jpastorehi I'm trying to run tshark remotely over ssh to pipe to my local wireshark, I'm getting an error that I do not have permission to the device? I've created the wireshark group, added myself to it, reconfigured the package, and I get a permission error relating to the device. How do you grant permission to a device? There is nothing in udev either which is strange...02:24
hallyn_smb: arges:  hm, is libvirt on xenial working for you? I get03:04
hallyn_Feb 16 19:03:53 sl libvirtd[2705]: Cannot set property Before, or unknown property.03:04
hallyn_Feb 16 19:03:53 sl libvirtd[2705]: Cannot open log file: '/var/log/libvirt/qemu/cdboot.log': Device or resource busy03:04
hallyn_(which i've seen mentinoed in a bug somewhere)03:05
hallyn_iirc it was a systemd machinectl problem03:05
hallyn_hm, i'l lhave to look into it tomorrow03:05
=== kickinz1|afk is now known as kickinz1
=== lordieva1er is now known as lordievader
=== lordievader is now known as Guest98681
smbhallyn_, Not sure I had been trying to do boots from CD. I'll check. Sounds a bit like either missing file or something wrong with aa profile(s)08:54
=== Guest98681 is now known as lordievader
lordievaderGood morning.08:56
Seven_Six_Twonot sure what happened, but most of my server's services stopped connecting. znc, apache, owncloud09:03
Seven_Six_Twoand oddly, my server and desktop are connected to the same router, and my desktop can ssh in to the server, and the desktop can browse the web, but my server can't ping google.09:04
Seven_Six_Twothe default gateway for both machines is the same09:05
SlingSeven_Six_Two: how are you trying to ping google?09:11
Seven_Six_Twoping google.com09:12
SlingSeven_Six_Two: try ping
Slingit that works, DNS is probably not working for you09:12
Seven_Six_Twothat doesn't work09:12
Slingwhat happens when you try?09:13
lordievaderDoes pinging the gateway work?09:13
Seven_Six_Twothat's so messed up. It's not doing anything09:13
Seven_Six_Twoyes, I get a response from the gateway's lan ip09:13
lordievaderSeven_Six_Two: What is the output of 'ip r'?09:13
Seven_Six_Twodefault via dev eth009:14
Seven_Six_Two10.8.0.0/24 via dev tun009:14
Seven_Six_Two10.8.0.2 dev tun0  proto kernel  scope link  src
Seven_Six_Two169.254.0.0/16 dev eth0  scope link  metric 100009:15
Seven_Six_Two192.168.1.0/24 dev eth1  proto kernel  scope link  src
Seven_Six_Two192.168.168.0/24 dev eth0  proto kernel  scope link  src
ubottuFor posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic.09:15
lordievaderI suppose you do dhcp?09:16
Seven_Six_Twoyes, static dhcp from the router09:17
lordievaderDoes the gateway run a firewall which is dropping icmp by any chance?09:18
Seven_Six_TwoI believe so09:18
Seven_Six_Twowell, it's not set to ignore or forbid, but icmp-flood attack filtering is enabled09:19
Seven_Six_Twooh dammit. that's it, isn't it. how did you figure that? the only thing I've changed recently is stats in the router, and didn't notice that icmp-flood filtering is only enabled with stats.09:20
lordievaderI guess I can conclude from this that you have access to the gateway? Can you ping from there?09:20
Seven_Six_Twocorrect. ping from router works09:21
lordievaderOdd. Well perhaps it is that. Can you ping from the problem host when the icmp-flood filtering is disabled?09:21
Seven_Six_Twooh udp-flood and tcp-syn-flood filtering too09:22
lordievaderThat shouldn't matter for ping ;)09:23
Seven_Six_Twolol. and my server is on blocked list09:23
Seven_Six_Twoin router. no it's like my router is seeing too many icmp packets and cutting off the server09:23
Seven_Six_Twoone sec though. I'll disable and test09:23
Seven_Six_Twoyeah, as soon as I remove the server ip from the blocked list, it starts working again.09:24
Seven_Six_Twoso the router is seeing too many of some packet type09:24
lordievaderHeh, and the firewall strikes once again ;)09:25
Seven_Six_Twothanks for the help. I made some changes, and know where to look if it happens again.09:26
Razvahey! is there any know Juju bug regarding network interfaces not named eth?09:40
lordievaderRazva: What version of Ubuntu is this?09:46
ikoniagoing to need to be more specific than "a bug regarding network interfaces not named eth"09:49
=== uaa is now known as Guest14860
=== Guest14860 is now known as damascene
=== IdleOne is now known as Guest13578
=== rxc is now known as Guest60283
=== shirgall is now known as Guest4271
=== damascene is now known as Guest58378
=== Guest58378 is now known as damascene
YamakasYhow large is the size of a mirror, without sources, for 14.04 ?10:14
YamakasYonly 64bit btw10:14
Razvalordievader 14. should we use 15?10:16
lordievaderRazva: No, I was asking since 15.10 uses a newer version of udev which uses bios dev names instead of the old naming scheme. Perhaps the udev of Trusty got updated to the same version though.10:20
lordievaderRazva: What kind of network names do you have now?10:20
smbhallyn_, As a follow-up on libvirt: I can boot kvm guests via libvirt. I realized that "cdboot" would be your guests name. I'd check whether /var/lib/libvirt/qemu is a directory owned by libvirt-qemu.kvm and writable at least for that user.10:42
Razvalordievader VLAN10:45
lordievaderRazva: What? We were talking about interface names right?10:46
=== 16WAAV1BC is now known as shauno
=== Lcawte|Away is now known as LewisCawte
=== extrememist is now known as extrememist|pc
=== extrememist|pc is now known as extrememist
Razvalordievader well it might be, I have no idea, but what I know is that with eno1 juju basically fails11:13
lordievaderRazva: Well look on Launchpad if this has been reported before, else make a new bug report.11:14
=== kickinz1 is now known as kickinz1|afk
=== jamespag` is now known as jamespage
=== mfisch is now known as Guest22953
=== shirgall is now known as Guest5617
=== med_ is now known as Guest37570
=== RickyB98|BNC is now known as RickyB98
=== ChanServ changed the topic of #ubuntu-server to: Ubuntu Server discussion and support | For general (not server specific) support, try #ubuntu | IRC Guidelines: https://wiki.ubuntu.com/IrcGuidelines | https://wiki.ubuntu.com/ServerTeam/GettingInvolved | Docs and resources: https://help.ubuntu.com/14.04/serverguide/ | Updates for CVE-2015-7547 have been released: http://www.ubuntu.com/usn/usn-2900-1/
=== Ursinha_ is now known as Ursinha
=== jelly-home is now known as jelly
=== jelly is now known as jelly-home
=== Guest22953 is now known as mfisch
=== mfisch is now known as Guest32940
spm_dragetSilly question, but after installing a fresh ubuntu-server and assigning a static IP and nameserver… where is the nameserver saved? Should it be in /etc/netwrok/interfaces… because I do not see it there…?15:13
freeone3000I have a Lenovo ThinkServer running 14.04 LTS. I'm trying to get the health of its primary drive raid, but I get: https://gist.github.com/freeone3000/4cc6e1633f22796de875 . (The server has a hardware raid controller with two SATA drives connected as SAS)15:21
jelly-homefreeone3000: which pci device is that controller?  lspci -nn15:24
freeone3000jelly-home: "04:00.0 RAID bus controller [0104]: LSI Logic / Symbios Logic MegaRAID SAS 2108 [Liberator] [1000:0079] (rev 05)"15:26
spm_dragetAh okay, /etc/network/interfaces is the correct thing15:30
dami0hi, i'm trying to test out SingleSignOn was going through the instructions and just wanted to ask about faking the domain names, if i put the server down in the /etc/hosts files, is that loopback for the server itself and the server's LAN ip on the clients?15:31
jelly-homefreeone3000: see if mpt-status works for that one15:32
freeone3000jelly-home: Always prints "ioctl: no such device", but man pages say it doesn't take an argument and should autodetect.15:34
freeone3000Did the autoload bit, tried specifying '-u /dev/sda' and '-u 0:40' but no change.15:36
jelly-homeno idea then, perhaps there are backports of newer smartmontools for trusty somewhere15:36
ivoksjelly-home: look at you :)15:36
jelly-homeyeah, ECHAN15:43
* jelly-home felt sorry for the LSI user15:43
leitaojgrimm, cyphermox: Any idea if docker on 16.04 will be compiled with golang (other than gccgo) for ppc64el?16:11
jgrimmleitao, the plan is to build docker1.10 with golang-1.6 for most architectures including ppc64el16:13
jgrimmleitao, do you have a preference on technical merits?16:13
leitaojgrimm, that is what we need. Thank you!16:14
leitao1.9 seems to have some issue16:14
leitaojgrimm, btw, any chance to move docker from universe to main?16:14
jgrimmleitao, no plans to do so16:15
leitaojgrimm, it means docker is not supported by Canonical, right?16:16
jgrimmleitao, community supported is the way its described -> https://help.ubuntu.com/community/Repositories/Ubuntu16:18
arrrghhhanyone have suggestions for something like Crashplan which has a Linux client?  I really liked Crashplan, until they decided to change their pricing structure and doubled everything since I last renewed 4 years ago...16:21
nacchallyn: not sure if you saw, but i updated the numactl merge reuqest yesterday16:29
naccif you could ack, that'd be great16:29
freeone3000arrrghhh: tarsnap16:31
arrrghhhfreeone3000, hm that's one I have not heard of before16:32
=== Guest32940 is now known as mfisch
hallynnacc: what's the url again?16:58
hallynsmb: uh, no, nothing to do with cds.  my old windows vista vm also doesn't start16:58
nacchallyn: one sec, https://code.launchpad.net/~nacc/ubuntu/+source/numactl/+git/numactl16:58
naccmerge branch16:58
hallynsmb: "Cannot set property Before, or unknown property." is the critical line I think.  (or I'm wrong and it's a red herring :)17:00
smbhallyn, hm odd. me and arges did notice no such errors.17:00
argeshallyn: is this only with windows guests?17:00
hallynarges: no17:01
hallynarges: smb: it probably ha snothing to do with libvirt itself, though, bc it seems to be a dup of https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/152907917:02
ubottuLaunchpad bug 1529079 in systemd (Ubuntu) "Can't start virtual machines after upgrade to Xenial" [High,Confirmed]17:02
argeshallyn: i take a look.. my uvtool vms have been working fine17:03
hallynnacc: thx, lemme do that right now17:03
smbarges, and I was using virt-manager. Only difference is that the host is a fresh Xenial install17:05
nacchallyn: thanks!17:07
hallynsmb: hold on;  fresh install did work, or did not work?17:08
smbhallyn, fresh install does work for me17:09
smbhallyn, you and the other reporter seem to use upgraded hosts17:09
hallynsmb: yup17:09
argesok yea I have been using fresh xenial installs17:09
argesso we need some upgrade testing I suppose17:09
hallynwell that gives ideas of where to start looking at least, that's good :)17:10
hallynbugs.debian.org is down?17:10
hallyncan the bot tell me about debian bug 814952 ?17:10
ubottuError: Could not parse data returned by Debian:17:10
hallynnacc: is debian bug 814952 the one you filed ?17:10
smbhallyn, and since the other bug was opened last year it already happened with an earlier update of libvirt ...17:11
smbhallyn, hm you happen to use special args for libvirtd (which might be ignored by the old service conversion)? Though I do read the right file now17:12
hallynsmb: yes but it didn't hapen until i upgraded to xenial.  i *think*17:13
hallynno special args,17:13
hallynroot      2705     1  0 Feb13 ?        00:00:00 /usr/sbin/libvirtd17:13
smbok, well still allows for breakage with the xml definitions which would not be new17:14
smbhallyn, have you tried to virsh edit the config and do a small change then write it back17:15
smbvalidation is done for writes... though the error messages for that used to be poor17:16
hallynnacc: LGTM, wish i could see the debian bug, but th17:19
hallynnacc: did you do a merge request against anything?  Should I just push the package to x?17:19
argeshallyn: can you dumpxml the faulty vm and pastebin it somewhere17:21
hallyni'll object to "the faulty vm" :-)17:22
hallynarges: just a min17:22
argeshallyn: well the one that libvirt cant parse17:23
argesand perhaps there is a qemu log somewhere to see if that cmdline it produced looks messed up17:23
hallynarges: it can't parse any of several I had.  I deleted all but one, and created one new one.17:24
hallynarges: the logfile has : libvirt:  error : libvirtd quit during handshake: Input/output error17:25
hallynnacc: hm, having trouble building a source pkg17:26
hallynwhy is debuild being stupid and trying to include .git?17:27
hallynthere msut be  a better way to build than to mv .git .. first17:28
hallynthat is to ohacky17:28
argesok i'll look17:30
hallynnacc: ffs can't you file a debian bug to fix the pkging to be 3.0 quilt? :)  kills me17:33
hallynanyway, pushing, thx17:33
teward|rbanffy: ping, if you're around and can give some guidance on the nginx package, that'd be great - gotta pick between two evils :/17:42
teward|rbasak: ^17:42
teward|sorry rbanffy for the mishighlight17:42
nacchallyn: :) i can do that for the next release17:43
rbasakteward|: around but otp. Do you want to provide a summary here please?17:44
teward|rbasak: FTBFS as is, we have to update the Lua moduel in -extras.  Two options: nitpick the fix from upstream which fixes the missing udp connector struct, or take from Debian's git repository in where they update the entire module to the release candidate version on openresty's github pages, which brings in additional changes to the Lua module (I don't have a complete list)17:45
teward|rbasak: not sure which is the correct approach, though i can get either done before FeatureFreeze once I get some guidance17:45
teward|Debian's updating of the package also incorporates the fix, apparently, though that will undergo its own build tests as well17:46
teward|rbasak: basically, i'm not sure which approach will work better, hence seeking some advice :/17:49
rbanffyNP, teward. It happens all the time.17:51
=== Guest86349 is now known as IdleOne
hallynsmb: arges: all right, sorry, focusing on the libvirt thing now18:15
hallynoh, one more thing first,18:15
argeshallyn: no problem, ive downloaded the xml files from you and the bug report, trying to look at what the differences are and repo on my end18:15
argeshallyn: ok no problem18:15
=== Guest91003 is now known as karstensrage
hallynarges: saving xml works fine, sothat's not the problem18:17
hallynlemme set up two vms, one with upgrade and one with clean install, and compare...18:17
argeshallyn: yea that's what I was trying to do, but hopefully that points us in the right area. wondering if its osmething to do with how the disk controllers are setup18:18
argesthat would be my swag18:18
=== Luke_ is now known as Luke
hallynarges: wtf.  failed to reproduce in a vm18:48
argeshallyn: yea i'm not having luck here either.18:50
argeshallyn: what did you upgrade from? wily?18:51
hallynarges: yeah, maybe i need to start from vivid, which my laptop apparently did18:54
hallyn(though not directly)18:54
hallyni'll try from trusty - upstart->systemd seems more likely to e a trigger18:54
argeshallyn: ok18:55
hallynarges: i think when we find the cause we're gonna laaaaaugh19:01
argeshallyn: : ) progres?19:02
hallynwiating on the trusty vm install19:03
hallynoh it's done19:03
argeshallyn: grabbing lunch... keep me posted19:07
rbasakteward: extras is in universe, right? So no security team main concern about whether to follow Debian's import or take from nginx upstream.19:13
tewardrbasak: correct - extras is in Universe19:14
tewardrbasak: though, once dynamic modules are applied, we're going to have to scramble to make sure some are in Main, some aren't19:14
tewardbecause it's not nice when that's taken into account19:14
rbasakteward: my remaining question would be about user visible differences. Is there a cost to committing to one over the other now? IOW, will any user have a reason to care if you take one approach now and switch later?19:14
teward(that's the future though)19:14
tewardrbasak: AIUI, no, there won't be any major reasons for Users to care.  However, one reduces the plugin-specific future merge delta by one tiny commit, the other reduces it to zero when merging in from Debian19:15
rbasakIf that's not a concern then I'd pick the easiest thing, which I guess you get to trade off. Following Debian is nice, but the benefit depends on how much we've diverged in this aresa.19:15
rbasakOTOH, cherry-picking a trivial patch may be far easier.19:15
rbasakThat's what we'd probably want to do after FF anyway.19:16
tewardrbasak: they're both available on git - patches are nice anyways :P19:16
tewardthough, every new release of nginx will have to get an FF exception19:16
tewardwhich is fine, but it's one of those annoying little headaches :)19:16
rbasakI thought we had an understanding with the release team on the plan? Or did that not get fully resolved?19:16
rbasakKeep ubuntu-release advised, certainly. I'm not sure if you need an explicit and formal FFe approval for each step though.19:17
tewardi'm not sure we did since the policy change, the TB discussion was one fo "everything still gets standard FF review before upload" iirc19:17
tewardrbasak: well, from this point on, each nginx will likely have some minor feature changes19:17
tewardso it'd need FFe if standard policy is followed19:17
rbasakAh, OK.19:17
tewardplus HTTP/2 bugfixes which don't affect the builds currently19:17
teward(because it's not enabled)19:17
tewardrbasak: i'm going to nitpick the minor change form upstream that enables building and doesn't implement an rc version of the third party module19:18
tewardseems like the lesser evil19:18
rbasakOK, that sounds good. Thank you for considering carefully, as always!19:19
tewardthat's my job :)19:19
tewardand thank you for being a sounding board and for the ongoing advice19:19
tewardit's always nice to bounce opinions off others first :)19:19
rbasakI forgot the other option was rc19:19
rbasakThat makes your choice sound even better.19:19
hallynarges: hm, not reproduced in trusty vm either.  so i guess i'll have to look more closely at my own setup.19:40
argeshallyn: yikes, ok19:41
=== jelly-home is now known as jelly
hallynarges: smb: systemd-machined, provided by systemd-container, is causing the problem21:42
argeshallyn: so your hunch in the bug earlier was corrupt21:46
hallynaccuse me of corruption?   I'll sue!21:46
hallynyeah i just don't get why systemd-machined woudl ship in systemd-container.  that's weird21:47
hallynmind you the bug may still actually be in libvirt21:47
hallynbut i find that hard to believe, as we're on 1.3.x21:47
hallynanyway with that found i'll lower it's priority and defer it :(21:47
argescool hopefully we can catch it before release21:51
jak2000are you there?22:35

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!