[00:06] <drab> speaking of bugs...
[00:06] <drab> patdk-wk: you're not hitting this one? https://github.com/systemd/systemd/issues/3374
[00:08] <drab> for some reasons it's not happening on one server class but it's happening on the other
[00:08] <drab> same i350 cards with igb module
[00:08] <drab> same /etc/network/interfaces
[00:09] <keithzg> Is there some sort of SSL caching mechanism? I've *definitely* configured a Dovecot instance to use a real certificate, but it keeps spitting out an older self-signed one that isn't configured to be used anywhere in the dovecot configs, and it's utterly baffling me!
[00:10] <drab> mmmh, I think I have renaming problems... it seems it tries to bring up the bond before the interfaces are renamed
[00:10] <drab> and the above systemd bug is just a red herring
[00:10] <drab> so it doesn't find any eth0
[00:22] <drab> uhm, kinda dodgy, but it seems if I move the bond stanza first, before the interfaces, then it works at boot too
[00:22] <sarnold> keithzg: are you sure you're working with the correct machine? the correct paths? try breaking something very badly in the config and make sure dovecot refuses to start with the bad config, and that your client testing it also reports failure to connect, etc..
[00:23] <drab> otherwise I'm seeing messages complaining that bond0 isn't ready and somehow it's not retrying/working out
[00:36] <patdk-lap> nope, don't believe I have
[04:42] <faekjarz> Hey there! Do i need a special kernel boot parameter to use SR-IOV?
[04:46] <patdk-lap> no
[04:46] <patdk-lap> but your nic and bios and driver all need to support it, and have it enabled
[04:48] <faekjarz> my nic is an i350 (check) bios? VT-d is enabled. should suffice (?)
[04:49] <patdk-lap> heh?
[04:49] <patdk-lap> vt-d has nothing at all to do with sr-iov
[04:49] <faekjarz> ok
[04:49] <patdk-lap> kvm, yes, sr-iov, no
[07:09] <lordievader> Good morning
[07:09] <lordievader> Didnt ethtool list support for that feature?
[08:20] <faekjarz> lordievader: talking to me? i didn't come across any iov related output of ethtool.
[08:21] <lordievader> No, more in general... thinking out load.
[08:21] <lordievader> loud even.
[08:26] <faekjarz> btw: i put my i350 in the wrong PCIe slot. The mobo manual doesn't have a block diagram, but i think it's connected to the PCH now. echoing 1-7 to 'find /sys -name "*numvfs"' now spawns a lot of VFs *happy*
[11:25] <adrian_1908> Not sure if this is the right place to ask, but does anyone know if 16.04 will eventually have PHP 7.1 through the regular streams?
[11:27] <blackflow> adrian_1908: regular streams? you mean the main, updates and security repositories? Probably not
[11:29] <adrian_1908> Yeah I meant those repos you mentioned. Ok.
[11:30] <adrian_1908> There looks to be a PPA, so if that's well done I imagine that would work too if I feel I want to upgrade.
[11:31] <blackflow> is there any significant benefit of 7.1 over 7.0? (asking as an admin of 7.0 systems)
[11:31] <blackflow> I mean, unless you're a dev and want to take advantage of the new language features.
[11:33] <adrian_1908> Beats me. I'm setting up a PHP server for the first time myself, and I noticed that several options have been deprecated in 7.1 (but not 7.0) so if there are saner design choices, one may wish to upgrade before 18.04 LTS comes out.
[11:35] <blackflow> usually deprecated options linger around for long time before they're completely dropped, in php. from what I see in the changelog, it's nothing significant
[11:35] <blackflow> in other words if you're on LTS and favor stability, no reason to rush over to 7.1
[11:36] <adrian_1908> ok!
[12:23] <daincredibleholg> good morning together
[12:24] <daincredibleholg> I was wondering if any of you uses Ansible to install graphite-web on your server? I am looking for a way to automate the step `graphite-manage syncdb` which requires super user details and the only way of doing this without interaction is to have a fixture file which would mean having the super user details in plain text in my version control. Nothing I necessarily fancy...
[12:43] <daincredibleholg> Ok, found a way. Will use fixture and use template vars to inject the fields.
[13:02] <blackflow> Need some help understanding the bugfix cycle. I reported a bug, and the fix was found, the bug is in "Fix released" status (#1673357), also tagged "server-next", with a note that the problem is fixed in Zesty. Will the fix be applied to Xenial? How/where can I track the progress of such fixes trickling down to LTS?
[14:22] <blackflow> Or, better question: What happens next to #1673357?
[15:45] <jancoow> Hi. I don't know what happends, but I was trying to restart docker
[15:45] <jancoow> docker.service, and now Nothing response anymore
[15:45] <jancoow> I can't even reboot: -bash: /sbin/reboot: Too many open files in system
[15:46] <jancoow> how could I reboot now?
[15:50] <chuck66> have you tried  $ sudo shutdown -r now?
[15:52] <jancoow> yup
[15:52] <jancoow> can't execute any command
[15:57] <blackflow> jancoow: systemctl -f reboot  ?
[15:58] <jancoow> in the mean time I walked downstairs to my server &^
[16:29] <chuck66> jancoow: were you able to reboot?
[16:33] <jancoow> with the reset button, yeah
[18:29] <drab> urm, anybody knows why zpool list and df -h would disagree so far as 3TBs?
[18:30] <drab> zpool list shows my pool to be 10.9TB, which is what I'd expect it to be with 6 disks in raidz2, so 4x3TB, with about 1TB lost, which is crazy, but fair enough
[18:30] <drab> however df -h says 7.1T
[18:36] <drab> oh, nm, sort of, at the very least zfs list and df agree
[18:48] <drab> nm, pebkac
[20:56] <Haris> hello all
[20:57] <Haris> I have mysql on 14.04. I stop it. It gets instantly restarted. Is that apparmor or watchdog or what watching/re-running it ?
[21:11] <bekks> Haris: How do you stop it?
[21:12] <Haris> by running /etc/init.d/mysql stop
[21:12] <compdoc> lies
[21:12] <compdoc> ooops
[21:24] <tomreyn> Haris: this suggests you have non standard behavior on this system.
[21:25] <tomreyn> check your process list for any watchdogs running there.
[21:39] <OerHeks> Haris, -d /etc/init.d/mysqld <start stop restart>  but normally one would use service to handle mysqld
[22:04] <Haris> hmm
[22:04] <Haris> I see
[22:13] <Haris> will check it in lab env
[22:13] <Haris> thanks all