/srv/irclogs.ubuntu.com/2013/01/03/#ubuntu-server.txt

F3Speechlooking for some help with a networking error using cli - basic prob is usb wifi adaptor will only connect to network is i reset my router00:45
=== slank is now known as slank_away
=== cpg is now known as cpg|away
genii-aroundThis is my last game, need to get going after02:20
genii-aroundMisdirect :(02:20
=== cpg|away is now known as cpg
=== cpg is now known as cpg|away
Free99Hello everyone. Trying to make a high-availability KVM setup, but the only thing that is giving me trouble is figuring out how to use iptables03:10
Free99the setup is pretty simple, two machines that replicate the disk images for the VMs via drbd, and the VMs running on both machines for quick failover03:11
Free99issue is, there would be an ip conflict for each VM... any suggestions?03:11
Free99my idea is to use arptables and iptables to block off all of one of the machine's VMs while still letting DRBD and administrative access through03:13
sarnoldFree99: wild guess here, but could you solve your problem by giving each vm two interfaces? give both vms a private IP and one vm a public IP, and use linux-ha or similar tools to failover to the other VM and takeover the public IP on its other interface?03:15
sarnold(you may be able to do it entirely with one interface, but two may be easier to use existing tools..?)03:16
Free99sarnold: well... I don't really know anything about linux-HA.. I have DRBD setup already, bridged networking, etc.. I'd just like to block everything that isn't going to the host VM, then use a sysctl to disable iptables and arptables if the other host goes down03:18
Free99I guess I'll go check on Linux-HA, but that implies adding new interfaces to every single one of my VMs, right?03:19
patdk-lapheh, would be so simple to use linux-ha (pacemaker)03:20
patdk-lapdoesn't need to03:20
patdk-lapall depends on how you do your iptables config03:20
Free99oh wait, I'm using linux ha, specifically heartbeat! lol03:20
patdk-lapthat is really old03:21
Free99patdk-lap: to be honest, haven't gotten to that point yet, still doing the DRBD setup, stuck on the iptables though03:22
Free99oh I see, I should have said pacemaker03:22
Free99my bad03:22
=== cpg|away is now known as cpg
=== cpg is now known as cpg|away
la_need help understanding what im doing in Aptitude05:53
la_i think it just so simple i do not understand05:55
=== cpg|away is now known as cpg
blim_hi, Does anyone know a guide on how to set the ToS (type of service) on a client os?10:45
jabba_hello11:17
jabba_i just setup a crypted partition (SR for a Xen-Server). and added it to the /etc/crypttab with the option "timeout=30". sadly the option seems to be ignored (as boot-process keeps going on) and i am not able to enter the passphrase at boot. anyone an idea what's wrong?11:17
jabba_referring to this bugreport: https://bugs.launchpad.net/ubuntu/+source/cryptsetup/+bug/46820811:18
uvirtbot`Launchpad bug 468208 in cryptsetup "cryptsetup passphrase prompt at boot not working if waiting too long (w/o usplash)" [Undecided,Triaged]11:18
jabba_is it possible that it's an upstart relatet problem?11:19
xnoxjabba_: cryptsetup unlocking is done in initramfs while talking to plymouth (these days)11:20
xnoxjabba_: I didn't experience it timing out, but I did experience three attempts only.11:21
xnoxjabba_: is your root file system encrypted or some other partitions?11:22
jabba_no, the encrypted partition in an SR (Storage Repository) for a Xen Hypervisor (XCP-XAPI). It is essential, that  this device is decrypted before the XCP-XAPI-Service starts.11:27
jabba_-in +is11:27
jabba_so is it possible at all to halt the boot-process for a defined time (waiting for the passphrase) and in the case of a timeout countinue booting without decrypting the partition?11:29
jabba_why is the timout-option not listed in the crypttab-manpage anymore?11:31
jabba_xnox: any idea?11:37
xnoxjabba_: so when root is not encrypted, indeed upstart jobs unlock the partitions. it should be accepting options.11:38
jabba_xnox: so upstart is responsible for calling the cryptsetup scripts, BUT the main problem is, that the boot process doesn't get interrupted, for the defined time (timeout=30 in crypttab).11:45
=== mcclurmc_away is now known as mcclurmc
xnoxthe same way it does when rootfs is encrypted.11:48
xnoxjabba_: where did you find reference to crypttab "timeout" option?11:49
jabba_http://manpages.ubuntu.com/manpages/hardy/man5/crypttab.5.html11:49
xnoxjabba_: but that's for hardy. Are you running hardy? that option is not present in lucid and later.11:50
* xnox is not entirely sure how cryptsetup is managed in hardy.11:51
jabba_xnox: the timeout option i missing in precise... is there a new way to handle this?11:52
jabba_*is11:52
* xnox is looking into the history behind this: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=49550911:53
uvirtbot`Debian bug 495509 in cryptsetup "cryptsetup: timeout option does not work anymore" [Important,Fixed]11:53
xnox"* Completely remove support and documentation of the timeout option, document this in NEWS.Debian."11:55
jabba_can't believe it... that's really sad11:57
xnoxwell your problem seems to be that it's way too quick, instead of non-existing.11:57
jabba_xnox: could you explain this please?11:58
xnoxjabba_: we are processing udev triggers and we should have approx. udev settle timeout to enter the passphrase which is ~30s long. yet earlier you said "i am not able to enter the passphrase at boot" meaning that even if there was a prompt, it vanished too quick.12:00
jabba_ok.. i must admit: "i can't follow" :/12:02
jabba_xnox: my udev settles to fast?12:03
xnoxjabba_: I guess I need to experiment with a setup similar to yours. E.g. regular install and one non-essential luks encrypted mountpoint.12:03
jabba_and that's why the passphrase-promt timeouts to early?12:03
xnoxI am speculating a little bit. But i am not spotting anything that will block boot to capture passphrase-promt.12:04
xnoxfor non-rootfs luks volumes.12:04
xnoxbug 51896412:05
uvirtbot`Launchpad bug 518964 in cryptsetup "cryptsetup can't get input at boot." [Undecided,Confirmed] https://launchpad.net/bugs/51896412:05
jabba_so i am lost :(12:07
jabba_xnox: did you change the State of the bug-report?12:10
xnoxno, it's been there since 2010, I assigned it to myself to check if it can be fixed or not.12:11
jabba_xnox: you're my hero :) , thanks!12:11
=== yofel_ is now known as yofel
=== cpg is now known as cpg|away
zulyolanda: ping can you have a look at https://code.launchpad.net/~zulcss/python-cinderclient/fixtures/+merge/14175113:38
yolandahi13:39
yolandaok13:39
yolandadone13:44
zulyolanda: thanks one more https://code.launchpad.net/~zulcss/python-swiftclient/ftbfs/+merge/14175413:46
yolandadone13:51
zulyolanda: if you want to have a look at whats wrong with glance and quantum be my guest :)14:02
yolandazul, what's the problem? i saw a fbtfs in your other reviews, is that the same?14:03
zulyolanda: glance patches need to be refreshed or dropped and quantum its the testsuite14:03
yolandaok, i'll take a look14:04
=== lamont` is now known as lamont
=== slank_away is now known as slank
=== Ursinha-afk is now known as Ursinha
LovelidgeIs anyone on this channel?15:03
Free99hey everyone, I have a bridged connection that I need to filter. Should I use iptables+arptables, or will ebtables do the job of both?15:45
patdk-lapheh?15:46
patdk-lapthat depends on what you want15:46
patdk-lapall 3 of those do totally different things15:46
patdk-lapif they did the same thing, they wouldn't all exist15:46
Free99yeah, that's kind of why I'm a little confused. I basically need to isolate my VMs which are connected via a bridge.. the VMs are replicants of the VMs on a different physical host, so there would be an ip conflict15:47
Free99the hosts watch each other using pacemaker, if one machine goes down, the other unblocks its VMs15:47
patdk-lapare you doing some kind of memory replication?15:48
Free99no, just DRBD. I had been wondering if that would work actually15:48
patdk-lapya, your whole setup is fail15:49
=== uksysadmin__ is now known as itarchitectkev
patdk-lapkind about this, what happens when two computers use the same harddisk?15:49
patdk-lapit causes complete corruption15:49
patdk-lapyour vm should only be running at one location at a time ever, then no harddrive corruption, no ip issues, ...15:50
Free99patdk-lap: so what could I do instead? I suppose I would just keep the VMs off on the other machine, keep their disk images synced though15:50
patdk-lapdrbd keeps the disk images synced15:50
patdk-lapjust have pacemaker start the vm on the other machine on failover15:50
Free99lol I dunno why I thought that'd work15:50
Free99well, just for my knowledge, if I need to filter these VMs, like for fencing or anything, how would I do that?15:51
patdk-lapdepends :)15:51
patdk-lapyou could use iptabls/ebtables/...15:51
patdk-lapall depends on what level/layer you want to do it at15:51
Free99getting conflicting advice saying iptables doesn't work on br0 and such, others saying it does15:52
Free99kind of don't know my dick from my elbow when it comes to networking haha15:52
patdk-lapheh?15:52
patdk-lapiptables works on everything15:52
patdk-laplets put it this way15:53
uvirtbot`New bug: #1095710 in linux-lts-quantal (main) "update-grub-legacy-ec2 does not consider 3.5.0-generic as valid for Xen (dup-of: 1005551)" [Undecided,New] https://launchpad.net/bugs/109571016:46
uvirtbot`New bug: #1005551 in cloud-init (main) "Quantal does not boot on EC2" [Critical,Fix released] https://launchpad.net/bugs/100555116:56
=== ppetraki is now known as ppetraki-busy
=== jfluhmann_ is now known as jfluhmann
=== uksysadmin__ is now known as itarchitectkev
=== samba35_ is now known as samba35
=== mcclurmc is now known as mcclurmc_away
uvirtbot`New bug: #1095757 in krb5 (main) "krb5 packages should be updated to v. >=1.10.2 to workaround bug with gssapi kerberos authentication" [Undecided,New] https://launchpad.net/bugs/109575718:41
tflgen2hi guys, i asked this question on the #ubuntu channel but it may have fallen on deaf ears. I've just got a new installation of 12.04 and am trying to set up a mail server following this guide: http://flurdy.com/docs/postfix/edition11.html but I'm having issues with SASL. It shows up as Error: authentication failed: no mechanism available18:53
sarnoldtflgen2: anything in the logs?18:57
tflgen2mail log shows "mailsrv postfix/smtpd[8927]: warning: unknown[192.168.40.180]: SASL PLAIN authentication failed: no mechanism available"18:58
genii-aroundSounds like a PAM issue18:59
tflgen2in /etc/pam.d/smtp:                auth required pam_mysql.so user=mail passwd=QQuz43 host=127.0.0.1 db=maildb table=users usercolumn=id passwdcolumn=crypt crypt=119:01
sarnoldtflgen2: looks like pam_mysql.so is part of that guide; do you have it installed?19:01
sarnoldtflgen2: does the guy say why he hsa both "mailPASSWORD" and "aPASSWORD"? Why are they different..?19:02
tflgen2sarnold: no idea. i just use the same19:02
sarnoldtflgen2: and can you connect to your mysql database on the command line with the specified username and password?19:03
tflgen2yep19:03
tflgen2hang on19:06
tflgen2for a sec there, thought I had missed some files to install……not the case19:11
tflgen2sarnold: i'm kinda at a loss of where to look here. what else can i look at regarding pam?19:22
sarnoldtflgen2: are there any errors in the auth log?19:23
=== ppetraki-busy is now known as ppetraki
tflgen2no, it doesn't even look like it hits the auth log.19:24
sarnoldyou could always strace it. it's not much fun (especially if its your first time iwth strace), but you can at least watch what it tries to do..19:25
tflgen2https://bugs.launchpad.net/ubuntu/+source/cyrus-sasl2/+bug/87544019:35
uvirtbot`Launchpad bug 875440 in cyrus-sasl2 "Cannot authenticate with saslauthd and mysql" [High,Confirmed]19:35
sarnoldtflgen2: wow. be sure to click the "oes this bug affect you?" thing..19:41
tflgen2https://bugs.launchpad.net/ubuntu/+source/cyrus-sasl2/+bug/875440/comments/61 that is the fix19:42
uvirtbot`Launchpad bug 875440 in cyrus-sasl2 "Cannot authenticate with saslauthd and mysql" [High,Confirmed]19:42
=== TheLordOfTime is now known as LordOfTime
=== LordOfTime is now known as TheLordOfTime
uvirtbot`New bug: #1095775 in maas (main) "maas 0.1+bzr482 (precise) incompatible with python-django 1.4 (cloud-archive)" [Undecided,New] https://launchpad.net/bugs/109577519:46
sarnoldtflgen2: oh, woo, so the guide is outdated and needs fixing. that's more pleasant. :)19:46
genewitchso i am trying to use the cloud live image, the http://localhost comes up but does not accept the password in the GettingStarted.txt file19:52
genewitchWhat up wit that19:52
tflgen2case sensitive?19:53
genewitchyeah it just says ubuntu:ubuntu12319:53
genewitchi can't even think where to look for the config file with that password in it so i can change it19:55
tflgen2"Point your browser to the public address of the openstack-dashboard node, "http://node-address/horizon" , login using admin/openstack"19:57
genewitchhorizon isn't a valid URI and that password combo doesn't work. it seems like the glance API isn't working on port 929219:58
tflgen2hrm19:58
genewitchyeah the stock commands don't work.19:58
genewitchIt's ok, i mean i am sure i can set it up manually, i was a bit enthused by being able to check it out this way19:59
genewitchoh, eth3 (the nic i am using) keeps getting unassigned20:01
genewitchso this is a networking issue perhaps20:01
genewitchand there it goes20:01
genewitchit keeps getting wiped out20:03
RoyKgenewitch: pastebin ifconfig -a20:03
genewitchwhat's the pastebin package called on ubuntu20:03
RoyKpastebinit20:04
genewitchi killed network-manager20:04
genewitchshould i not have done that20:05
genewitchhttp://paste.ubuntu.com/1492939/20:05
RoyKif you reconfigured /etc/network/interfaces and restarted first, yes20:05
genewitchno20:05
RoyKeth[012] not configured, eth3 configured with rfc1918 address, what do you want to know?20:06
genewitchRoyK: It kept getting UNconfigured.20:06
genewitchlike 4 times.20:06
RoyKpastebin /etc/network/interfaces20:07
genewitchnothing in there but lo20:07
RoyKwell, then configure it20:08
RoyK!interfaces20:08
RoyK!network configuration20:08
RoyKno idea how to get that from the bot - google ubuntu network configi20:08
genewitchi know how to manually configure a network. My point was that's not in the GettingStarted guide.20:09
genewitchnice: failed to get http://169.254.169.254/latest/meta-data/public-keys20:11
genewitchthis is a bang-up job.20:11
genii-aroundLooks like no dhcp and it falls back to the 169.254.x.x20:27
genewitchgenii-around: no, 169.254.169.254 is a metadata server20:31
genewitchi guess this cirros image is meant to boot on amazon cloud, and openstack doesn't have a metadata server by default20:32
genewitchoh, it does have a metadata server, just doesn't have curl :-p20:33
tflgen2genewitch: i just booted into the livecd of ubuntu cloud on virtual box and was able to get an instance of cirros to boot20:35
genewitchyes, i have 3 running.20:38
genewitchi'm testing ephemeral20:38
tflgen2oh, my bad20:38
genewitchit is KVM >.<20:39
LovelidgeQuestion to all. Where can I get documentation on how to install mono on uBuntu server?20:55
LovelidgeHello?21:04
guntbertLovelidge: does http://askubuntu.com/questions/11921/how-to-install-mono-on-a-server help?21:07
tflgen2has anyone here migrated from courier to dovecot? I'm gonna try and give dovecot a try to see if the imap behavior is better.21:10
guntberttflgen2: I never "migrated" - and dovecot gave no reason for complaints21:13
tflgen2guntbert: i guess i'm not actually looking to migrate as this is a fresh install. Doing this for a customer, they currently are hosted elsewhere and using outlook with pop3. I'm hoping to get them on my server with courier/dovecot (preferably imap).21:15
tflgen2do you have a good tutorial on installation of dovecot on 12.04?21:15
_KaszpiR_Question - I have a home server running ubuntu-server LTS on some quite old motherboard, SATA (AHCI) in bios, booting form single drive21:23
genewitchthat's not a question21:24
_KaszpiR_now, I'd like to migrate to  new motherboard that uses EFI - any suggestions what can go wrong/21:24
genewitch_KaszpiR_: nothing, if you turn EFI off21:24
lastninjawitam można po polsku21:24
_KaszpiR_hm21:25
_KaszpiR_ok, looks like I've found some more info21:28
=== alaing_ is now known as alaing
guntberttflgen2: sure - see https://help.ubuntu.com/12.04/serverguide/dovecot-server.html :)21:33
=== cpg|away is now known as cpg
tflgen2gah, dovecot sucks :( can't even get it to start "dovecot: config: Warning: Killed with signal 15 "21:56
patdk-lapheh? dovecot is great22:01
patdk-lapI switched to it in 2003 I think, and I wouldn't call anything else currently close to it22:02
tflgen2i can't get the damn thing to start :)22:03
patdk-lapwell, it has a lot of config options22:03
patdk-lapuse dovecot -n, to see what you did22:03
tflgen2! thanks for that22:03
ubottutflgen2: I am only a bot, please don't think I'm intelligent :)22:03
patdk-lapubottu, oh, go get a life22:04
ubottupatdk-lap: I am only a bot, please don't think I'm intelligent :)22:04
tflgen2patdk-lap: ok, so that helped me weed out some crap. now i'm staring at fatal: unknown database driver 'mysql'22:07
patdk-lapthat is a debian/ubuntu thing22:08
patdk-laplikely didn't install the dovecot-mysql package22:08
tflgen2dovecot --build-options |tail -n422:08
tflgen2Mail storages: shared mdbox sdbox maildir mbox cydir raw22:08
tflgen2SQL driver plugins: mysql postgresql sqlite22:08
guntberttflgen2: how did you install dovecot? I had no troubles at all22:08
patdk-laphmm22:09
tflgen2turns out =i didn't have the dovecot-mysql package installed….22:13
tflgen2 /facepalm22:14
uvirtbot`New bug: #1095840 in nova (main) "No logrotate config for nova.log" [Undecided,New] https://launchpad.net/bugs/109584022:16
=== cpg is now known as cpg|away
anti-neutrinohi guys23:25
anti-neutrino I have a ubuntu-11.10 (Desktop) installation.. which I was using as a server (not using any gui feature of it)23:25
anti-neutrinocan I just rip off the ubuntu-desktop package .. to make it as efficient as ubuntu-server23:26
anti-neutrinoas I see on the ubuntu website .. they merged the generic and server kernl after 12.04 .. but I have 11.10 installation23:26
anti-neutrinothis machine was suppoed to be only for testing .. but then we want to use it in the production too23:27
anti-neutrinoplease suggest .. thanks!23:27
sarnoldanti-neutrino: feel free to uninstall any packages you don't use.23:28
sarnoldanti-neutrino: you'll make most impact by removing applications that are using the most amount of memory, but that can be a bit difficult to quantify23:29
anti-neutrinothanks for the reply @sarnold23:29
sarnoldanti-neutrino: for my little pandaboard, I just asked upstart to not start X, it seemed like a nice half-way point. it doesn't run, but the libraries are all there in case I want to ssh -X an application some day.23:29
anti-neutrinoyeah I am removing all the unwanted services .. (making list  of resource hungry services from the top command)23:30
anti-neutrinoyeah I can do that too ..23:30
anti-neutrinoso once again a naive question .. if kill X  .. it pops up again23:31
anti-neutrinoi understand this should be in something like 'startup'23:31
sarnoldaha, here we go: http://upstart.ubuntu.com/cookbook/#override-files23:31
sarnoldanti-neutrino: the book says these override files were added in upstart 1.3; if so, something like: echo manual >> /etc/init/lightdm.override    ought to save you a few hundred megabytes of RAM/swap23:32
anti-neutrinocool... thanks a lot sarnold23:34
anti-neutrinoI can try this on one of the VMs ..23:34
anti-neutrinoyou guys are awesome .. this seems to do the trick :)23:35
=== cpg|away is now known as cpg
sarnoldanti-neutrino: you may also wish to look at the powertop output and find which processes are responsible for the majority of your CPU wakeups23:39

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