[00:00] can someone help me figure out why apt-get --print-uris is not printing the uris [00:04] ponyofdeath: Are you using it like: apt-get update --print-uris or some other way? === Lcawte is now known as Lcawte|Away [00:05] https://bpaste.net/show/00320d77491e [00:05] genii: [00:06] my manpage suggests --print-uris is supposed to work with apt-get update and apt-get source [00:06] yup [00:06] whats funny is that it works fine on another of my servers [00:07] ah.. that was "also works".. [00:08] i wonder if vagrant is somehow supressing the output or something [00:11] Does not work here with install option in any arrangement, like apt-get --print-uris install blah , apt-get install blah --print-uris , apt-get install --print-uris blah [00:12] ( 15.04 ) [00:12] 12.04 server i have it works [00:13] this 12.04 vagrant box no worky [00:14] works on 14.04 too [00:16] ponyofdeath: Actually, I just found that if the package is already installed, it will not get to the point of processing the --print-uris [00:17] yup thats why i use --reinstall [00:17] but i test with packages not installed [00:17] It works when I used it with a package that was not already installed, but not with a package that is [00:18] ( whether reinstall was used or not) [00:24] ponyofdeath: Actually, it works when I do it like: apt-get install --reinstall binutils --print-uris [00:24] i think i figured it out [00:24] Interesting [00:24] vagrant-cachier [00:24] it cache apt stuff [00:24] and aperently messes with print-uris [00:39] hi has Canonical combined Ubuntu server with MapR? I found one reference to it. Haven [00:39] Haven't found anthing else [00:40] as a platform [00:45] I am trying to figure out if its MapR or Apache Hadoop that is combined with Ubuntu-server as a platform. My idea is that since I have a lot of experience with Ubuntu, that I might learn the version that is for the Ubuntu platform as a specialty...thanks in advance [00:47] transhuman_: hopefully useful https://jujucharms.com/big-data [00:47] thanks for that link [03:15] rbasak: so my current idea is to "break" the connection between some of the PHP packages I need so I can at least get one updated, in order to build the other at the right level (and not pull in any PHP5 dependencies) -- does that seem reasonable to you? We can talk about it more in the morning, if you're around [04:37] I just want to say ubuntu server is a turd! [04:59] ofaq: thanks === m1dnight1 is now known as m1dnight_ === dw2 is now known as dw1 [07:40] Hi [07:41] Hi Can any one provide steps to install Openstack HA using autopilot deployment === Lcawte|Away is now known as Lcawte [09:26] Good morning. === Piper-Off is now known as Monthrect [10:46] nacc: so you have some kind of dependency loop? That doesn't surprise me. We had a similar issue with php-json. If so then yes, break the loop by temporarily. Please make a note as we'll need to ask an Ubuntu archive admin to do the same thing when we do it in the archive. [11:24] kickinz1: how is the ntp merge going? Can I assign bug 1512980 to you? [11:25] bug 1512980 in ntp (Ubuntu) "Please enable PPS in the Ubuntu build of ntpd" [Undecided,In progress] https://launchpad.net/bugs/1512980 [11:26] rharper: fyi, bug 1535076. I've merged it with your merge bug. [11:26] bug 1535951 in strongswan (Ubuntu) "duplicate for #1535076 Please merge strongswan 5.3.5-1 (main) from Debian unstable (main)" [Undecided,Confirmed] https://launchpad.net/bugs/1535951 === cpaelzer is now known as cpaelzer_afk [11:49] rbasak, I've applied the old deltas, forgotten ones, and missed one, I'm now verifying that they are all necessary. Yes you can assign to me. [11:58] OK, thanks. === cpaelzer_afk is now known as cpaelzer === Lcawte is now known as Lcawte|Away [12:51] <|TheWolf|> Hi! [12:52] <|TheWolf|> I just tried to join #ubuntu, but apparently I'm banned :-O I have no idea why, is there any way to find out and/or get me unbanned from there...? [12:52] #ubuntu-ops [12:52] <|TheWolf|> thx! [12:52] I think [12:52] !banned [12:52] If you have been banned it is probably because you have not gone along with what is acceptable !behaviour. If you're not sure what acceptable !behaviour is please see !Etiquette and http://wiki.ubuntu.com/IRC/Guidelines - If you think the ban was a mistake, please join #ubuntu-ops [13:03] <|TheWolf|away> What I'm actually here for: Am I right in assuming that the patches for CVE-2016-0728 have not been released yet? It's a bit confusing because https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1534887 says that it's fixed in 3.13.0-76.120, and that the fix is released, but the current version according to http://packages.ubuntu.com/search?suite=trusty&searchon=names&keywords=linux-generic is 3.13.0.76.82 [13:03] Launchpad bug 1534887 in linux (Ubuntu Xenial) "CVE-2016-0728" [High,Incomplete] [13:03] |TheWolf|away, yes, patched 12 hrs ago === |TheWolf|away is now known as |TheWolf| [13:04] maybe 16 hrs.. but the fix is certainly out [13:07] <|TheWolf|away> OerHeks : and this patch should already be available via the package managers? [13:08] just update, or to be sure, apt-dist-upgrade [13:08] err apt dist-upgrade === |TheWolf|away is now known as |TheWolf| [13:13] <|TheWolf|> that's what I did, but I only get the older 3.19.0.47.32, not 3.19.0-47.53, which includes the patch. might that be a caching problem with the mirror of my hosting provider? [13:14] You ran apt-get update first, I presume [13:14] <|TheWolf|> yes [13:14] Where do your sources point to? It might be that your local mirror is slightly out of date. [13:15] maybe your mirror is not synced yet [13:15] try the main [13:16] <|TheWolf|> according to the output of apt-get update, it fetches from both http://mirror.hetzner.de (the mirror of the hosting provider) and http://de.archive.ubuntu.com and http://security.ubuntu.com [13:17] <|TheWolf|> I'll try and check if the mirror is the problem, thanks for your help so far! [13:20] De.archive.ubuntu.com-archive http 128 Kbps One day behind [13:20] https://launchpad.net/ubuntu/+archivemirrors [13:34] <|TheWolf|> I've tried a number of other mirrors now (which are supposed to be up to date and verified in the last hours), but none of them have the latest version. I guess I'll just have to wait a couple of hours [13:36] figured I would ask again for clarification is apache hadoop the preferred and supported version of hadoop on Ubuntu server or is it one of the other three MapR, Cloudera, or Hortonworks, I know canonical has partnered with one of them just not sure which [13:36] rbasak: thanks; in the request, IKEv2 frag, I'll need to confirm, but I don't think that fix itself is in 5.3.5, rather a bit newer than 5.3.5; we could look to also cherry pick that in if it's going to help users out [13:37] rharper: I'm a little skeptical about cherry-picking anything in a security-sensitive package. Maybe we should unmerge that bug then? [13:37] it's not security [13:37] it's a feature [13:38] It's a security-sensitive feature though, isn't it? [13:38] The whole of IKE is security sensitive. [13:38] I don't feel qualified to say yes or no [13:40] The danger is that either we introduce a security issue, or the security team have trouble patching a future security issue because the later fix conflicts (either literally or logically) with our cherry pick. [13:40] looks like, there's already some IKEv2 frag support in 5.3.5 [13:40] I was looking at the changes since 5.3.5 was tagged, and there are some additional updates, but the bulk of the feature was introduced in 5.2.1 it appears from strongswan/NEWS [13:40] so sounds like the user should have something to work with [13:41] so, no need to cherry pick anything newer at this point [13:41] post merge, we can look at individual bugs [13:41] OK, great! [13:42] while I have you; I've got a 5.3.5 with Ubuntu changes in it; it installs fine, so going to look for testers and then work on re-assembling the ubuntu changelog and a debdiff to pick over; === devil is now known as Guest77896 === DavidDuffey is now known as dduffey === Opcode90- is now known as Opcode90 === SierraAR is now known as Sierra === beisner- is now known as beisner === AndrewMC` is now known as AndrewMC === stevenroose|BNC is now known as stevenroose === spicypixel_ is now known as spicypixel === frediz_ is now known as frediz === erisrenee_ is now known as erisrenee === remix_auei is now known as remix_tj === pitastrudl_ is now known as pitastrudl [14:27] how to check the number of users present on a server ? [14:27] hi [14:28] sertyui: w, who, users [14:37] thanks [14:37] now users give me 3 times my username [14:37] what that it means ? [14:41] Anything you recommend for centralized update management? Something that for example can pull in multiple apt repositories and then allows updates to be picked and deployed [14:45] Landscape is Canonical's commercial answer. [14:49] We're a bit reluctant of Landscape as its very expensive [14:54] in that case you one brilliant solution which not at all expensive [14:54] i mean opensource [14:58] well how to check the number of user on my system ? [15:23] sertyui: I've told you. w and who will both list where those users are coming from. Each (v)tty session is actually a separate login, so thats why a user may show up twice. [15:23] you can awk|uniq to sort it [15:29] and pts means ? [15:30] pseudoterminal [15:31] means [15:31] putty ? [15:32] or kvm ? [15:34] no, just any terminal created that isn't one of the ctrl-F1-like terminals [15:37] i don't understand [15:37] on virtuall server [15:37] when you see pts1 what it means ? [15:39] sertyui: it is a terminal created by ssh or perhaps by screen or tmux. [15:39] ok [15:39] fine === frickler_ is now known as frickler [15:40] i got a service called assp on my server [15:40] sometimes suddently the service get stopped [15:40] i m using server : Ubuntu 14.04.3 LTS \n \l [15:41] so my question [15:42] is : on the server how i can identify what is the root cause of that incident ? [15:55] rbasak: yep, it's a build-depends immediate cycle between phpab and phpunit, which prevents me from using the updated pkg-php-tools (which pulls in php7.0 rather than php5.x) [15:55] rbasak: will work on that today === dannf` is now known as dannf [16:23] If you admin users a linux machine, and some of them are developer and needs to edit some files that is related to service - let's say tomcat - and its directories only edited by a user called "tomcat" ... and you need to give him access, what is the best practice to do that ... give him a limited sudo or set access list for the tomcat directory or create a group and add him to that group and change the group of the tomcat directory then ch [16:23] ange the permission to rwx? [16:23] Good evening everyone, I'm in need of some help with the management of Juju and MaaS, can anyone help me? [16:23] eahmedshendy: use ACLs [16:24] TJ-: why, can you give a short explanation when I choose between the three options? [16:25] My problem consists in the Juju bootstrap, because it tells me when bootstrapping that there aren't avaiable tools. When I use sync-tools to download them, it tells me it can't because the environment is not bootstrapped [16:25] I'm in a sort of dead lock I can't figure out [16:26] (I'm following the MaaS setup documentation, from https://maas.ubuntu.com/docs1.8/juju-quick-start.html) [16:28] ...Anyone? [16:28] eahmedshendy: you could put the developer in a group that has write access to the directory entries/files, or you could leave default permissions as-is and use 'setfacl' (1) to add additional fine-grained permissions for users/groups [16:29] TJ-: I prefer this ACL way ... but I wanted to learn if there is a reasons that makes me select one more than the other in my case [16:30] eahmedshendy: It generally depends on whether the underlying file-system can support ACLs, and if so, if the fine-grained permissions it provides are what is required. [16:33] TJ-: got it, thank you [16:34] eahmedshendy: the old addage of "Keep It Simple, Stupid" applies - in other words don't over-complicate the solution unless you have to [16:35] TJ-: mmm, sorry I didn't get it .. I still can't difference between it gets complicated or not [16:37] eahmedshendy: 'complicated' is for you, the admin, to judge. If you can make do with just using user/group membership and permissions then ACLs would over-complicate it [16:38] TJ-: got it, for me I need to get much practice on ACL use :) [16:39] eahmedshendy: yes. "man setfacl" "man getfacl" and "man acl" will help understand it [16:39] TJ-: Thank you for your help [16:39] :) [16:42] So... Has anybody any idea on how to help me? [16:42] In my opinion, ACLs add too much complexity, I'd either add the users to the tomcat group or create a new group and grant access via limited sudo or similar [16:42] depends on how many files and the details [16:43] I build things based on how easy they are to figure out when you don't know they exist [16:44] so when you have to come back to it five years later, you can tell how it's set up [16:58] Is there really no one who could help me? [17:00] ServantGrunt: looks like the key -server folks aren't around right now. There's usually a good complement of them [17:00] isn't server meeting on wed? [17:01] probably busy with that [17:03] ServantGrunt: I suggest you ask your question, rather than waiting for someone to volunteer to answer a question they don't know yet. [17:03] ServantGrunt: also you could try the #maas or #juju channels. [17:04] eahmedshendy: I don't think it's appropriate to allow admins to edit files directly on a server except in an emergency. Otherwise you lose the ability to roll back or recover from a disaster. Get them to commit changes they want somewhere else, and arrange a process to get it live. === Monthrect is now known as Piper-Off [17:18] rbasak: good point [17:26] I quitted before by mistake so I'm gonna repeat my question just to be sure: I need help about juju and ubuntu server with maas. When I juju bootstrap it tells me there are no tools available. When I juju sync-tools it tells me it needs the environment to be bootstrapped. That's a dead lock I can't figure out, what should I do? [17:33] ServantGrunt: I know there's a "juju bootstrap --upload-tools" option, or at least there used to be. But it depends on your environment I think. Try asking in #juju. [17:33] ServantGrunt: have you been following https://maas.ubuntu.com/docs/juju-quick-start.html === cpaelzer_ is now known as cpaelzer_afk === Piper-Off is now known as Monthrect [17:51] hallyn, you around? [17:53] I'm running into an issue migrating VMs from P->T. should have been fixed by https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1374612 [17:53] Launchpad bug 1374612 in qemu (Ubuntu Trusty) "[FFE] add pc-1.0-precise machine type" [High,Fix released] [17:54] seems that I have two pc-1.0-qemu-kvm machine types: http://pastebin.com/StVfrj9F [17:55] and here's the exact error I'm seeing: http://pastebin.com/YvCsY4YW [18:09] current machine type on precise is pc-1.0-qemu-kvm === lifeless_ is now known as lifeless [18:14] swat30: hm [18:15] swat30: can you pb the full vm xml? [18:15] sure, mind if I DM ? === realcliluw is now known as cliluw [18:17] swat30: you can try, not sure how well my end will take it [18:17] just meant the link === Chrisfu- is now known as Chrisfu === funkyHat_ is now known as funkyHat === spicypixel_ is now known as spicypixel === stevenroose|BNC is now known as stevenroose === dasjoe_ is now known as dasjoe === Exagone314 is now known as Exagone313 === med_ is now known as Guest65103 === fidothe_ is now known as fidothe === marlinc_ is now known as marlinc === Azelphur_ is now known as Azelphur [18:42] Yes TJ- I've been following the guide [18:42] And in the guide it makes it easy, but if I follow the commands execution as the guide does, I get those errors, solving nothing [18:43] rbasak I've tried using that option but I couldn't solve anything either. [18:46] ServantGrunt: do you have firewall rules that might prevent communication with canonical servers? iirc the bootstrapping process pulls simplestreams data from canonical to describe the available images, juju tools, etc. and uploads those into your cloud's storage services thingy.. not sure how it works when using juju to bootstrap the cloud itself, but presumably maas provides the storage.. === jelly-home is now known as jelly === Lcawte|Away is now known as Lcawte === cpaelzer_afk is now known as cpaelzer_ [19:43] sarnold I'm pretty sure no firewall is blocking comunications between my server and canonical === tomaw is now known as 02HAAAAAD === 02HAAAAAD is now known as tomaw [20:12] coreycb` o/ - ok, pushed oslo.messaging 1.4.1-0ubuntu1.2~cloud1 package from juju-staging to juno-proposed in the cloud archive. [20:12] beisner, thanks [20:13] coreycb`, yw, thx sir === bekks_ is now known as bekks === Monthrect is now known as Piper-Off [23:54] hmm, I got this kingston ssd here that I cant write to [23:54] can read the files off it just fine, === Guest77896 is now known as devil_ [23:54] which filesystem are you using? [23:54] tried dd it, no effect after dd finishes the partitions show right up [23:55] it has ext4 on it [23:55] But you arent try to use dd while the filesystem is mounted, are you? [23:55] *trying [23:56] nope [23:56] are there any errors in dmesg? [23:56] just poped the drive into a workstation that dualboots ubuntu and windows, tried formating it in both [23:56] the filesystem just comes right back each time [23:57] How do you see that it is coming back? [23:57] ls [23:58] ls on what? [23:59] lol