/srv/irclogs.ubuntu.com/2013/04/23/#ubuntu-server.txt

delinquentmeis there a quick and easy way I can find out how much RAM a machine is capable of running  + how much it currently has?00:00
delinquentmevia command line?00:00
sarnolddelinquentme: sudo dmidecode -t memory    will probably show you what you need00:01
delinquentmeYeah I just ran it ... is it consistent?00:02
delinquentmebc I dont believe the output I'm getting00:02
sarnoldit's always worked for me..00:02
delinquentme90% sure I've only got 2 slots ... it tells me I've got 400:02
sarnoldbut I've only had a handful of motherboards, compared to the vast numbers available...00:02
delinquentmeIt's important to note that Dmidecode reports system hardware information as described in the BIOS and does not scan your hardware, so in some cases the output can be wrong.00:03
sarnoldheh, could be :/ you can always compare against crucial.com's database00:03
delinquentme^00:03
sarnolddelinquentme: you could also try sudo lshw -c memory -- I don't see a "maximum" though..00:04
adam_gjamespage, verification-done for that openstack folsom batch. http://people.canonical.com/~agandelman/2012.2.3-verification/   first thing tomorrow,  gonna update the meta-bug to include all new change log entries, tag verification-done where needed and post logs01:00
adam_gjamespage, i noticed though that none of the rebased packages that made it into proposed were built with 'debbuild -v$prev_proposed_package', so bug references from the original uploads might be a bit screwy / missing in the sru tracker?01:02
adam_gjamespage, ttyt01:02
=== wedgwood is now known as wedgwood_away
Arrickhow does one force ssmtp to send email as an address such as noreply@domain.com02:10
JeudiCan UFW be used to set up a network firewall on a soho LAN?  UFW 0.29.3-1.04:41
=== security is now known as baba
=== smb` is now known as smb
=== Jikan is now known as Jikai
=== Jikai is now known as Jikan
jamespageadam_g, great - yeah - the rebase without -v was my mistake - Daviey is aware08:11
Davieyjamespage: strictly, it wasn't your mistake - as we checked that lack of -v was infact OK08:12
Davieybut I think any uploads i do, i will still do -v :)08:12
_rubenok .. this is mighty annoying .. got a fresh vm with heartbeat installed, but for some reason it refuses to create /var/run/heartbeat/register (or it does create but gets removed again), imparing its connectivity stuff08:35
koolhead17burp..08:58
=== baba is now known as cod3r
no0tichi10:47
no0ticI'm trying to understand the basics of openstack and would like to set up a test environment using one or two IBM blades. I read various articles on openstack docs and ubuntu docs, and I am trying to follow https://help.ubuntu.com/community/UbuntuCloudInfrastructure. I don't get where hypervisors are concerned. Where will the vm instances run? I have a little experience with cloudstack,10:56
no0ticwhere an external controller commands multiple hosts with hypervisors like xen or esx and starts user vm instances on various hosts. I would like to use XCP as the basic layer of my environment but I don't know where to start from.10:56
andygraybealno0tic, i only use kvm11:25
andygraybealso i'm not much help :) .... libvirt/kvm :)11:25
andygraybealit's super easy.. i don't know any of the words you used ;)11:25
=== highvolt1ge is now known as highvoltage
=== highvolt1ge is now known as highvoltage
Davieyjamespage / smoser: I just noticed raring-server-amd64.iso is 707MB.. we wanted to keep it sub 700. Can one of you deal?12:48
jamespageDaviey, cjwatson is tweaking the shipped seed to reduce that12:49
jamespagewe discussed in release earlier12:49
jamespageDaviey,  (python-svn + miscfiles + openssh-blacklist*) got the chop12:50
Davieyjamespage: bah, i am 5 steps behind today. :)12:50
jamespageDaviey, yeah - my jetlag is bad as well12:50
=== cod3r is now known as Guest34392
andygraybealhey guys, i need some handholding with app armor ... i'm reading here: https://help.ubuntu.com/10.04/serverguide/apparmor.html13:49
andygraybeali need to give libvirt access to eth113:49
andygraybeali have a /etc/apparmor.d/libvirt ... but i'm afraid of anything13:50
andygraybealerr.. welll i have no clue13:50
rbasakandygraybeal: have you read https://wiki.ubuntu.com/DebuggingApparmor#Fixing_profile_bugs? If you can get apparmor complain messages logged, then aa-logprof walks you through profile modifications to permit those actions. I've only done it once and can't really help you further though.13:52
andygraybealrock on13:52
andygraybealthank you13:52
andygraybeali'm eyeballing this: usr.sbin.libvirtd file right now :)  and i will read that guide you posted.13:53
andygraybeali can't break anything, this is my mail and www server :)13:54
andygraybealrbasak, i don't think there is a bug, i think there is a configuration problem13:56
andygraybealbut maybe yuo know this too13:56
rbasakandygraybeal: sure - you might want to do something different from the default configuration. But either way you need to get a modified profile on your system, so to that point the same documentation should suffice for both cases. See the last bit (about /etc/apparmor.d/local/) of that section for details of how to apply this as a local configuration modification only.13:58
jdstrandandygraybeal: you might be interested in http://wiki.apparmor.net/index.php/Libvirt14:02
=== wedgwood_away is now known as wedgwood
andygraybealthank you jdstrand14:05
andygraybeali think the problem might be that i'm just an idiot14:05
andygraybealand i have a stupid network configuration that bridges my eth1 and i didn't realize it till now14:05
andygraybealand not an apparmour thing14:05
andygraybealbut this libvirt apparmor thing is nice!14:05
andygraybealthe document14:05
andygraybealaah jdstrand, ah neat your name is in the article!14:06
andygraybealprofessional turd polisher, nice.14:06
jdstrandheh, yes. taglines are fun14:07
Skaaghello guys, I have a problem upgrading my kernel:14:37
Skaaglinux-server : Depends: linux-image-server (= 3.2.0.38.46) but 3.2.0.40.48 is installed14:37
SkaagI tried apt-get -f install, it doesn't help14:38
beanSkaag: 38.46 is earlier than 40.48... no?14:39
ogra_Skaag, what command did you use exactly ?14:41
ogra_note that apt-get upgrade will neither remove existing packages nor install new ones ... you need to use apt-get dist-upgrade in that case14:42
SkaagI see14:44
SkaagI was just trying to upgrade actually14:44
Skaagi'll try a dist-upgrade and see if it yields different output14:44
ogra_tight, you want dist-upgrade14:44
ogra_*right even14:44
Skaagdoes the channel have an official pastebin you guys like?14:45
andygraybealaah okay... so it does seem to be app-armor14:46
andolSkaag: http://paste.ubuntu.com might be appropiate :)14:46
=== chuck__ is now known as zul
Skaaghttp://paste.ubuntu.com/5595740/14:53
Skaagboth dist-upgrade and -f install, fail to fix the issue14:53
SkaagI may have to remove something first, but what?14:54
ogra_did you apt-get update first ?14:54
ogra_to update the package lists14:54
SkaagI believe I did. I will feel really stupid if I didn't. hold on.14:55
Skaagsame problem14:55
Skaag(after updating)14:55
Skaaglet's see if aptitude can do better14:58
=== rickspencer3_ is now known as rickspencer3
beanSkaag: it looks liek you're trying to install an older kernel than what you have.15:09
Skaagyes15:11
Skaagbut how do fix this? :)15:11
Skaagaptitude is going forward, so far15:11
Skaaghopefully it will manage to overcome this15:11
andygraybealjdstrand, well i found that the apparmor profile is set to allow me to have rw access to my eth1:   "/sys/devices/pci0000:00/0000:00:1c.1/0000:05:00.0/0000:06:00.0/resource" rw,15:17
andygraybeal  "/sys/devices/pci0000:00/0000:00:1c.1/0000:05:00.0/0000:06:00.0/config" rw,15:17
andygraybeal  "/sys/devices/pci0000:00/0000:00:1c.1/0000:05:00.0/0000:06:00.0/resource0" rw,15:17
andygraybealdo i need more than this?15:17
andygraybealthat is in the specific file for the libvirt machine, and it says the file is auto created with the libvirt process15:17
jdstrandandygraybeal: do you have any apparmor denials in /var/log/kern.log on the host?15:20
andygraybealjdstrand, yes, lemme post to that pastie bin15:21
andygraybeali don't understand this devils talk15:21
andygraybealthis is the device: /sys/devices/pci0000:00/0000:00:1c.1/0000:05:00.0/0000:06:00.0/vendor"15:22
andygraybealthis is the pastie: http://paste.ubuntu.com/5595791/15:23
andygraybealsomething about requesting a mask and then getting denied the request?15:25
jdstrandI think that is harmless, but can you shutdown the vm, then add to /etc/apparmor.d/libvirt/libvirt-c2059dbd-b96a-bb40-3c11-6dd10877acef the following:15:25
jdstrand/sys/devices/pci0000:00/0000:00:1c.1/0000:05:00.0/0000:06:00.0/vendor r,15:25
jdstrandthen start the vm15:25
andygraybealaah 'vendor'15:26
andygraybealinside the brakets?15:27
jdstrandyes15:28
jdstrandunder the '#include ...' lines15:28
peteyhey how would i search for a specific file?15:29
peteylike an sql file15:29
andygraybealjdstrand, i'm moving forward, now i have another error... but no problem in my kern.log!15:32
andygraybeali will research the other error and will get back if i cannot figure it out.15:32
andygraybealthank you jdstrand!15:33
beanpetey: locate "name of file"15:33
Skaagstill no luck with that kernel upgrade15:33
andygraybeali think that solved that bit of the problem.15:33
beanSkaag: you mean kernel downgrade...15:33
Skaagaptitude fails the exact same way15:33
Skaagyes15:33
jdstrandandygraybeal: ok. I'm thinking it isn't an apparmor issue. however, it would be nice to address that denial. can you file a bug using 'ubuntu-bug libvirt' with exact steps to reproduce?15:33
andygraybealSkaag, how did you get into that state?  (i'm curious, not judging.. i'm like some guy living out of the trenches.. not knowing what is coming at me or what is going away!)15:34
andygraybealjdstrand, i'm in 10.04 ....15:34
andygraybealjdstrand, i would thikn everyone would scream to use 12.04 at me.15:34
andygraybealjdstrand, whatever you did.. cleared up the kernel.log!15:34
beanSkaag: "apt-get install linux-image-server=3.2.0.38.46 linux-image-headers=3.2.0.38.46" first maybe15:35
jdstrandyeah, that would probably not be fixed in an SRU. but you know how to silence the denial now15:35
andygraybealSRU ?15:35
Skaagtrying15:38
andygraybealjdstrand, this is the new error from my libvirt machien log:15:38
andygraybealhttp://paste.ubuntu.com/5595828/15:38
SkaagE: Version '3.2.0.38.46' for 'linux-image-server' was not found15:38
Skaagandygraybeal: I can only theorize: Many dist-upgrades, without rebooting after kernel updates, and then erasing older kernels from /boot/ because it was running out of space...15:39
andygraybealSkaag, that sounds like what i do!15:39
andygraybeali'm glad i'm not in your position.15:39
Skaagmy position isn't that bad... ;015:39
SkaagI mean everything is working, but I do want upgrades to continue working15:39
jdstrandandygraybeal: that is unfortunately not a useful error message due to the way libvirt reports errors. I suggest disabling apparmor for libvirt for now, then try to troubleshoot the problem. when you get it worked out, reenable apparmor in libvirt, and try again15:40
andygraybealjdstrand, how do i disable apparmor in libvirt?15:40
Skaagand where would my uptime be, if I rebooted after every new ubuntu kernel?! :)15:40
jdstrandit is in that wiki page. but the easy temporary way is to do: sudo /etc/init.d/libvirt-bin stop ; sudo apparmor_parser -R /etc/apparmor.d/usr.sbin.libvirtd ; sudo /etc/init.d/libvirt-bin start15:41
andygraybealthank you  jdstrand15:41
jdstrandyou can then verify it is disabled by doing 'virsh capabilities | grep apparmor'15:42
=== xcrracer_ is now known as xcrracer
andygraybealjdstrand, i typed in: " libvir: Security Labeling error : error calling aa_change_profile() " and went to the bug tracker and your name is all over it!15:43
andygraybealubuntu is so powerful because of this alone; being about to talk face to face to people like you.15:44
andygraybealit is great.15:44
jdstrandthat error message is potentially misleading. it could be just the last thing that went wrong before libvirt couldn't go any further with it15:44
peteybean: thanks15:44
andygraybealokay, i understand15:44
jdstrandthat improved in later releases, but with lucid, it can be interesting debugging libvirt15:44
jdstrandandygraybeal: heh, thanks15:45
beanpetey: ?15:45
beanoh the locate15:45
peteyhey its not working15:45
peteyim trying locate "xxx.sql"15:45
andygraybealjdstrand, yes, i am hoping to get this to work because it is the crux of my setup at work... to run LTSP from a virtual machine.15:46
andygraybealelse i have a lot of backpedaling to do!15:46
beanpetey: do "sudo updatedb" first15:46
andygraybealjdstrand, stopping libvirt like this will it kill my running virtual machines (my web and mail servers?)15:48
Skaagok so I did this: dpkg -r linux-server && dpkg -r linux-image-server15:48
Skaagnow dist-upgrade works15:48
andygraybealSkaag, congrats man!15:48
Skaagwill my server die on next reboot? :-)15:48
roaksoaxhalvors: around?15:48
roaksoaxerr15:48
roaksoaxsorry15:48
roaksoaxhallyn: around?15:48
hallynroaksoax: yeah15:49
SkaagI mean am I not supposed to be running a 'server' kernel?15:49
roaksoaxhallyn: would you know why i would get something like this: maas@canaima:/home/ubuntu$ virsh list --all15:49
roaksoaxerror: failed to connect to the hypervisor15:49
Skaaglooks like I'm running a 'generic' kernel right now15:49
roaksoaxhallyn: for that particular user, while with other users it works just fine?15:49
hallynroaksoax: bc that user isnot in group libvirtd?15:49
jdstrandandygraybeal: no, they shouldn't shutdown. it is possible the libvirt might lose track of them, but if you don't mess with the monitor file, I think it'll be ok. you could try just the apparmor_parser line. that may work15:49
roaksoaxhallyn: it has sudoers access to virsh15:50
andygraybealjdstrand, rock! thanks15:50
roaksoaxhallyn: however, this is an upgrade from quantal -> raring, and it stopped working in raring15:50
roaksoaxjamespage: did you have to do anything special for maas to be able to control virsh?15:50
roaksoaxvmass that is15:50
hallynroaksoax: what does 'sudoers access to virsh' mean?15:50
jamespageroaksoax, yes - I had to patch the power control template to use sudo15:50
hallynoh, you're running virsh as root through sudo?15:51
jamespageroaksoax, and add a sudoers config to let it do that15:51
roaksoaxjamespage: ahhh15:51
roaksoaxhallyn: that's it then :)15:51
roaksoaxhallyn: sorry for the noise :)15:51
jamespageroaksoax, patching the power control template would have been lost on upgrade15:51
roaksoaxjamespage: yeah...15:51
roaksoaxI'll commit that upstream15:51
jamespageroaksoax, I refer you to bug 107346315:51
uvirtbotLaunchpad bug 1073463 in maas "User-configurable files are not installed in /etc" [High,Triaged] https://launchpad.net/bugs/107346315:51
hallynroaksoax: ok, cool, ttyl :)15:51
jamespage:-)15:51
jamespagejust poking you!15:52
roaksoaxjamespage: that bug requires upstream support.. julian said he would work on it :)15:52
andygraybealjdstrand, apparmor is confirmed not running, by way of your virsh command and grp.  libvirt now says this: http://paste.ubuntu.com/5595856/15:52
andygraybeal*grp = grep15:52
andygraybealwhat the heck does "no IOMMU" mean?15:53
jdstrandandygraybeal: ok, so we've confirmed it isn't apparmor. unfortunately, I'm not familiar with the issue you are seeing with pci-assign15:53
jdstrandhallyn: have you seen that before ^15:54
andygraybealjdstrand, okay thank you for holding my hand up until now.  i appreciate it.15:54
andygraybeali will keep googling.15:54
hallynlooking15:54
hallynjdstrand: well, if you dont' have IOMMU you can't do pci assign right?15:55
jdstrandI guess not "If your hardware doesn't have an IOMMU ("Intel VT-d" support in case of Intel - "AMD I/O Virtualization Technology" support in case of AMD), you'll not be able to assign devices in KVM." - http://www.linux-kvm.org/page/How_to_assign_devices_with_VT-d_in_KVM15:56
jdstrandandygraybeal: ^15:56
andygraybealhm.. i have intel15:56
andygraybealxeon15:56
andygraybealmaybe that's not what they are talking about though15:57
andygraybeali pasted my kern.log in case it is helpful too : http://paste.ubuntu.com/5595870/15:57
jdstrandandygraybeal: the page I quoted has other info on how to see if your motherboard and/or cpu supports VT-d15:57
andygraybealokay cool15:57
andygraybeali assigned USB device on this same machine15:58
andygraybealwould that mean anything?15:58
andygraybeali don't think i have assigned PCI device yet.15:58
andygraybeali will read linux-kvm page15:59
andygraybealxeon is not listed with VT-d support  .... omg i want to cry16:00
andygraybeali'm still reading ti seems that just because xeon is not listed on that page doesn't mean it doesn't nhave VT-d support16:04
andygraybealah it's a function of the chipset.. too16:04
andygraybealit isn't often i run into a hardware limitation.....16:05
andygraybeallike this i mean.16:05
andygraybealjdstrand, well... i run this command: 'dmesg | grep -e DMAR -e IOMMU' and nothing comes up... so depressing16:08
andygraybeallooks like i need to do some peaking around my BIOS too...16:09
hallynandygraybeal: is this a network card you want to assign?16:09
andygraybealyes it is.16:09
andygraybealmaybe i just do a bridge...and leave it to that, eh?16:10
hallynyup16:10
andygraybealgood deal, not the end of the world then.16:10
andygraybealbut i think i'm going to take a walk around the block to chill out a bit.16:11
andygraybealokay back16:20
RoyKseems there's a bug with neted raids on ubuntu raring, probably on precise as well. similar is reported for mint13. if I create two RAID5s and then a RAID0 on top of those, the latter won't be started after a reboot. this works on lucid16:26
=== mrjazzcat_ is now known as mrjazzcat
InteliWaspif i install 13.04 beta 2 today, is there a way to update it to the official version after it's release?17:07
sarnoldInteliWasp: I'd expect 'apt-get dist-upgrade' to do a good enough job..17:24
Pici!final17:24
ubottuIf you install a development version of Ubuntu Raring and keep up with package updates, then you will be upgraded to the official release of 13.04 when it comes out. To make sure, type « sudo apt-get update && sudo apt-get dist-upgrade » in a terminal.17:24
InteliWaspok thanks17:29
=== baffle_ is now known as baffle
RoyKhttps://bugs.launchpad.net/ubuntu/+source/mdadm/+bug/117194517:34
uvirtbotLaunchpad bug 1171945 in mdadm "Nested RAID levels aren't started after reboot" [Undecided,New]17:34
roaksoaxjamespage: does this make sense to you? http://paste.ubuntu.com/5596176/18:03
adam_gjamespage, Daviey  verification-done for https://bugs.launchpad.net/ubuntu/+source/nova/+bug/1116671  and any other related ubuntu bug tasks18:11
uvirtbotLaunchpad bug 1116671 in cloud-archive "Meta bug for tracking Openstack 2012.2.3 Stable Update" [Undecided,Confirmed]18:11
roaksoaxsmoser: I'll be making this change for raring: http://paste.ubuntu.com/5596235/18:20
roaksoaxdo you agree?18:20
smoseronly that i dont think it shoudl be necessary18:20
smoserare you sure that template file is being rendered?18:20
smoserrather than another one?18:20
roaksoaxsmoser: yes, it is being rendered, i just dunno why it doesn't add what's required18:21
roaksoaxsmoser: give me a sec and I'll show you18:21
=== cod3r is now known as baba
jamespagezul: bug 116321820:11
uvirtbotLaunchpad bug 1163218 in nova "Typo in nova-cells upstart script" [Undecided,New] https://launchpad.net/bugs/116321820:11
zulyeah its fixed in bzr20:11
jamespagezul, coolio20:11
zulill do one last upload tomorrow20:12
jamespagezul, ack20:12
=== glebihan_ is now known as glebihan
cellofellowI'm trying to set up DKIM verification on Postfix. I have a site with several (~50) domains, and it'd be ideal to manage those domains on the mail servers' OpenDKIM using the DB, but I've not found any comprehensive information on actually doing that.21:16
=== broder_ is now known as broder
Davieyjcastro: watching your workshop video.. seemed you did a great job kicking it off21:27
jcastroDaviey: in hindsight I say "ummm" way too much21:29
jcastroit's like my crutch word21:29
Davieyjcastro: not as bad as you think.  I hadn't noticed until you said that21:30
sarnolduhoh, now Daviey won't be able to ignore it again :)21:31
jcastroit's this thing where you think a talk went awesome until you see yourself on video21:31
cellofellowjcastro: no need to be hard on yourself.21:31
jcastroand then it's like "God, what am I saying!"21:31
sarnoldjcastro: I'm lucky to have avoided that so far :) hehe21:31
Davieyjcastro: been there. :)21:39
Davieyjcastro: Could be worse, you could trip on a cable on the stage - and make a fool of yourself. :P21:40
jcastrothat was one of the best talks I've ever seen. It was awesome.21:40
andygraybealhow do i re-run the ldap client auth configuration?  .... dpkg-reconfigure .... ??21:47
andygraybeali tihnk i typed in the wrong host uri21:47
andygraybealoh oh i got it21:48
JeudiCan UFW be used to set up a network firewall for a small lan?21:52
jdstrandJeudi: yes, it can, though you have to know a bit about iptables syntax atm. see 'man ufw-framework' for details21:53
Jeudijdstrand: thanks.  I'll read that and see how I fare.21:54
=== wedgwood is now known as wedgwood_away

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