update feb of this year as still an issue. oi [19:14] Launchpad bug 713385 in netcfg (Ubuntu) "netcfg/choose_interface=auto fails to find the right interface" [Medium,Confirmed] [19:19] Ok now I'm lost.. After reboot I can't login again.. but I think it's only the first terminal on f1.. I will verify but right now my screen is flooding with conjure-up messages. [19:20] Btw the first terminal cursor is always screwed up. Instead of being at the end of the word login, it's always under the L [19:20] It it usually still works [19:21] But it [19:21] eww [19:22] Is this because of development version? The flooding? [19:22] Can I stop it? [19:36] Ok sarnold. How do I turn off all the logging to stdout? [19:37] stokachu: ^ [19:38] All containers running after reboot. [19:40] Oh thanks dpb1 I did tag the wrong one oops sorry. [21:18] Ok server 16.04.2 still have a login problem. Intermittent authentication failure. Auth log just says authentication failure. If I keep trying to login eventually it lets me. Sometimes I enter username and hit enter and it resets asking for login again. [21:20] resetting the login prompt is normal after a timeout on the password [21:31] still cannot login after an install? === Epx998_ is now known as Epx998 [22:53] Dpb1 it reset as soon as i hit enter after putting in login name. [22:54] Epx998 intermittently. Dpb1 tag me so it alerts my phone. Sometimes I can login sometimes I cant.