[07:46] <lordievader> Good morning
[08:48] <sliddis> If I change an option in a DHCP-server (eg dns server), and a client has a lease. After half the lease-time the client will ask if it still can use it's address, right? Would the client also apply the new DNS servers from DHCP at the same time? Or does the client have to release and renew completely to receive the new DNS?
[08:50] <rbasak> I think the renewal will cause it to pick up the new DNS servers (through DHCP options present in the renewed lease) but I'm not absolutely sure.
[08:50] <rbasak> The client doesn't *have* to renew at half the lease time. Technically it could wait all the way to the end.
[08:51] <bekks> sliddis: No, it just asks wether the lease is still valid.
[08:51] <bekks> sliddis: For applying new options, you need to request a new IP address.
[09:03] <sliddis> bekks: are you sure? If the client asks if the lease is valid, the server never sends options?
[09:03] <bekks> I am pretty sure of it, yes.
[09:04] <rbasak> In that case a smart DHCP server should refuse the renewal and force the client to take an entirely new lease.
[09:04] <rbasak> No idea whether that happens or not though,.
[09:06] <sliddis> It seems though like the clients have recieved the new dns server. I dont get any firewall hits on the old dns server. Although not verified.
[09:06] <sliddis> I am using a mikrotik routerOS as dns-server
[09:06] <sliddis> I am using a mikrotik routerOS as dhcp-server*
[09:07] <sliddis> not dns -_-
[11:08] <chRi____> Hi all! That's my first activity here so please be kind ;)
[11:08] <chRi____> I'm normaly doing webdevelopment, but since a lot of my colleagues don't bother, I also administer most of the webservers and I'm loving the out-of-the-box configuration that ubuntus vanilla lamp-stack offers...
[11:08] <chRi____> But now I have a strange problem:
[11:08] <chRi____> I installed a Ubuntu Server 14.04.3 LTS as a virtual machine inside a KVM/libvirt infrastructure.
[11:08] <chRi____> My Images are qcow2 and I enabled LVM for my ext4 fs.
[11:08] <chRi____> Strange thing and I'm not sure if it is really Ubuntu related, you guys have to decide:
[11:08] <chRi____> I (re-)installed a php application via composer several times and had the strange result, that all the executable bits where removed from scripts that need to be run right after the php-apps installation. What might be the problem?
[11:08] <chRi____> The git-repo that gets cloned via a packagist-link, has the executable bits, but the decompressed files don't... Any Ideas?
[11:08] <chRi____> I'm able to run other scripts, so at least I assume, there is no "noexec" mount flag set or so...
[11:08] <chRi____> I would be really glad if someone could help or has also experienced such behaviour in the past!
[11:08] <chRi____> Best Regards,
[11:08] <chRi____> Christian
[11:13] <bekks> git doesnt keep track of executable permissions, since thats irrelevant for a RCS.
[11:15] <rbasak> Yes it does.
[11:15] <rbasak> $ git diff
[11:15] <rbasak> diff --git a/foo b/foo
[11:15] <rbasak> old mode 100644
[11:15] <rbasak> new mode 100755
[11:16] <rbasak> It's only the executable bit mind. I had actually set foo to 0700 in that example.
[11:16] <bekks> Then how does it handle permissions for two different client OS?
[11:16] <rbasak> No idea. All I'm saying is that it does track the executable bit (but not whether it's assigned to user, group or other - it seems to just be one bit)
[11:17] <bekks> ah ok.
[11:17] <chRi____> @rbasak: that's what I also exprienced until yesterday
[11:18] <chRi____> excutable bit was still there
[11:18] <rbasak> chRi____: I don't know what composer is.
[11:18] <rbasak> chRi____: but your problem lies in the area of how you deploy it
[11:18] <chRi____> its a package manager for php
[11:19] <rbasak> You are welcome to try here but you might find more people are likely to know the answer in a PHP related channel or something.
[11:19] <rbasak> It doesn't sound like your problem is Ubuntu-specific.
[11:28] <chRi____> You maybe right... But I'm asking here, because I have an Arch Workstation at home and the the same composer-based installation leaves the executable bits intact
[11:29] <chRi____> And I can't remember that I had this problem before, when I used 14.04.1 LTS or so...
[11:35] <rbasak> There could be a bug in Ubuntu then, or a bug in composer in how it handles an Ubuntu system.
[11:36] <rbasak> Somebody who knows composer will need to pin it down.
[11:36] <rbasak> Is composer packaged in Ubuntu or are you installing it from an external source?
[11:40] <chRi____> since it really is just an php application archive it's installed via curl and you can than install it "globally" by simlinking the composer.phar
[11:40] <chRi____> I will try to get in  touch with the composer-minds!
[11:40] <chRi____> thanks!
[12:48] <alximych> hi! I intstalled fail2ban and each time I start the service I receive an email notification that fail2ban was started. I have already tried to create a senamail.local, mail.local, sendmail-common.local, etc. files and override actionstart = (nothing), but I still get notifications. How do I disable service started/stopped mail notifications?
[12:50] <smo80> hi
[12:52] <smo80> i have a new fujitsu primergy tx 1330 m1 server... with damn hardware raid controller.... first i made a software thru ubuntu installer, no problems but when i reboot i don t have anything to choose in the bios to boot on, so i created a raid array thru the raid controller interface on boot... it add an raid arrray"
[12:52] <smo80> option to the bios
[12:53] <smo80> but once i return to ubuntu installer i can t make the partitions as i want... i m trying lvm, i can create my group and logical volumes but don t know what to choose at the grub-install , /dev/????
[12:54] <smo80> i tried /dev/sda, /dev/mapper/md126 and others it always fail
[13:00] <alximych> *solved my issue by creating 9 .local files...
[13:06] <pmatulis> morning
[13:27] <smoser> hey. an someone who has a "fresh"-ish ubuntu run 'echo $TERM' from a gnome-terminal ?
[13:27] <smoser> i'm pretty sure i've not touched that, but one can never be sure.
[14:06] <rbasak> smoser: xterm on gnome-terminal on a system that booted Vivid but is actually now running Wily (not rebooted yet).
[14:07] <rbasak> Though it's not really a fresh machine. Upgraded for many years. I don't recall ever touching terminal stuff though.
[14:07] <smoser> and ?
[14:07] <rbasak> "xterm"
[14:23] <fishcooker> i have 2 dns cache, 1 server take so long to resolve yahoo.co.uk, but the another one take short time to resolve, how to inspect the case?
[15:48] <soc42> hi
[15:48] <soc42> i want to configure xdebug settings on ubuntu 12.04, but can't find the right .ini file to change these settings
[15:49] <soc42> the .ini-files i found only containt the path to the xdebug.so module, but no xdebug config options... ?
[15:52] <soc42> as i am doing a $> php -i the current xdebug settings appear, but where the heck are they set?
[15:52] <soc42> i checked all files, which appear for $> php --ini
[15:55] <Odd_Bloke> smoser: Argh, just upgraded to wily and now I'm hitting https://bugs.launchpad.net/ubuntu/+source/gnome-terminal/+bug/1463072
[15:56] <smoser> :)
[15:56] <Odd_Bloke> smoser: I'd be tempted to say that we should be doing something to make this less painful for our users.
[15:56] <Odd_Bloke> Though as I am one of those users, I am somewhat biased. :p
[16:00] <Odd_Bloke> In fact, I might just stop using gnome-terminal.
[16:08] <rbasak> hallyn_: need any help with bug 1490110?
[16:13] <hallyn_> rbasak: yes i do.  got any idea what's going on there?
[16:14] <rbasak> hallyn_: no idea, I've not looked apart from confirming the reproducer. I can start digging though I only have 45 minutes left today. I can carry on looking tomorrow.
[16:14] <arrrghhh> Hey all.  Still having issues with syslog, kern.log, auth.log etc being completely empty and untouched files.  I found a bug which looks like it's related, but the 'fix' didn't work for me.  https://bugs.launchpad.net/ubuntu/+source/rsyslog/+bug/1059854
[16:15] <Overand> Is there any reason I should or shouldn't use ext2/3/4 for /boot?
[16:15] <Overand> (Do we need /boot anymore? I'm looking at using LVM, so, mayube?)
[16:16] <jpds> Overand: No, grub can do LVM
[16:16] <Overand> Also, thoughts on using XFS vs EXT4?  This is 14.04
[16:16] <jpds> Overand: You only need a /boot when you do encrypted /
[16:16] <jpds> Overand: XFS is useful for lots of small files I think
[16:16] <Overand> 75% of my systems are 14.04 with ext4, ~5% are 12.04, and ~15% are CentOS 7 with XFS
[16:16] <hallyn_> rbasak: it kinda rings a bell of when i converted libvirtd to systemd,
[16:16] <jpds> Overand: For /boot, you could probably just do ext4
[16:17] <Overand> (I more meant xfs for / or /var maybe, but I don't really care one wya or another)
[16:17] <hallyn_> so system was systemd, but no systme job so sysvinit job was running.  on upgrade, systemd lost track of the sysvinit job.
[16:17] <hallyn_> but, that cant be the case really...
[16:17] <Overand> Just building out a new template, wondering what the best-practices are these days.  Defaults on Ubuntu have been EXT4 so I've been using EXT4, likewise XFS on CentOS 7
[16:18] <RoyK> jpds: XFS historically was rather bad on lots of small files, but I think most of that has been fixed now
[16:18] <Overand> I'll probably stick with ext4, but I suppose I'll want to enable xattrs unless there's some raeson not to
[16:18] <RoyK> jpds: Apart from that, the only reason I'd recommend ext4 over xfs is if you ever would want to shrink the filesystem
[16:19] <RoyK> Overand: I don't know any reason not to use xattr - it's been stable for years
[16:20] <RoyK> the main reason redhat switched to XFS was for big-ish data - ext4 isn't very good on large filesystems when you have to do an fsck
[16:20] <Overand> These will probably just be fairly bog-standard LAMP servers usually
[16:20] <RoyK> I had to fsck a 9TiB fs some weeks back, and it took ~8 hours
[16:20] <Overand> At least so it seems based on our trajectory
[16:21] <Overand> RoyK: Yeah, I ahve a fairly large ext4 volume at home.  RAID 6 setup with 5 2TB SATA disks
[16:21] <Overand> like 85% of the array is sliced up as raid 6, actually, the rest is RAID 10 and raid 1
[16:21]  * Overand really wishes everyone ever involved in ZFS could get into a room and sign a licensing change
[16:22] <jpds> Overand: Probably easier to use btrfs
[16:22] <RoyK> Overand: hehe - you can use zfsonlinux, though
[16:22] <RoyK> jpds: really?
[16:23] <RoyK> jpds: last I checked, raid[56] on btrfs was still experimentally
[16:23] <jpds> RoyK: I read these days that it performs better than zfs
[16:23] <jpds> RoyK: Yeah, the raid[56] part is the only flaky bit
[16:23] <RoyK> I'm not sure
[16:25] <jpds> http://www.diva-portal.org/smash/get/diva2:822493/FULLTEXT01.pdf
[16:25] <arrrghhh> although bug 1059854 doesn't really match up with my issue... rebooting the server/rsyslog service I still do not get any logging...
[16:27] <Overand> the raid5/6 part is the paort i want
[16:27] <Overand> well, raid 6/7
[16:27] <Overand> (coupled with the checksumming features and such)
[16:28] <Overand> god there are so many friggin' filesystems out there
[16:29] <quantic> "raid 7" interesting.
[16:29] <Overand> https://en.wikipedia.org/wiki/List_of_file_systems#Disk_file_systems
[16:29] <Overand> quantic: raid 6 isn't going to be that useful for long
[16:29] <Overand> they need to start naming them differently though
[16:29] <Overand> it should be raid 6+1 or somesuch
[16:30] <Overand> http://www.zdnet.com/article/why-raid-6-stops-working-in-2019/
[16:30] <quantic> yeah, raid 7 doesn't actually exist.
[16:31] <quantic> Except for one proprietary thing from one company.
[16:31] <jpds> RAID10 is what seems to be suggested these days
[16:31] <quantic> RAID10 for active data, RAID6 for archive data.
[16:33] <quantic> RAID6 can be used for active data if it's bulk, non-write-heavy data like a fileserver.
[16:33] <quantic> But if you want to know what pain is, run a database server off of RAID6. :P
[16:34] <RoyK> jpds: depending on use
[16:35] <RoyK> quantic: depends on the database - a typical LAMP server's database can reside happily on r6
[16:37] <RoyK> quantic: it all depends on how much iops you need
[16:39] <quantic> RoyK: I think that that's kinda obvious... or I wouldn't have mentioned write-heaviness as a decision point.
[16:42] <RoyK> quantic: most people don't databases that demanding
[16:43] <quantic> RoyK: That's true. Sorry, I'm used to high-demand databases. Enterprise environments and that.
[16:44] <RoyK> quantic: I'm managing some database servers at work (college with 25k students and 2k employees) and we don't have much that requires r10, even if we use it for them
[16:44] <quantic> RoyK: By high-demand I mean write-heavy as HELL.
[16:44] <quantic> RoyK: What's the read-write load of your database? Just curious.
[16:48] <RoyK> quantic: I haven't really checked - sar -b reports 10:1 write/read on the postgres thing, and something 3:1 write/read on the mysql thing
[16:48] <RoyK> wait - that's bytes
[16:48] <RoyK> a lot more writes according to sar's tps data
[16:53] <qman__> I've been using ZFS on Linux for over a year now, there were some bumps in the beginning but it's been pretty stable since
[16:54] <qman__> And those bumps were stuff like manually adding a boot wait so it'd assemble properly
[16:54] <RoyK> qman__: tbh, I haven't seen any issues at all after my PSU was replaced with something big enough to handle all the drives :)
[16:54] <qman__> And tweaking the arc size
[16:55] <arrrghhh> so... any ideas on my logging issue?  nothing is getting to syslog, auth.log, kern.log... they are 0b files which haven't been touched in a while.
[16:55] <RoyK> arrrghhh: lsof -p `pidof rsyslogd`
[16:56] <RoyK> arrrghhh: perhaps it's not been properly reloaded after logs were rotated
[16:56] <arrrghhh> yea I found a bug which suggested changing 'reload' to 'restart'... but that didn't make a difference in my case
[16:56] <RoyK> arrrghhh: a killall -1 rsyslogd should make it reload its config, close its open files and reopen them
[16:57] <arrrghhh> wow that lsof command put out a lot.  let me redirect to a file
[16:57] <RoyK> paste it to pastebinit ;)
[16:57] <RoyK> that is
[16:57] <arrrghhh> oh.  9662: command not found
[16:58] <arrrghhh> lsof: status error on 9662: No such file or directory
[16:58] <RoyK> lsof -p `pidof rsyslogd` | pastebinit
[16:58] <RoyK> did you forget -p ?
[16:58] <arrrghhh> derp I did -l
[16:58] <arrrghhh> what is wrong with me today...
[16:59] <arrrghhh> http://paste.ubuntu.com/12886930/
[16:59] <arrrghhh> w/sudo http://paste.ubuntu.com/12886934/
[17:00] <arrrghhh> hm.  doesn't seem to be interacting with any of the /var/log 'system' files.
[17:00] <arrrghhh> mythtv is logging properly tho, I can confirm that
[17:01] <RoyK> arrrghhh: did you do something interesting to the rsyslog config?
[17:02] <arrrghhh> RoyK, well I did make a change after I noticed the issue based on that bug 1059854
[17:03] <arrrghhh> hm I didn't try this "create 640 syslog adm" to /etc/logrotate.d/rsyslog
[17:03] <arrrghhh> I swapped "reload" with "restart"... I think there was a different bug I was looking at yesterday
[17:03] <RoyK> arrrghhh: which version of ubuntu is this?
[17:03] <arrrghhh> 14.04
[17:04] <RoyK> never seen that issue...
[17:05] <arrrghhh> evidently I haven't looked at system logs in a while, it almost appears this has been an issue since the upgrade
[17:05] <arrrghhh> I went from 12.04 to 14.04 last summer
[17:06] <RoyK> arrrghhh: please pastebin /etc/rsyslog.conf and the files under /etc/rsyslog.d
[17:06] <arrrghhh> http://paste.ubuntu.com/12886986/
[17:06] <marka13> Does anyone know if Landscape/Autopilot have their own channel?
[17:07] <arrrghhh> http://paste.ubuntu.com/12886992/
[17:08] <arrrghhh> http://paste.ubuntu.com/12886994/
[17:08] <arrrghhh> http://paste.ubuntu.com/12886996/
[17:08] <arrrghhh> http://paste.ubuntu.com/12886998/
[17:10] <RoyK> arrrghhh: try to stop rsyslogd and start it manually with -d
[17:11] <arrrghhh> hm stopping the service and the process is still running... kill -9 tiem
[17:11] <arrrghhh> marka13, I just searched, there is #ubuntu-autopilot
[17:12] <arrrghhh> RoyK, http://paste.ubuntu.com/12887022/
[17:13] <RoyK> arrrghhh: check lsof again
[17:15] <arrrghhh> hm looks about the same...
[17:15] <arrrghhh> http://paste.ubuntu.com/12887042/
[17:15] <RoyK> can you try to move the mythtv file aside and restart rsyslogd?
[17:15] <arrrghhh> interesting that rsyslog is still running but it stopped putting text into the terminal...
[17:16] <arrrghhh> RoyK, so mv /var/log/mythtv/mythbackend.log 'somewhere else'?
[17:17] <RoyK> arrrghhh: /root or something
[17:17] <RoyK> erm
[17:17] <RoyK> no
[17:17] <RoyK> not the log
[17:17] <arrrghhh> /tmp?
[17:17] <RoyK> the one in /etc/rsyslog.d/
[17:17] <arrrghhh> oh 10-4
[17:17] <RoyK> arrrghhh: don't use /tmp, some distros wipe that from time to time or even use tmpfs, so never store anything valueable there
[17:18] <arrrghhh> alright put it in ~
[17:18] <RoyK> (which probably is /root)
[17:18] <RoyK> make sure the process is dead before you try to start things again
[17:19] <arrrghhh> http://paste.ubuntu.com/12887066/ startup
[17:19] <arrrghhh> yea I kill -9'd it
[17:19] <arrrghhh> lsof http://paste.ubuntu.com/12887068/
[17:19] <RoyK> run 'logger "what's happening to my system???"'
[17:20] <arrrghhh> syslog is still 0b
[17:20] <arrrghhh> I assume that should insert to syslog?
[17:20] <RoyK> it should indeed
[17:20] <RoyK> Oct 21 19:20:46 roysk roysk: what's happening to my system???
[17:21] <arrrghhh> ya empty... looking at perms, is this correct?
[17:21] <arrrghhh> -rw-r-----  1 root      adm       0 Aug  8  2014 syslog
[17:21] <arrrghhh> shouldn't it be syslog adm?
[17:21] <RoyK> it's root:adm here
[17:21] <arrrghhh> hm.  File hasn't been touched since 8/2014 :/
[17:22] <RoyK> remove the log files and try to restart rsyslogd again
[17:22] <arrrghhh> k
[17:23] <RoyK> (or move them away)
[17:24] <arrrghhh> yay lsof shows it's accessing them
[17:24] <RoyK> perhaps time for a little fsck ;)
[17:24] <arrrghhh> and it appears to have a non-zero size
[17:24] <arrrghhh> ehm
[17:24] <arrrghhh> it shows syslog:adm now... probably because of that change I made
[17:24] <arrrghhh> -rw-r-----  1 syslog    adm     397 Oct 21 11:23 syslog
[17:25] <arrrghhh> Oct 21 11:23:28 nas rsyslogd-2039: Could no open output pipe '/dev/xconsole': No such file or directory [try http://www.rsyslog.com/e/2039 ]
[17:25] <arrrghhh> that may be an issue....
[17:26] <arrrghhh> but auth.log is working, kern.log
[17:26] <arrrghhh> what else did I delete other than syslog
[17:26] <arrrghhh> well those are the main ones I care about.  odd issue... thanks RoyK !
[17:27] <RoyK> arrrghhh: I suspect an old logrotate config may change the user to root after its run
[17:28] <arrrghhh> https://bugs.launchpad.net/ubuntu/+source/rsyslog/+bug/940030 << per that bug I changed /etc/logrotate.d/rsyslog
[17:28] <arrrghhh> I added create 640 syslog adm
[17:28] <RoyK> arrrghhh: ok
[17:28] <RoyK> arrrghhh: at least now you know where the error was - if it happens again, please call
[17:29] <arrrghhh> :) thanks!!!
[17:29] <RoyK> np :)
[17:48] <marka13> does landscape/autopilot have their own channel?
[17:49] <arrrghhh> marka13, I just searched, there is #ubuntu-autopilot
[17:49] <arrrghhh> (fyi I said that almost an hour ago :P)
[17:50] <marka13> sorry didn't see it
[17:50] <arrrghhh> np, I did ping you hoping you would
[17:50] <arrrghhh> it was mixed in with my troubleshooting mess
[17:50] <marka13> ya saw that
[17:50] <marka13> and I've been in TS hell myself  ;-)
[19:51] <atralheaven_> Hello, I need someone outside of china, to test my openvpn connection. I can't connect to it I want to know if it is because of country firewalling
[19:52] <rbasak> I'd ask someone you know personally! Do you really want some stranger connecting to your VPN? Similarly, I don't want to connect to some stranger's VPN, as a malicious one could do weird things to my system.
[19:54] <jpds> atralheaven_: You know that the great firewall doesn't block SSH at all?
[19:55] <atralheaven_> rbasak: I just want to see if it connects, and the ip is right, you don't need to browse with it! and I will revoke the certs so its not a problem
[19:55] <atralheaven_> jpds: Im using it now, but I need openvpn, it was working.... now it doesn't
[19:56] <jpds> atralheaven_: Why do you need openvpn when you have working ssh?
[19:56] <atralheaven_> jpds: openvpn is faster and works easily everywhere.
[19:56] <atralheaven_> may just someone help me to test it :(
[20:06] <atralheaven_> is it something risky?
[20:08] <rbasak> atralheaven_: an OpenVPN server can send arbitrary routes that the client will put into its routing table, for a start.
[20:09] <rbasak> It might be possible to disable that at the client end, but I wouldn't recommend that anyone connect to an untrusted OpenVPN server for security reasons.
[20:09] <atralheaven_> rbasak: what about testing on a vm?
[20:10] <rbasak> Also the server might be able to route via a client, too.
[20:10] <rbasak> So nobody can be sure that you aren't intending to use the client's Internet connection for nefarious purposes.
[20:10] <rbasak> That applies to a VM, too.
[20:11] <atralheaven_> rbasak: what about connecting just for a second? to see it connects, and doesn't fail
[20:11] <rbasak> Ask someone you trust.
[20:11] <rbasak> (and who trusts you)
[20:11] <atralheaven_> if I had someone outside of here who could help me I wouldn't come here :|
[20:13] <rbasak> Buy some instance time on EC2 or something.
[20:13] <atralheaven_> if I run ssh socks on my system, and then connect to openvpn, if the problem be from the country firewalling, should I be able to connect?
[20:15] <atralheaven_> for using openvpn with obfsproxy, do I need 2 servers? I have one vps
[20:46] <ponyofdeath> anyone have any ideas why using ssh -f -N -D 8080 to set up a tsocks proxy times out after tyring to use it?
[20:46] <ponyofdeath> i have tsocks and the tunnel set up