[00:45] <tomreyn> tarvid: it's been the 'default' location on ubuntu (and debian) for the past 10 or 15 years
[00:46] <tomreyn> you can use file system permissions + sftp chroot'ing to restrict things, same as in #/home basically
[03:43] <RoyK> tomreyn: somewhere.tld/~user != somewhere.tld/user
[03:43] <RoyK> tomreyn: for the former to work, you just need the userdir module
[03:46] <RoyK> tarvid: just try to enable userdir
[09:53] <jonah> Hi can anyone help. Each evening my server's mysql crashes and fills the log file with 2016-10-23T09:20:42.318890Z 0 [ERROR] Error in accept: Socket operation on non-socket
[09:53] <jonah> which in turn causes services to go offline
[09:53] <jonah> Does anyone know a fix? I'm on Ubuntu 16.04 LTS server
[10:12] <PCdude> I have installed openstack with conjure-up today. When I rebooted the system it is no longer active and commands like "lxc list" just hang and dont give any info
[10:12] <PCdude> How can I solve this?
[10:12] <PCdude> stokachu: maybe you have an idea?
[10:34] <tomreyn> jonah: https://www.percona.com/forums/questions-discussions/mysql-and-percona-server/44502-error-in-accept-socket-operation-on-non-socket suggests it is possible to reproduce this on percona mysql server 5.7.12-5 on debian by, as run on the server itself, 'telnet localhost 3306'. Can you?
[10:37] <tomreyn> if so, i'd suggest posting a bug report including your mysql configuration ('ubuntu-bug mysql-server' *may* do so for you) and ubuntu version. as a workaround, downgrading to mysql 5.6 may be an option (since these reports all refer to 5.7),
[10:47] <tomreyn> another workaround *may* be to only connect to the server using TCP connections (to port 3306) and to set the 'socket' option to a location where it will not be expected, preventing socket access to the server. but this is likely more error prone and complex to configure than a downgrade.
[10:50] <tomreyn> before doing anything at all, use mysqltuner or mysql tuning primer or similar to review your current configuration.
[16:28] <LostSoul> Hi
[16:28] <LostSoul> Anyone around?
[16:32] <RoyK> LostSoul: there's usually someone around, yes
[16:32] <RoyK> !ask
[16:32] <hfp> Hi, I'm currently running an old Debian, and I'd like to switch to Ubuntu. My plan is to setup Ubuntu server in a VM and then somehow copy it in place of the Debian. I have a few questions: 1. I'd like to make a backup. The system lives on / (80GB SSD) and I have 1TB of media in /dev/sdb3. I'd like to nackup / by booting a LiveCD and taking an image with `dd`. If I screw up, I can restore with dd and
[16:32] <hfp> everything will be like nothing happened. Correct? 2. I have VMs running on the Debian host using QEMU (latest for Debian, but old in other universes). How do I transfer these? I was thinking of copying their disk images and somehow extract their configuration in an XML file. Is that enough? 3. What's the best way to run VMs in Ubuntu Server, which software should I use?
[16:33] <RoyK> don't dd it - just rsync the lot and setup the new VM with the stuff you need
[16:35] <hfp> RoyK: Sorry I'm not sure I follow. I'd like to backup the system SSD in such a way that I could just wipe it all and restore it exactly the way it was before I started screwing around. If I rsync the system SSD, then I'll have to reinstall Debian and then copy the files over to restore it, correct?
[16:35] <LostSoul> RoyK: Good to see you here :)
[16:36] <LostSoul> I have problem with setting up bridge to LXC containers
[16:36] <LostSoul> They see each other on brigde IP but are  no accessible for internal network (for example I can't access them from my laptop)
[16:36] <RoyK> hfp: I see - just use dd or ddrescue to create an image file on your data disk and then reinstall
[16:36] <LostSoul> My lxc config for container: http://paste.debian.net/hidden/4aa6df9f/ my brigde: http://paste.debian.net/889024/
[16:38] <RoyK> LostSoul: I haven't used LXC, but I use this brige setup with KVM http://paste.debian.net/889027/
[16:39] <LostSoul> Thanks RoyK
[16:41] <hfp> RoyK: I just thought of something. I could image the Debian machine, install Ubuntu, and then run the Debian as a VM within Ubuntu; the Debian VM running its original Virtual Machines while I transfer everything to Ubuntu. I'm not sure how long it's going to take me and I'd like to keep services running while I do it. The VMs are not very resource intensive so hopefully they'd be ok with the VM in a VM
[16:41] <hfp> penalty
[16:46] <RoyK> hfp: I'd recommend against it - it'll be easier to just migrate the old system to a new one by just reinstalling and moving the config
[16:48] <hfp> Right, so I make a backup with dd, I install Ubuntu server in a VM, and then I rsync the fully configured Ubuntu VM to Debian's / ? Or do I wipe Debian, install and configure Ubuntu, install VirtualBox, and setup the VMs from their drisk images?
[16:48] <hfp> RoyK: ^
[16:54] <RoyK> hfp: just setup a new machine and configure it like the old one
[16:55] <hfp> RoyK: I see, thanks
[16:57] <RoyK> hfp: it may seem to take longer, but it'll work, and you might even learn a little by it
[16:58] <RoyK> hfp: trying to move things between distros by rsyncing configs may work, but may also fail, because of different versions of things - so better setup the new vm from scratch
[16:59] <hfp> RoyK: Yes, that makes sense. What's the best way to run VMs in Ubuntu? VirtualBox? It's a headless home server if that makes a difference.
[17:01] <RoyK> hfp_: kvm
[17:01] <LostSoul> Promiscuous mode was not turned on... damn..
[17:01] <LostSoul> Thanks RoyK
[18:04] <hfp> I have an easy question: if a file is created on one mchine, and belongs to some user, can you chown it on another machine and make it belong to whichever user you want on that other machibne? I'd say yes, and so you can't losr files that way unless they're encrypted but I just want to make sure
[18:43] <oxkipo> Hi Im getting an error on conjure-up when deploying openstack on a single node fresh ubuntu 16.04 after updates "Exception: Unable to find /home/stack/.local/share/juju/accounts.yaml" Please may someone help me?
[19:10] <hfp> Where should I mount a partition used to store data? Does it go in /mnt, /media, or somewhere else?
[19:11] <hfp> It's not on a removable drive, it's on an internal SATA drive
[19:18] <patdk-lap> anywhere you want
[19:18] <patdk-lap> personally I like /var/lib/mysql
[19:20] <hfp> I put it in /media/storage
[20:23] <hfp> I'm having fun with NFS permissions now... I set no_root_squash in /etc/exports, but I don't know how to make it so that network clients with different UID and GID can also write to the NFS share. I remember it was a pain last time around but I can't remember how I solved it. Any pointers?