[04:38] Has anyone else had issues with mysql-server on Ubuntu 16.04 where the socket isn't being created? [06:19] someone who knows a answer to this problem : https://askubuntu.com/questions/910562/no-mail-log-when-sending-with-postfix-ubuntu-17-04 [06:48] nobody ? === Malediction_ is now known as Malediction [08:30] someone who can help me figure this out : https://askubuntu.com/questions/910562/no-mail-log-when-sending-with-postfix-ubuntu-17-04 === vamiry_ is now known as vamiry [10:03] jamespage: I saw you abandoned ticket 72 with the new OVS and so much more [10:04] jamespage: big issues or just test done and further uploads will be on a different path [12:22] cpaelzer: no issues; just needed todo a followup on most of those packages for a rebuild for sqlalchemy 1.1.x [13:23] Hi. Is there any fault-tolerant replicated reliable filesystem like GlusterFS and CephFS that works on Ubuntu 16.04 and Ubuntu 14.04 clients _and_ servers? [13:23] how would I figure out which packages to remove to get an already running 16.04 server into a minimal package state? [13:34] let me try a different question, if I run apt-cache showpkg openssh-server and remove all the packages that are NOT listed, will my server still run? [15:02] Kyoku: try debfoster for more targeted pruning. Just removing everything that's not a dependency of openssh-server is not a good idea at all [15:05] thanks seveas, I'm basically trying to create a sub 100MB footprint instance in EC2 which is proving to be a pain, the raw http://cloud-images.ubuntu.com/releases/xenial/release/ubuntu-16.04-server-cloudimg-amd64.tar.gz is around 1GB [15:06] rcj can you point me to the tools / documentation you use to create http://cloud-images.ubuntu.com/releases/xenial/release/ubuntu-16.04-server-cloudimg-amd64.tar.gz please? i want to try to build a tiny image, like the mini iso but in EC2 === JanC_ is now known as JanC [15:56] Kyoku: that'll prove to be quite tricky. Kernel+initramfs is nearly 50mb already [15:56] s/nearly/over/ [15:58] hurricane:~$ du -hc /boot/*-4.10.0-20* | tail -n1 [15:58] 64M total [15:58] that's kernel+initramfs+abi+System.map+efi [16:00] Kyoku: is the a better why to that goal, then "because"? [16:05] nacc yes, mainly because of a project requiring 500-1,000 instances per region so it's an extreme optimization to reduce all types of resource usage and costs, and improve speed of moving things around and firing them up etc. so far i've got the footprint down to 764M with debfoster [16:06] Kyoku: The basic Ubuntu cloud image is produced from https://launchpad.net/livecd-rootfs with the ubuntu-cpc target, but that is difficult to run outside of the launchpad-buildd (lp:launchpad-buildd) environment. [16:06] after disabling all logging and auto updates that might actually be small enough - it's at /dev/xvda1 991M 764M 161M 83% / right now [16:08] Kyoku: wouldn't containers be better than VMs for such a workload? [16:08] And the image is build from the cloud-image seed http://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu.zesty/view/head:/cloud-image (which is a minor addition above ubuntu-server http://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu.zesty/view/head:/server [16:10] nacc i'm not certain about containers, if they can be allocated dedicated ip's then maybe [16:10] thanks rcj [16:11] Kyoku: they definitely can [16:12] do you know if lxd can handle dedicated ip's nacc? [16:12] the external elastic kind [16:12] Kyoku: https://github.com/lxc/lxd/issues/2083 [16:12] Kyoku: maybe? [16:13] Kyoku: i'm not sure how/if it works in an ec2 sense === led2 is now known as led1 === vamiry_ is now known as vamiry === genpaku_ is now known as genpaku === yokel_ is now known as yokel [19:38] hi [19:38] i'm an administrator of a large FTP server, and we now also provide the ubuntu cloud images [19:39] we managed to find an upstream rsync source, but lack a contact address or mailing list [19:39] launchpad also does not offer a way to setup our mirror [19:39] so, could you provide some admin contact? [19:39] right now there's a permission error on the master server, which we're unable to communicate [19:40] C-Otto: contact user for the cloud images? [19:40] Odd_Bloke: --^ ? [19:41] nacc: was that directed to me? [19:41] C-Otto: which server are you trying to sync from ? [19:41] C-Otto: i was clarifying your question, yes, and I was pointing someone who could help if your answer is yes [19:41] rcj: --^ as well, i think? [19:42] nacc: cloud-images.ubuntu.com [19:42] nacc: ok. then yes [19:42] ikonia: sorry, that's for you: cloud-images.ubuntu.com [19:42] C-Otto: ok [19:43] C-Otto: rsync ? [19:43] ikonia: yes [19:43] heh I didn't know those did rsync. neat. [19:43] C-Otto: responds for me [19:43] ikonia: yes, i know - it's a permission problem [19:43] ikonia: besides, i'd like a contact address for future issues, too :) [19:44] C-Otto: what permissions problem, it's syncing fine for me [19:44] ikonia: lost+found cannot be accessed [19:45] thats not a problem [19:45] lost and found shouldn't be synced [19:45] ikonia: you don't have to read my error mails, so I don't think you can decide on that [19:45] that's a filesystem recovery directory [19:45] you should have that in --exclude [19:45] you should not expose that to the internet [19:45] C-Otto: I can decide on that [19:45] you should exclude it [19:46] i'd rather not mirror projects that pose such work on me [19:46] then don't mirror it then [19:46] but before we take a detour, i'm here for a contact address - could you provide it to me? [19:46] if you can't be bothered to type --exclude=lost+found - don't mirror it [19:46] C-Otto: I can't be bothered working with someone who poses such work on me [19:47] ikonia: could you provide a contact address? [19:47] C-Otto: i've pinged the folks that I think can help, they will respond when they can, i hope [19:47] nacc: thank you [19:47] C-Otto: it's detailed in the wiki [19:47] C-Otto: I can't be bothered working with someone who asks me to read the wiki page for him [19:47] ikonia: could you point me to that? my collegue was unable to find it [19:47] C-Otto: no [19:47] hm. [19:47] C-Otto: in the same way you can't be bothered to type --exclude [19:47] I can't be bothered to type the url [19:48] maybe anyone else reading this is in the mood to help? [19:48] you could read it yourself [19:48] that seems to be a key opion [19:48] option [19:48] i don't even know what 'the wiki' is, i'm not an ubuntu user [19:48] you found the IRC channel [19:48] I'm pretty sure you could find the wiki [19:49] ikonia: i feel that your energy could be invested in more helpful responses [19:49] C-Otto: I provide more helpful responses to people with helpful attitudes [19:50] i'm quite helpful, i'd say [19:50] ikonia: I really don't think we have a wiki page about mirroring the cloud images. where is it? [19:50] sarnold: it's the overall mirror maintainers [19:50] ikonia: I don't think that's a safe assumption. [19:51] sarnold: it's a canonical contact detail for the ubuntu mirrors, I'd say that would get the ball moving pretty well [19:58] C-Otto: Hi, I'll try to help you out. I'm part of the team that produces the images which show up on cloud-images.u.c. However, that web service is run by the same folks that run the Ubuntu archive mirrors. That is to say, I can't keep lost+found from showing up, but the folks in #ubuntu-mirrors can. [19:59] C-Otto: and I'll open a ticket internally to see about removing that entirely. [20:02] thank you [22:26] does @hourly in cron run exactly at 1:00, 2:00, etc? [22:27] it should, yes [22:38] rcj: thanks! === blackboxsw is now known as blackboxsw_bl === blackboxsw_bl is now known as blackboxsw_bbl === CodeMouse92 is now known as CodeMouse92__