[00:03] <Patrickdk> not sure why you would want to disable that
[00:06] <Delemas> Because it also has no way to define the CA files to use so it always fails.
[00:07] <Delemas> So an upgrade resulting in no SMTP or IMAP connections. Very useful webmail...
[00:07] <Delemas> If Ubuntu had a modern squirrelmail release I could just define the options and move on...
[00:08] <Delemas> I guess I'm going to have to package it myself...
[00:09] <cryptodan_laptop> Delemas: https://help.ubuntu.com/community/Squirrelmail
[00:10] <Delemas> That doesn't help.
[00:11] <cryptodan_laptop> thats what came up when I googled squirrelmail ubuntu 14.04
[00:11] <trippeh> to be fair, it doesnt look like squirrelmail has actually made a release since then
[00:11] <trippeh> its all "just use a snapshot plz kthx"
[00:11] <Delemas> squirrelmail on 14.04 works because it doesn't using php 5.6
[00:12] <Delemas> They have a stable snapshot.
[00:12] <cryptodan_laptop> cant drop back to 5.5?
[00:12] <Delemas> We are actually running a snapshot now only it was taken in 2012...
[00:13] <Delemas> i.e. this if the version in universe: 2:1.4.23~svn20120406-2
[00:13] <Patrickdk> using 5.6 is pretty daring, 5.6 is like just released a few months ago I think
[00:14] <Patrickdk> oh, universe, that isn't ubuntu supported
[00:14] <Patrickdk> hmm, guess it will be a year old soon
[00:15] <Delemas> 3 years by my count...
[00:15] <trippeh> my personal outdated-in-ubuntu pet peeves is amavisd-new and wireless-regdb (both in main)
[00:15] <Delemas> wordpress would be another...
[00:15] <Patrickdk> how? release date of php 5.6.0 is aug 2014
[00:16] <Delemas> Ah that makes sense I thought you meant squirrelmail...
[00:16] <trippeh> when debian oldstable ships newer versions than 15.04 ;)
[00:24] <sarnold> trippeh: the requestsync program can file bug reports about those things for you :)
[00:26] <trippeh> pretty sure they all have "new upstream release" bugs on them at least
[00:26] <trippeh> I heard wireless-regdb is beeing worked on now
[00:27] <sarnold> \o/
[00:27] <trippeh> finally we can have working 802.11ac ;)
[00:27] <sarnold> fancy, I still think of G as the neat new thing
[00:31] <trippeh> well it sorta works, just without its killer feature, the extra performance ;)
[00:38] <trippeh> oh god squirrelmail is still using sourceforge?
[01:03] <drocsid> I'm having trouble increasing ulimit -l max locked memory       (kbytes, -l) 64
[01:03] <drocsid> in ubuntu server
[01:03] <drocsid> in /etc/security/limits.d/elasticsearch.conf
[01:03] <drocsid> I set elasticsearch   - memlock unlimited
[01:03] <drocsid> and I  tried sysctl -p
[01:03] <drocsid> I also tried sudo sysctl --system
[01:04] <drocsid> and logging out and back in
[01:04] <drocsid> but locked memory max remains at 64kbytes. How do I change this in ubuntu-server?
[01:05] <sarnold> drocsid: check your /etc/pam.d/ files to make sure you didn't accidentally typo the name /etc/security/limits.d/elasticsearch.conf when you set it up
[01:06] <drocsid> sarnold: I just asumed that I could place that file in limits.d, maybe I should just set this in limits.conf instead?
[01:07] <sarnold> drocsid: hmm, maybe you should be able to..I didn't see anything referring the limits.d directory inthe pam_limits manpage so I hadn't realized it already existed.. :)
[01:07] <sarnold> drocsid: .. but low and behold, I've got the limits.d directory too. hah.
[01:08] <sarnold> drocsid: dno't forget that the limits set via the pam_limits module will only affect whatever login mechanisms are configured to use it, via /etc/pam.d/..
[01:08] <drocsid> sarnold: from my understanding it's there so you don't goof and mess up all the system limits. But this isn't working
[01:09] <sarnold> drocsid: it's both way simpler than that and way more complicated than that :/ it only affects whatever login mechanisms are configured to use pam_limits via /etc/pam.d/
[01:09] <drocsid> Is /etc/security/limits.conf part of pam_limits / pam ?
[01:09] <sarnold> yes
[01:10] <sarnold> drocsid: so if it is started via an upstart or sysvinit script, pam_limits is never run and those files are never parsed anyway
[01:10] <sarnold> but if you're logging in as elasticsearch via ssh and need this locked memory limit raised, that'll work fine
[01:11] <drocsid> hmm so then since I start the process using the init script as root, let me see what it does.
[01:11] <sarnold> drocsid: try this: http://upstart.ubuntu.com/cookbook/#limit
[01:11] <drocsid> didn't know this was that complicated. :(
[01:11] <sarnold> the thing is, _something_ needs to run to set those limits.
[01:11] <sarnold> otherwise the limits are inherited from the parent process
[01:12] <sarnold> well,that's backwards; limits are inherited from parent processes, and if it happens to run as root, and happens to be configured to use pam and pam runs pam_limits....
[01:12] <drocsid> looks like these lines exist in the init script
[01:12] <drocsid>       if [ -n "$MAX_OPEN_FILES" ]; then
[01:12] <drocsid>                 ulimit -n $MAX_OPEN_FILES
[01:12] <drocsid>         fi
[01:12] <drocsid>        if [ -n "$MAX_LOCKED_MEMORY" ]; then
[01:12] <drocsid>                 ulimit -l $MAX_LOCKED_MEMORY
[01:12] <drocsid>         fi
[01:12] <sarnold> the end result is that it feels elegant and simple once you know it; and until you know it, it feels incredibly complicated and terrible. :)
[01:12] <drocsid> so I will try setting those vars in /etc/default
[01:13] <drocsid> and looks like it will set them maybe
[01:13] <sarnold> nice, are those variables inherited from an /etc/default/ file hopefully?
[01:13] <sarnold> woo
[01:14] <drocsid> it's a big pain in the ass as far as I'm concerned, these simple configs differ between all the distros and init scripts. Setting it one way doesn't work in another environment. :(
[01:15] <drocsid> maybe I'm just not getting it right, but I thought that /etc/security/limits.conf was supposed to set the limits across the system...
[01:17] <drocsid> sarnold: thanks
[01:17] <sarnold> drocsid: I can understand :) it took me years before I came to terms with PAM...
[01:17] <sarnold> drocsid: all fixed?
[01:18] <drocsid> yeah, the init script takes care of it.
[01:18] <sarnold> nice.
[01:42] <Locke2002> There is a gap in my kern logs... How can that happen?
[01:43] <sarnold> Locke2002: of what sort? it's quite common for the kernel to go weeks without logging anything
[01:44] <sarnold> my currently running kernel has a period of 139 hours with nothing to report :)
[01:45] <Locke2002> kern.log.1 ends on May 28, kern.log starts today after an unplanned reboot of my VPS. My SSH session showed root sent shutdown, another log mentioned stopping a service for runlevel 15.
[01:45] <Locke2002> maybe my VPS provider did a reboot?
[01:46] <sarnold> it's unusual for vps providers to not at least mention it to you, but there have been some serious xen / qemu issues since then
[01:46] <Locke2002> I know it wasn't too sudden, I got all my fail2ban shutdown notifications
[01:48] <patdk-lap>  depends on the vps provider
[01:48] <patdk-lap> my cloudatcost node randomly reboots all the time
[01:48] <Locke2002> haha, ok
[01:48] <patdk-lap> I have only ever received one notice about it
[01:48] <sarnold> with a name like that, it's not too surprising.. :)
[01:48] <patdk-lap> even amazon, same deal
[01:48] <patdk-lap> they give notice on hardware failure though
[01:49] <patdk-lap> but I have aws instances randomly reboot, without warning or notice
[01:49] <patdk-lap> on aws
[01:49] <Locke2002> is it not unusual that kernel wouldn't have anything to log while shutting down?
[01:50] <patdk-lap> who said it shutdown?
[01:50] <sarnold> dang, I don't have any shutdowns in my logs :/
[01:50] <sarnold> patdk-lap: he had a shutdown notification from root in an ssh session
[01:50] <patdk-lap> ah
[01:50] <Locke2002> my fail2ban notifications, they were sent out as it was going down
[01:50] <patdk-lap> ya, you won't get a kernel shutdown notice
[01:50] <patdk-lap> it will be readonly by then
[01:51] <Locke2002> ok
[01:51] <Locke2002> thanks guys :)
[01:51] <sarnold> Locke2002: usually the best notice you get that a reboot happened is the huge logspam at 0.xxxx seconds in the log
[01:51]  * patdk-lap uses uptime :)
[02:01] <histo>  22:01:57 up 70 days, 1 min,  2 users,  load average: 0.00, 0.01, 0.05
[07:36] <lordievader> Good morning.
[08:29] <histo> *yawn&
[10:23] <trijntje> Hi all. I want to start using zfs on ubuntu. Can somebody point me to a good resource to get started? I have no experience with zfs
[10:33] <Sling> trijntje: did you try google? :) the ubuntu wiki has a nice zfs article with a ppa
[10:36] <trijntje> Sling: I found the ubuntu wiki article, but I didn't find it very usefull, its pretty general. I was hoping someone here would know of a best practice/tutorial document for running zfs on ubuntu
[10:36] <trijntje> I tried the server guide as well but I couldn't find any mention of zfs
[10:37] <Sling> what kind of best practice do you mean? zfs kinda does all of its work on its own
[10:37] <Sling> also do you want have a zfs root partition or just separate
[10:45] <trijntje> just separate, I need to store some expensive data and make sure it doesn't silently get corrupted
[10:45] <trijntje> Sling: I guess I'm not too sure what I mean, thats the problem when you are completely new to a subject ;)
[10:46] <Sling> trijntje: zfs is mainly self-healing
[10:46] <Sling> as long as you keep an eye on the status of your pool/volumes using monitoring (im sure there is a nagios plugin for this) then you're good
[10:47] <bekks> trijntje: https://pthree.org/2012/04/17/install-zfs-on-debian-gnulinux/ - thats the starting point for ZFS on Linux.
[10:54] <trijntje> bekks: thanks, I'll go and read that
[15:03] <vikram1> Hi all i need to migrate my server from ubuntu 14 to ubuntu server 15 plz help me
[15:03] <lordievader> vikram1: sudo do-release-upgrade?
[15:05] <vikram1> vil my all database vil b safe ryt bcoz i installed on ubuntu desktop 14 LTS and needed to upgrade to Ubuntu Server 15 LTS
[15:05] <lordievader> vikram1: Read the release notes.
[15:07] <vikram1> release uipgrade will uopgrade me to desktop release. I want to upgradee / migrate to server edition.
[15:08] <lordievader> vikram1: do-release-upgrade only upgrades the packages that are currently installed.
[15:08] <vikram1> it'll upgrade to desktop to desktop version i need to migrate to desktop edition to server edition
[15:09] <lordievader> What?
[15:10] <vikram1> i have installed my server on Ubuntu 14 LTS desktop edition now. i need to migrate to Ubuntu 15  Server edition and not desktop edition
[15:11] <teward> vikram1: there's no distinction really between "Desktop Edition" and "Server Edition" to the upgrader.
[15:11] <vikram1> will migrating from Desktop edition to server edition increase the performance ?
[15:11] <teward> `do-release-upgrade` will upgrade everything as is, ignoring the image you've installed with, and doing the upgrade on a package basis
[15:12] <teward> vikram1: removing the GUI portions might help a little, but without further details on your environment it's impossible to figure out what you're trying to achieve other than an upgrade
[15:12] <vikram1> Thanks a lot for teward
[15:12] <teward> vikram1: from apt and do-release-upgrade's perspectives, it's upgrading based on the migration path from trusty -> utopic -> vivid
[15:12] <teward> not the image you used
[15:13] <teward> (however 14.04 LTS -> 15.04 will require one additional upgrade, 14.04 LTS -> 14.10, then 14.10 -> 15.04)
[15:13] <teward> unless you absolutely need 15.04 for your environment, i'd suggest keeping the LTS
[15:13] <teward> (or, fresh install after backing up the data and determining the applications you need to install)
[15:14] <deadrat> Vikram1 must've installed the default unity desktop in his server. Now he is wondering if changing to Ubuntu server increases performance.
[15:14] <teward> deadrat: he's installed the Desktop iso
[15:14] <teward> deadrat: he's also asking two things:  "How do I migrate to Server Edition" and "How do I go from 14.04 LTS to 15.04"?
[15:14] <deadrat> Yes..
[15:14] <vikram1> What would suggest me as i m using it for internal usage abt 100 users max for a social networking site.
[15:15] <teward> deadrat: so, removing the GUI bits is not too hard, nor is removing the other apps provided he has a list of what to remove
[15:15] <teward> the upgrading will be two-stage though
[15:15] <teward> and with each stage he runs the risk of explosions
[15:15] <deadrat> Let him stay in LTS till next lts release and maybe he can remove the desktop packages that are not required.
[15:15] <teward> agreed
[15:16] <teward> and for a server, performance will probably increase since the GUI overhead wouldn't be existing
[15:16] <deadrat> Yup.. So unity and its dependencies must be removed.
[15:17] <teward> as well as libre office and the other installed applicatoins
[15:17] <teward> i should really diff the ubuntu-desktop package and the default server install to find how to purge the GUI components xD
[15:17] <deadrat> And other packages like software centre , Firefox etc.
[15:18] <teward> mhm
[15:18] <teward> hence the diffing of the desktop package and the server install xD
[15:18] <teward> (brb)
[15:19] <deadrat> Where can he find the list of packages included?  . they are in release notes, I guess..
[15:20] <teward> deadrat: i think it's a large enough list that it's not in the release notes
[15:20] <teward> IDK i'd have to check
[15:21] <deadrat> Vikram1 should be able to figure that out.
[15:22] <teward> mhm
[15:23] <vikram1> thanks a lot guys
[15:25] <deadrat> U r welcome.
[16:29] <mgz_> hiya. is there a reason the base ubuntu server image doesn't run ntpd? seems you'd like to avoid clock skew by default.
[16:37] <tonyyarusso> mgz_: While often a good idea, it's certainly not necessary, and having unnecessary daemons running with open ports is avoided in default configurations.
[17:06] <rbasak> mgz_: >= Vivid runs systemd-timesyncd now.
[17:06] <rbasak> mgz_: we were going to start running ntpd by default, but no need now.
[17:07] <teward> does anyone know what the default charset is for a MySQL table on 12.04?
[17:07] <teward> or do I have to hunt things down?
[17:07] <mgz_> rbasak: aha, thanks!
[17:07] <trippeh> rbasak: good. newer upstream ntpd supports client only modes without gross config hacks though.
[17:08] <mgz_> so, still an issue for charms for now but goes away next lts
[17:08] <trippeh> ntpd is the only reason our servers run with conntrack enabled (for UDP only, tho)
[17:11] <trippeh> ntp in ubuntu is also on my list of outdated in ubuntu pet peeves :)
[17:11] <trippeh> although its also outdated in debian, so..
[18:18] <bitfury> hey guys, I'm trying to get a rackspace cloud VM going with an older version of Ubuntu (14.04.01 LTS) since they only have .02 I'm booting from an iPXE tool they have available but the only option I see is to use a "preseed" config file.. I have no idea what this is but would it be possible to specify the version I want install along with the ISO to use in there?
[18:20] <sarnold> once you get to needing a preseed file, the ISO has already been selected and booted..
[18:21] <bitfury> :( it's annoying they don't allow uploading an ISO and using it to build your VM..
[18:21] <bitfury> lame
[18:22] <bitfury> and importing an image into their cloud environment is a pita
[18:23] <sarnold> I'm surprised they don't have the 14.04 or 14.04.1 images available. :/
[18:23] <bitfury> I know :(
[18:23] <lordievader> What is wrong with the .02?
[18:24] <bitfury> nothing wrong, this is a test box I'm building and we would like to keep same version as production
[18:31] <lordievader> Ah. In that way... Hmm.
[18:34] <bitfury> could I somehow boot from .02 then downgrade .01?
[18:34] <bitfury> to
[18:37] <sarnold> it'd probably just be a matter of selecting the proper kernel metapackage, linux-image-generic, linux-tools, linux-generic, etc..
[18:38] <bitfury> :\ sounds like hell
[18:39] <sarnold> heh, I can understand; it'd be nicer to just find a 14.04.1 image in their cloud offerings in the first place.
[18:39] <bitfury> yeah..
[18:40] <sarnold> but downgrading a 14.04.2 to the earlier kernel is probably a much easier route than dealing with preseed files.
[18:41] <bitfury> this would be another option: http://www.rackspace.com/knowledge_center/article/creating-an-ubuntu-1310-image-for-the-rackspace-open-cloud-0
[18:41] <bitfury> but looks like a pita too..
[18:41] <bitfury> dealing with kickstarter files, XenServer etc
[18:42] <bitfury> damnuuuuu rackspaceee
[18:44] <bitfury> are changes from .1 to .2 significant? I read a different kernel is used
[18:45] <bitfury> would version of packages be different on default installation of both?
[18:45] <sarnold> bitfury: the kernel and X11 family of packages are newer; the .2 kernel will be unsupported sometime in 2016, so you'd have to deal with that eventually. most users would never notice the difference, but it depends what you're going to do with the machine..
[18:47] <sarnold> .. the last diagram on this page describes what I mean about the kernel being end-of-lifed before the rest of the release: https://wiki.ubuntu.com/Kernel/LTSEnablementStack
[18:48] <bitfury> idea is to have this box running our app for demos, testing is done locally by a developer (using a windows box)
[18:49] <bitfury> I have another local VM running .01 where I test security and package updates
[18:49] <lordievader> bitfury: Dd that over?
[18:49] <bitfury> but I wanted to use this box to test all of this stuff
[18:49] <bitfury> instead of having a local VM
[18:49] <bitfury> guess I can keep it and then use .02 for demos
[18:50] <bitfury> what's DD lordievader
[18:50] <lordievader> dd, a bitcopy tool.
[18:54] <bitfury> lordievader: hmm ok, 60G transfer to a remote server though..
[18:54] <bitfury> across the internetz
[18:55] <bitfury> never done it but I assume SSH is involved and that will add overhead.
[18:55] <bitfury> might take a while..
[18:56] <bitfury> fuck it, will roll with .02
[18:56] <lordievader> Please watch your language.
[18:57] <bitfury> sorry
[18:57] <bitfury> thanks guys
[19:06] <sarnold> have fun bitfury :)
[19:10] <bitfury> sarnold: I'll try ;)