=== pleia2_ is now known as pleia2 === adam_g` is now known as adam_g [01:04] rharper`: all tests pass with that qemu merge, so pushing. === Bray90820_ is now known as bray90820 === bigjools_ is now known as bigjools === jelly-home is now known as jelly === zopsi_ is now known as zopsi === rharper` is now known as rharper === sarnold_ is now known as sarnold === athairus is now known as afkthairus === remix_auei is now known as remix_tj === iberezovskiy|off is now known as iberezovskiy [09:42] What port should be opened to do sudo git clone? [09:46] Yuri4_: from the client? [09:46] RoyK^, from terminal [09:46] and why sudo git? git works well as a normal user [09:46] Yuri4_: also depends on the url format you clone, could be http, https, ssh - all different ports [09:46] sudo git clone git@github.com:TelescopeJS/Telescope.git [09:46] Permission denied (publickey). [09:47] don't use sudo with that [09:47] and it is not a port issue [09:47] and you probably need just the right key [09:47] What key? [09:47] Yuri4_: is this your repo? [09:48] RoyK^, no it is a public repo that guide says to copy [09:48] Yuri4_: git clone https://github.com/TelescopeJS/Telescope.git [09:48] Like "step 2 Clone this repository locally:" [09:48] RoyK^, thank you! [09:49] You are supposed to add your nick at nick@github.com:blabla ? [09:49] you only need to login to git if you're member of the project and submitting code [09:49] Yuri4_: as RoyK^ said you don't need user for just clone and can use https, other wise if you need authentication e.g. to submit, this is about your issue https://help.github.com/articles/error-permission-denied-publickey/ [09:50] cpaelzer, RoyK^ thank you === RoyK^ is now known as RoyK === _degorenko|afk is now known as degorenko [10:54] I edited crontab to run a python script every minute: "* * * * * python /home/shay/scripts/my_script.py" [10:54] If I run the very same command ("python ...") directly, it works. I can see under /var/log/syslog the CMD executed every minute, [10:54] I even tried to output (>>) to a file, and the file gets created, and remains empty. Eventually the script doesn't get executed which is my problem. any idea what else I can check? [10:54] var sys log shows this every minute: Jun 16 13:21:01 elk CRON[2903]: (shay) CMD (python /home/shay/scripts/my_script.py) [10:55] You are not using a full path to the python binary. [10:55] i tried that [10:55] i had /usr/bin/python [10:55] same problem [10:56] You need that, in every case. Cron does not use the users environment. [10:56] And all paths inside your script need to be absolute as well. [10:57] oh, so doing something like open('filehere.txt') is not good? [10:57] because i do that often in my script [10:58] Norbin: just chmod +x my_script.py and add #!/usr/bin/python in the first line of the script [10:58] Norbin: open('/full/path/to/file.txt') [11:00] Norbin: does your script ensure its current working directory is as expected before using relative paths? [11:04] nope it does not, fixing that right now [11:07] works, did both the absolute path and the chmod [11:07] cheers [12:17] with regards to dovecot I have one question. If it's using mbox format, and the mbox doesn't exist yet in /var/mail/${USERNAME} but INBOX is configured to be in that directory, will Dovecot attempt to create the mailbox for INBOX before having mail delivered to it? [12:19] teward: it will, and don't use mbox [12:20] teward: use maildir - it scales [12:20] mbox sucks [12:20] RoyK: yes it does. [12:20] it's like 30 years old [12:21] RoyK: existing setup uses Postfix to receive and deliver mail to mailboxes, will that work with dovecot in maildir format? [12:22] RoyK: remember, I inherited this migration project from someone else :P [12:22] the less I have to 'switch over' in the short term the better [12:24] old distro too? [12:26] mhm [12:26] new issue: can't get mail box data from old -> new without new locking when rsync maxes the CPUs [12:26] s/locking/locking up/ [12:26] so...... [12:27] configs are already moved. [12:27] locking up? [12:27] RoyK: CPU softlock [12:27] I/O wait? [12:27] oh [12:27] CPUs peg at 100% on all 4, and seizes up [12:27] bad kernel, perhaps [12:27] RoyK: well, i reapplied the latest kernel now [12:27] so we'll see this evening when I rsync data over [12:27] don't we all love that sort of jobs [12:27] but if, with the mbox config, we can just "fire it up" with empty INBOXes that may solve things [12:27] it's for a home-grown listserv solution so it was never 'built' to scale it seems [12:28] and it involves Linux / MSSQL stuff [12:28] so... *shivers* [12:28] RoyK: but if mbox will recreate the INBOX if it doesn't exist, I can just say "SCREWIT!" and switch to the new server, keeping a backup of the old mbox files [12:28] :P [12:29] teward: if the customer/boss/something allows for that, fine [12:29] there's a meeting today :P [12:29] but yes. [12:29] short of a drive-to-drive sync via disk IO instead of network/rsync/scp i have no other options [12:29] with USB passthrough'd drives and what not *shivers* [12:31] (VMware house, otherwise we'd be using qemu or something) === lamont` is now known as lamont [12:56] I have some ssl keys I need for an app. is there a special place where they are usually saved? [12:57] web servers often store theirs in /etc/ssl/private -- I don't know if that overlap would be beneficial or harmful for your application, but it's a starting point [13:00] sarnold, kk thanks! === dannf` is now known as dannf [13:51] Hey guys, I would like to lear how you create Ubuntu 16.04 AMI at cannonical, since is not possible to the import-image api call. The latest supported version is 15.10 [13:51] thanks [13:54] fernando_: have you found this yet? https://cloud-images.ubuntu.com/locator/ [13:55] Yes sarnold, thanks btw. But I want to have custom partitions with LVM [13:56] So I'd like to know how cannonical guys build their own AMIs for Ubuntu 16 [13:56] Odd_Bloke: ^ [13:56] Odd_Bloke: worth an FAQ entry somewhere? [13:58] RoyK: barring the evil method of indexing that we have which requires reaching into one of the inboxes temporarily, is there a way to completely trim old messages from mbox-format storage? Assuming a conversion to maildir is in "phase two" of the migration. [14:00] RoyK: (the evil indexing requires us to keep the messages around, so 'clean' mboxes for all the accounts is a no-go but we need to have only one of the inboxes remain intact) [14:02] fernando_: The way we build images is, unfortunately, pretty obscure (and depends on some services that aren't publicly available); I can point you at the scripts, but you may be better served by another way of producing an image. [14:02] fernando_: (We do things the way we do because we want to produce a very generic image that makes few assumptions about what users will do with it; you are probably not constrained in the same way) [14:02] fernando_: Why is the import-image API call unusable? [14:05] Odd_Bloke: import-image is the API call [14:05] AWS documentation says Ubuntu 16 is still unsupported [14:05] beisner, qemu 1:2.2+dfsg-5expubuntu9.7~cloud5 is ready to promote to kilo-updates when you get a chance [14:05] fernando_: Oh? Can you give me a link to that? [14:06] What I'm thinking is to mounting ebs volume, use qemu-nbd to mount vhd and then dd to the clear EBS [14:06] does that make sense? [14:07] http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/VMImportPrerequisites.html#vmimport-operating-systems [14:07] fernando_: A VHD which you've created locally with your LVM configuration? [14:07] yes [14:07] with VirtualBox [14:07] teward: just convert it to maildir [14:08] beisner, also everything in mitaka-staging is ready to promote to mitaka-proposed [14:08] fernando_: I expect that is just out-of-date documentation; for example, Ubuntu 15.04 is out of its support lifetime. [14:08] fernando_: (As are 12.10, 13.04, and 13.10, by several years) [14:08] hah, out of that list only two are still supported [14:09] ups sorry guys [14:09] bad link [14:09] let me update it [14:10] http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/VMImportPrerequisites.html [14:10] and in fact [14:10] the api call returns [14:10] fernando_: That's the same link. :) [14:10] "Unsupported kernel 4.4" [14:10] Ahh, interesting. [14:11] fernando_: What type of image are you trying to import it as? [14:11] fernando_: (I've never used this command, apologies if my questions don't make sense :) [14:11] Is the right link, isn't it? "Ubuntu 12.04, 12.10, 13.04, 13.10, 14.04, 14.10, 15.04" [14:12] of that list only 12.04 and 14.04 are still supported [14:13] I converted vmdk to ova and then did the call [14:13] fernando_: Would you be able to pastebin the command that you ran? [14:15] esta bueno [14:15] oops sry [14:15] Odd_Bloke [14:15] yes sure [14:20] Odd_Bloke: http://pastebin.com/RdjfkjM2 [14:21] no error? [14:21] RoyK: okay... how do I do that when I cant get it to the new system [14:22] again, stuck with the E:NoRSYNC issue [14:22] (CPU softlockup) [14:22] call vmware and ask em why it falls over? [14:24] teward: just google it [14:24] teward: that sucks - something wrong with the hardware? [14:25] i'm assuming kernel, I was able to replicate the issue on a Wily box on completely different hardware [14:25] teward: pull out the disk and put it into something else (like a shredder) [14:25] * teward shrugs [14:25] heh [14:25] I'll fix it heh [14:25] just gotta salvage that one mailbox :P [14:25] teward: got a backup of the data? [14:26] teward: last time I moved a mail server I just backed it up and restored from backup (using bareos, a bacula fork) [14:26] RoyK: if we do it's inside the copy of the VM [14:26] and not 'standalone backup' [14:26] i didn't set up the environment heh [14:27] well, it's a VM, setup a new one and attach the disk to that and rsync it [14:29] fernando_: Thanks, and what error do you see? [14:30] RoyK: again, not my call. If i had eternal god power here, I'd create a separate VMdisk, format it the filesystem I need, put the data there, detach it, put it on the new VM that already exists, copy data that way [14:30] then tell the gods to do that [14:31] it's not like it takes a lot of time [14:31] yep [14:31] probably a Monday task [14:32] :P [14:32] * teward has a long weekend coming up, if he can push that to Monday he will :P [14:32] you taking tomorrow off [14:33] teward: it'd take me about five minutes :P [14:36] Odd_Bloke: Unsupported kernel version 4.4... [14:36] which does not happen with trusty [14:36] RoyK: yes. It would. :P [14:36] if I had that power :) [14:36] oops i need to update these VMs... eheheheheheh [14:37] teward: heh - I work with vmware - it's not hard ;) [14:37] RoyK: so do I - my two hypervisors at my apartment (HAIL GODMODE!) are VMware [14:37] it's a VMware house here [14:37] but i don't have the VMware access [14:37] the one who does? [14:37] out until tomorrow [14:37] and I'm not here tomorrow :) [14:37] fernando_: If you could paste the entire command and output together, it would make it much easier to point our EC2 contact at the issue. :) [14:37] (here at work, not necessarily 'not here on the internet') [14:38] teward: you could ask the admin to give you access to your vm, though [14:38] Ok, I will run it again :) [14:38] fernando_: Thanks. :) [14:38] teward: if things are setup with SSO correctly, it *really* shouldn't be a problem [14:38] RoyK: we can discuss system setup another dat [14:38] day* [14:38] outside my purview to comment on how things're set up :p [14:38] that said... I love my own to Vmware hypervisors :) [14:40] we have 15ish nodes in our clusters [14:40] works rather well [14:41] even when one guy (that was me, btw) managed to pull out the wrong blade once (just started working there) - the VMs just popped up on other nodes in the cluster [14:43] wow, nice [14:43] Odd_Bloke, takes some time, hang on a second. Thanks, btw :) [14:44] fernando_: Sure, no rush. :) [14:45] * RoyK was sweating a bit right then - wanted to hard reboot a blade that was troublesome and just yanked it and had a ping running at it which just continued and found that - oops - wrong blade centre - called my boss .... that was just two months after I started working there [14:45] coreycb, qemu pushed to uca kilo-updates re: bug 1587039 [14:45] bug 1587039 in Ubuntu Cloud Archive kilo "aio: strengthen memory barriers for bottom half scheduling" [Undecided,Fix released] https://launchpad.net/bugs/1587039 [14:51] Odd_Bloke: http://pastebin.com/mTFpeXw1 [14:54] fernando_: Brilliant, thanks! [14:54] thanks to you :) [15:03] beisner, thanks and I see the packages moving to mitaka-proposed \o/ [15:06] coreycb, * @ mitaka-staging (http://pastebin.ubuntu.com/17397848/) pushed to mitaka-proposed in uca [15:07] beisner, awesome [15:27] jamespage: I think I already reported this issue but in case I missed it, when running tempest/master and your newton repo: http://logs.openstack.org/88/330188/3/check/gate-puppet-openstack-integration-3-scenario002-tempest-ubuntu-xenial/d5cefca/console.html#_2016-06-16_13_33_34_856 [15:28] I'm trying to get either icinga or icinga2-classicui working in Xenial. Under both, the menu on the left displays for a moment and then slides up and is invisible. Anyone seen this? [15:31] Apocope: no - perhaps try #icinga - haven't used it much for years... moving to zabbix now === pitastrudl_ is now known as pitastrudl [15:35] RoyK: Yeah, maybe. I'm suspicious that it's something to do with the packaging, it's not really anything to do with the functioning itself. I have it running fine on 12.04. [15:36] really, no iea [15:36] idea [15:36] Apocope: i can try and help you out, otp right now [15:36] Apocope: i don't know much about either, but i am probably responsible (php7 migration) for any breakage :) [15:37] oh - php changes [15:37] those usually break things [15:44] RoyK: :) [15:45] nacc: Ok, I'll poke around a bit.. [16:06] Apocope: spinning up a container [16:07] nacc: Ok, in icinga, it's the menu.html that's displaying the unwanted behaviour. I copied the menu from the 12.04 install, it does the same thing. In fact, it's exactally the same between the distributions. [16:07] Apocope: right, it's probably some underlying thing -- what's the URL normally? I know nothing about icinga :) [16:08] nacc: It's $SERVER/icinga/menu.html [16:09] Apocope: thanks [16:09] Apocope: give me about 5 minutes to get setup [16:09] Apocope: if i can reproduce it, i'll probably ask you to file a bug :) [16:15] Apocope: what's the admin username? [16:15] Apocope: ah, icingaadmin, of course :-P [16:16] Apocope: reproduced [16:16] (with icinga) [16:16] nacc: Glad it's not just me. [16:16] Apocope: can you file a bug please? [16:16] Apocope: i'll try and figure out how to debug :) [16:17] nacc: Ok. [16:17] Apocope: feel free to subscribe me, as well (same lp username as my nick) [16:17] Apocope: have you tried 14.04, btw? [16:18] nacc: Yes. I believe it works properly there. Let me verify. [16:18] Apocope: ack, i'm checkign too [16:19] Apocope: also, thanks for bringing this to our (my) attention ... [16:19] Hey all, I'm looking for some help with a new Landscape (on premises) install [16:20] (also new to IRC, so sorry in advance for anything annoying I do!) [16:23] So, Landscape appears to be working fine, (for example I can update packages), but when I click on one of the alerts, I get a System Error, an apology, and very long OOPS ID. Any thoughts? [16:33] nacc: I've filed a bug and subscribed you. Thanks. [16:34] Apocope: thanks, trusty does seem to be fine [16:47] Apocope: hrm, the contents of icinga-cgi are quite different [16:47] i wonder if it's a js issue [16:47] nacc: I wouldn't be suprized. [16:48] Apocope: digging into it, like i said; probably not my highest priority, but will do my best [16:48] Apocope: unfortunately nothing is throwing any errors, so it's tricky to debug :) [16:48] nacc: I understand exactally how it goes. [16:49] Apocope: in chrome, the developer console indicates the size of the menu is 1073 x 0 :) [16:50] That doesn't seem right. [16:50] nope, and it's a computed value, i think ... i wonder how that's happening === iberezovskiy is now known as iberezovskiy|off === sivir_ is now known as sesevi === afkthairus is now known as athairus [19:16] coreycb: good afternoon! === iberezovskiy|off is now known as iberezovskiy [19:17] coreycb: we're facing issues with OpenStack packages that try to start services when running apt-get. It's a problem if you deploy your API in WSGI with Apache for example (Nova, Ironic, etc). [19:17] you need to stop nova-api service before starting apache otherwise you have binding issues [19:18] of course we can do that with automation tools but I thought this feedback would help you to understand why it's not helping to start services automatically [19:41] EmilienM, didn't this discussion already occur? [19:41] coreycb: indeed [19:47] I have a Dell Poweredge 850 (kinda old, 4 years-ish I think) that I ran the 16.04 installer on, and after the reboot it goes past POST and then to a solid black screen (At 1280x1024) and does some hard drive access but never goes anywhere else. [19:48] I'm trying to figure out what is happening to it, but it doesn't do any network traffic, let alone start SSH, and I get not keyboard response [19:48] I have also tried to login without a monitor to confirm I'm doing it right, and just reboot. This doesn't seem to do anything either [19:49] No errors from the grub install, and it is going to the right hard drive when it does so [19:49] Any ideas to point me at? === iberezovskiy is now known as iberezovskiy|off [20:19] it seems like LP: #1544296 is marked as affecting the wrong package [20:19] Launchpad bug 1544296 in symfony (Ubuntu) "IPv6 autoconfiguration adds default route to Neutron tenant networks on qbr- interfaces" [Undecided,Confirmed] https://launchpad.net/bugs/1544296 [20:34] !nomodeset | TodPunk [20:34] TodPunk: A common kernel (boot)parameter is nomodeset, which is needed for some graphic cards that otherwise boot into a black screen or show corrupted splash screen. See http://ubuntuforums.org/showthread.php?t=1613132 on how to use this parameter [20:36] I shall try that. I saw that before and it seemed unrelated, but I presume we'll find out === tinoco is now known as tinocoff === spammy is now known as Guest95181 === Guest95181 is now known as spammy === esde_ is now known as esde