/srv/irclogs.ubuntu.com/2013/03/04/#ubuntu-server.txt

=== NomadJim_ is now known as NomadJim
=== roasted_ is now known as roasted
=== VD is now known as Guest18424
=== histo1 is now known as histo
=== locust is now known as Guest48387
=== xcrracer__ is now known as xcrracer
=== tyhicks` is now known as tyhicks
=== vooze_ is now known as vooze
SyriaHello!! I have deleted all the users in the phpmyadmin by accedent!! How can create a new user agian?? I can't log in.09:44
andolSyria: http://dev.mysql.com/doc/refman/5.0/en/resetting-permissions.html#resetting-permissions-unix09:46
Syriaandol:  After submitting the query I got this "Rows matched: 0  Changed: 0  Warnings: 0"09:53
SyriaI don't think that I have a "root" user now, I have deleted all the users.09:53
Syriaandol:  Can I remove it compeletly and re-install it agian from the scratch please? I don't have important information like data bases or something.09:55
=== frojnd_ is now known as frojnd
andolSyria: You can do that too.09:58
SyriaLost the connection!10:06
dchevakhello all how are you10:29
dchevakI wish I new about IRC before changing anyway I was running ubuntu 12.04 server on 2 pc's for a small wireless intranet anyway switched one to CentOS 6.3 I always seem to have problems with ubuntu updates with apt-get10:32
dchevakcan someone give me a link to search for IRC chat rooms looking for BIND NAME server ISC maybe I should go there maybe they have it posted on there web site10:34
psivaajamespage: hallyn: Reported a new bug on lxc test cases, for a failure in SimpleLxcTests: bug 114411110:51
uvirtbotLaunchpad bug 1144111 in ubuntu-test-cases ""Error: Failed copying lxc package contents in container test-raring" in SimpleLxcTests test step" [Undecided,New] https://launchpad.net/bugs/114411110:51
zuljamespage:  quantum has testrepository support now maybe we should consider switching the packaging to use it? (i havent tested it though)13:56
jamespagezul, if you would like to test and propose it please feel free :-)13:57
zuljamespage:  ack13:57
hallynpsivaa: yeah i hadn't had a chance to verify, but some comments after stgraber's PEP8 cleanup patch made me think he might have accidentally undone a fix14:17
hallyni'll merge a fix, thanks14:17
zuljamespage:  yeah not going to happen14:18
hallyn(i'll address stgraber's other comments - which i hadn't yet - at the same time :)14:21
psivaahallyn: ack, thank you14:21
g0tchahey guys, everytime i try to install something using apt-get i get this msg http://paste.ubuntu.com/5585178/14:21
g0tchacan someone help me out with this? i tried to run 'apt-get -f install' and 'apt-get autoremove' and still the same problem14:22
stgraberhallyn: oops, sorry about the breakage... I thought the diff looked reasonable here after I handled the conflict, but apparently not...14:23
hallynstgraber: it's a long string of chrooted commands, eye-numbing.  completely understandable :)14:24
hallynpsivaa: fix should be pushed.  are you goign to be doing a new run very soon, or should i?14:35
psivaahallyn: i could run once the fix goes to  lp:ubuntu-test-cases/server. Until then our tests wont pick the fix, but if you want me to run it against your personal branch, i could try14:37
hallynpsivaa: i pushed the fix to there14:37
psivaahallyn: ahh ok, ill run the tests then. thanks14:38
hallynpsivaa: thanks.  (i had to nuke themachine i was last testing on :)14:39
psivaahallyn: np, the tests are running now, I'll let you know the results once they are done14:40
=== wedgwood_away is now known as wedgwood
lydaLooking at the ubuntu cloud server images here: http://cloud-images.ubuntu.com/precise/current/14:45
lydais there an easy way to go from the .img files to an AMI?14:45
lydain other words can i configure server images locally, run them through local testing and then release them as AMIs?14:46
=== lynxman_ is now known as lynxman
SpamapSlyda: IIRC there's a tool that does exactly that15:24
SpamapSlyda: curious though, why not customize an EBS root instance and then just snapshot it and turn it into an AMI?15:26
lydaSpamapS: that's a valid way to do things of course, but I'd like to create an instance that i can use on open stack as well.15:32
lydain addition i'd like to use it for running tests locally - without fears of incurring huge costs.15:32
eagles0513875_hey guys i am looking for documentation on how to setup sieve on ubuntu 12.04 does anyone have a good link on how to configure it15:40
lydawhy bzr, why?15:41
eagles0513875_?15:41
eagles0513875_that came out of nowhere lol15:41
lyda(sorry, nothing against bzr, but the rest of the world has chosen git.)15:41
eagles0513875_lyda: canonical seem to enjoy being different. hwo did you find setting up openstack on ur system? when i read the open stack documentation i find it rather daunting and intimidating15:42
lydai find it a bit of a pain.  we don't really have a decent setup yet.15:43
eagles0513875_lyda: have you taken a look at cloud stack15:43
eagles0513875_citrix open sourced it under the apache license nto sure if that would be any easier15:43
lydanot really, no.15:43
lydais it an openstack setup tool?15:43
lydathey often seem worse than openstack!15:44
eagles0513875_no15:44
eagles0513875_concept i would say is similar to that of open stack but its focus is infrastructure as a service15:44
eagles0513875_lyda: i guess it depends on the project15:45
lydaah.  well, an amazon compatable api is important.15:45
lydaand honestly i think open stack is the right direction.15:46
lydaspecifically i'm interested in the lxc "virtualisation."15:46
lydaha!15:46
eagles0513875_:) lyda thanks for confirming with me though that im not the only one who finds open stack setup and configuration daunting15:46
lydasudo pip install bzr15:46
eagles0513875_bzr is python based O_o15:47
lydagit clone bzr::lp:cloud-init15:47
eagles0513875_O_o15:47
eagles0513875_lyda: are you setting up your own instance of launchpad15:48
lydai'm just checking out the code.  but with git, not bzr.15:48
eagles0513875_interesting15:48
eagles0513875_i have heard that launchpad isnt an easy thing to work with or setup15:49
eagles0513875_lyda: are you looking for a bug tracker of some sort?15:49
lydanope.15:49
lydainternally we use jira.15:50
lydaand for SCM we use gitlab.15:50
eagles0513875_nice :) im hosting a few personal repos on a virtual server of my own but i think im going to start using bitbucket more15:50
eagles0513875_with bitbucket you have a choice between mercurial and git15:50
eagles0513875_lyda: if you looking for an external bug tracker checkout redmine as I have heard alot of good things about it15:51
lydai have two oss projects and i actually use all the major git code hosters so contributors can use what they feel comfy with.15:52
psivaahallyn: the LXC tests have another failure, http://pastebin.ubuntu.com/5585410/. Please let me know I should report a separate bug15:53
lydasee the README for https://github.com/lyda/misspell-check15:53
eagles0513875_lyda: have you used any web frameworks or anything seeing as you like using python15:54
lydai've been trying out bottle (framework) and waitress (for serving).15:55
lydathey seem to work well.15:55
eagles0513875_does anyone know if i install the dovecot-sieve package on 12.04 if that will effect the current operation of my mail server? or will i still continue to to get emails and send them15:55
eagles0513875_lyda: if you like python check out django its is probably the most complete in terms of documentation15:55
lydait is, but it also seems kind of heavy.15:56
lydai have played with it.15:56
eagles0513875_lyda: what webserver did you try it out on15:56
eagles0513875_i need to when i have the time migrate all myservers away from apache prefork15:57
eagles0513875_that thing is a memory hog and implement php-fpm15:57
lydafor waitress i just used it directly for an internal deployment.15:58
lydai think if i was actually deploying it i'd put it behind ngnix or apache.15:58
eagles0513875_what exactly is waitress like a pos system for restaurants or something15:58
eagles0513875_what about lighttpd lyda i have been told that has virtually 0 memory footprint somehow15:58
=== matsubara is now known as matsubara-lunch
sliddjurCan someone help me out with enabling pam module cracklib to set password complexity rules? I've come to the point where I dont know what im doing anymore... :s16:11
sliddjur./etc/pam.d/common-password:password     required        pam_cracklib.so retry=3 minlen=8 difok=3 dcredit=1 ucredit=1 lcredit=1 ocredit=116:12
hallynpsivaa: d'oh.  typo.  sorry16:12
hallynoen sec,16:12
sliddjurI thought this would enable password rules, but it doesnt seem to be enabled when I change my users password (not with root)16:12
hallynpsivaa: fix pushed, thanks16:13
psivaahallyn: ok thanks, running again16:14
SpamapSlyda: perhaps have a look at https://github.com/stackforge/diskimage-builder16:19
SpamapSlyda: it will output a raw image that, I think, will boot just fine on ec216:20
rbasakjamespage: hey! I've prepared an upload for ipmitool in bug 1074443 that I'd like to get in before feature freeze. Should I put this in the sponsor queue as usual in the hope that it'll go in on time, or could you review/upload it for me please?16:23
uvirtbotLaunchpad bug 1074443 in dell-poweredge "Please upgrade ipmitool to 1.8.12 for 13.04" [Wishlist,In progress] https://launchpad.net/bugs/107444316:23
SpamapSeagles0513875_: also re "why bzr" its not so much that canonical enjoy being different as the fact that bzr and git grew at the same time... git in the kernel community (and later everywhere else) and bzr in canonical and its projects....16:24
SpamapSlyda: ^^16:24
eagles0513875_SpamapS: what advantages does it have over the already established ones such as git svn mercurial16:25
SpamapSlyda: anyway, since you are interested in openstack .. di-b is nice.. and developed using the same model.16:25
SpamapSeagles0513875_: bzr is already established too ;)16:25
SpamapSeagles0513875_: its just not widely adopted16:25
eagles0513875_what advantages does it have over the main stream ones  SpamapS16:25
SpamapSeagles0513875_: bzr doesn't ever rewrite history unless you explicitly say "delete that commit" and it can only pop them off the top.16:25
SpamapSeagles0513875_: instead it folds them down into commits underneath the larger ones16:26
eagles0513875_interesting but technically isnt that what the others do as well?16:26
SpamapSeagles0513875_: it also is *FAR* easier to learn how to use the whole bzr toolset. If you've never used bzr or git, bzr will take maybe a week to master.. git .. well I've been using git since December and I know f*** all about it.16:26
SpamapSeagles0513875_: forget svn, its not even in this conversation.16:27
eagles0513875_interesting16:27
eagles0513875_lol16:27
SpamapSeagles0513875_: hg I know very little about16:27
eagles0513875_lyda: have you looked at juju and charms for what you are doing to help ease your deployments of frameworks etc16:27
SpamapSeagles0513875_: git does rewrite history.. you can delete commits going back forever w/ rebase.16:27
SpamapSheh... IMO juju is not quite there for "easing deployment of frameworks"16:28
SpamapSIt eases deployment of infrastructure like a champ16:28
SpamapSdatabases, monitors, logging, queues, all good. frameworks.. well.. just go try the django charm. :-/16:28
rbasakTechnically git doesn't rewrite history, and you can't delete commits. You only create new history and generally choose to forget about old branches by moving branch tips. What I find interesting is that when we say we "merge" an Ubuntu package with a newer Debian package, what we actually do is in git parlance a "rebase" (despite what we might commit in bzr UDD).16:31
SpamapSrbasak: yeah, the way bzr is abused for separate tracking of upstream source and debian changes is definitely more git's territory16:34
SpamapSrbasak: you are tainted by git mastery .. git masters *HATE* bzr16:35
SpamapSand, I get it now16:35
SpamapSbecause git has all the tools to do the appropriate surgery on your VCS to make it be representative of what you want.16:35
SpamapSbzr hides them or outright makes them impossibly hard to find, because bzr is a perfectionist16:35
rbasakI think the problem is that bzr is very poorly documented for git masters. It hides what's really going on. So yeah, exactly what you're saying.16:35
SpamapSI also have the bad fortune to be dealing with OpenStack's limited git workflow, which isn't helping me love git ;)16:36
rbasakAll the documentation is workflow based, which works until I have an unusual workflow. Then there is no documentation, and all I can do is ask others what they do. With git, the documentation is extremely details always enough (if you're willing to delve into the data structures, which is easier with a CS background).16:37
jamespagerbasak, ok - I'll take a look16:41
jamespagerbasak, is it actually a merge from upstream or a new upstream release?16:55
rbasakjamespage: I meant upstream as in not Debian. New upstream release that I've merged with previous Ubuntu package.16:55
rbasakAs you're asking, I guess the answer is "new upstream release". I guess I shouldn't call it a merge then in this case?16:56
jamespagerbasak, yeah - I'll fix it up locally and upload for you16:57
rbasakThank you!16:57
jamespagerbasak, uploading now16:57
jamespagerbasak, thanks for picking that up16:58
jamespagerbasak, build failure on powerpc - https://launchpad.net/ubuntu/raring/+source/ipmitool/1.8.12-0ubuntu117:24
rbasakjamespage: :-(17:24
rbasakjamespage: looking at it now. Looks like a simple endianness/porting issue.17:25
tkeithI upgraded a package and it asked me what I want to do about changed configuration and I hit N for "keep current config". Now I want to know what the changes were. How do I get apt or dpkg to give me that option again?17:36
rbasaktkeith: sounds like a conffile prompt. Usually you'll end up with .dpkg-dist or .dpkg-old files in the same directory as the conffile affected. I'm not sure how to identify which conffile was affected after the prompt though. Searching for files with a .dpkg-dist suffix in /etc might be your best bet.17:54
tkeithrbasak: Ah, thanks! I knew which file it was, and there's a .dpkg-disk17:54
rbasakGreat!17:55
smoserhallyn, i dont know if you're aware and it is by design or not18:32
smoserbut i think that default memory size to kvm changed18:32
smoseri think at one point in ubuntu it was 256 maybe.18:32
smosernow it seems like 12818:32
smoserbut i could be just mis-remembering18:32
hallynsmoser: yeah, kirkland carried a patch setting it actually to 356 i think,18:32
hallynno you're right18:32
hallynsmoser: we can re-add that patch if it's needed for something, i'm just trying to get delta from debian as small as possible18:33
hallyni do agree default size is useless :)18:33
hallynshout if you want the patch back18:33
smoserhallyn, perhaps try to push that to debian?18:34
smoserto keep the delta small18:34
mattraeanyone familar with apt-mirror? my first download completed but i wanted to run again to just get anything that updated. now its saying its downloading the same amount of data i initially downloaded. anybody know if this is actually an incremental download?18:34
hallynsmoser: lemme ask there real quick18:34
mattraei don't want to re-download everything if i can avoid it18:34
tasslehoffon my synology-nas it was possible to have an encrypted volume that I could decrypt and share via samba. how can I do the same on ubuntu? must I manually do the decrypt, and then restart samba with an added share on the mount point?18:58
tasslehoffor are there more elegant solutions?18:58
sarnoldtasslehoff: how did you give the key to the synology to decrypt the data?19:00
tasslehoffsarnold: I think I only had to give a password when creating the share19:01
tasslehoffso, not sure exactly how secure it was19:01
sarnoldtasslehoff: aha, so it stored the key somewhere?19:01
RoyKor used the password hash as the key?19:02
sarnoldtasslehoff: you could re-create that with ecryptfs or dm-crypt and store the key on the server somewhere, so it doesn't require your interaction to mount it every boot...19:02
sarnoldRoyK: ah, another good possibility19:02
RoyKmeaning !secure19:03
tasslehoffI would have to make sure samba starts after the encrypted folder is mounted?19:05
tasslehoffIs truecrypt a good solution as well? Have used it on my desktop before, but never on a server.19:09
* tasslehoff will google19:09
=== shadeslayer is now known as evilshadeslayer
=== matsubara-lunch is now known as matsubara
guntbertWhere resides the setting if I want be prompted for non-LTS upgrades or not?19:25
guntberts/if/whether/19:26
patdk-wk/etc/update-manager/19:27
guntbertpatdk-wk: thx so much :-)19:27
guntberthmm - two systems (12.04.02 LTS), both have "prompt=lts", one prompts me for 12.10 on login, one does not19:31
zuljamespage:  ping i filed the MIR for python-json-patch and python-json-pointer fyi20:02
hallynstgraber: triggered a few pre-existing bugs in my cgroup patch, so the patchset i'm sending out is growing...  on the bright side it'll survive all the tests i can throw at it20:29
hallynstgraber: interestingly, lxc-destroy isn't yet lxcpath aware (be careful with that!).  (patch to fix that going out now)20:30
stgraberhallyn: oh, so apparently I just used rm and not lxc-destroy when doing lxcpath testing ;) good catch20:34
hallynsmoser: mjt says he intended to send a patch to up default memory size upstream.  so we'll get this fixed through debian at least - thanks21:06
utlemmingsmoser: Bug #114521521:06
uvirtbotLaunchpad bug 1145215 in cloud-init "cloud-init mangles sources.lists if mirror can't be found" [Undecided,New] https://launchpad.net/bugs/114521521:06
smoserutlemming, what release is this ?21:10
utlemmingprecise21:10
utlemmingI booted this on a non-EC2/Openstack cloud21:11
axisyshow to install netbackup client?21:18
axisysdont see one21:18
axisysshould this question be in offtopic ?21:18
smoserutlemming, it seems there is a bug there, but i dont understand how it would ocur with the /etc/cloud/cloud.cfg that is packaged.21:22
utlemmingsmoser: I am overriding the cloud-config. See: http://paste.ubuntu.com/5586309/21:24
RoyKaxisys: heh - call support ;)21:24
smoserbuti i think your file inside the image is busted.21:26
smoseras it should look like: http://paste.ubuntu.com/5586315/21:26
utlemmingsmoser: is the apt_mirrors syntax deprecated?21:28
axisysRoyK: heh21:29
smoserutlemming, yes. https://bugs.launchpad.net/cloud-init/+bug/1006963 has the correct way to do that.21:32
uvirtbotLaunchpad bug 1006963 in cloud-init "sources.list configuration does not cover security" [Medium,Fix released]21:32
smoserit was intended for what you did to work, but clearly it did not.21:33
utlemmingsmoser: ack, thanks for the clarification21:33
smoserutlemming, fwiw, i cannot reproduce this.21:45
smoserit works as expected on openstack.21:47
hallynstgraber: hm, (testing now, but) does /etc/init/lxc.override work in lucid?21:47
stgraberhallyn: not sure, I can't remember using those back then21:47
utlemmingsmoser: interesting. I'm doing another test21:47
utlemmingsmoser: I'll give you access once its booted21:48
smoserhttp://paste.ubuntu.com/5586373/21:48
hallynstgraber: oh, sorry, that wasn't the problem...21:50
jcastrosmoser: do you have any generic cloud talks? We have that cloud talk in Ann Arbor on Wednesday btw.22:21
jcastrosmoser: I was just going to do Juju but I think some info about Guest would be welcome.22:22
aleza84how can I connect to a wireless lan wpa2 encrypted via console?22:51
tsaavikAnyone know if/where there are docs on the various default user groups in ubuntu?22:57
tsaavikYes, there is checkout:23:08
tsaavikhttps://wiki.ubuntu.com/Security/Privileges23:08
sarnoldtsaavik: neat, thanks23:09
azbyinhi all..23:22
azbyinwhat is the essential difference between  %admin ALL=(ALL) ALL     and   %sudo   ALL=(ALL:ALL) ALL  in the sudoers file?23:23
sarnoldazbyin: depends on the contents of the 'admin' group vs the 'sudo' group23:24
sarnoldoh, there's the (ALL) vs (ALL:ALL) difference, too...23:24
azbyini have identical sudoers files on my local workstation and a uec image running on a vm. both os are precise lts. on my local workstation my user is 'azbyin' and am in the 'sudo' group. on the cloud machine the user is 'ubuntu' and is in the 'admin' group23:24
azbyinnow, my workstation asks for my passwd when i try any command with sudo (with the obvious credential caching (for 15 mins?)) while on the cloud machine ubuntu is able to run any sudo command without request for passwd23:25
azbyinfrom my understanding of reading the sudoers manpage, it should ask for a password _unless_ NOPASSWD is present in the rule23:26
azbyinso, i don't understand the difference in behaviour23:26
bmoylesis there a defaults line in sudoers that specifies anything around authentication?23:26
azbyinoh, also.. on my workstation my user is uid 1000, on the cloud machine ubuntu user is uid 99923:26
bmoylesyou can turn off authentication globally in sudoers23:26
azbyinbmoyles, as i mentioned above, the sudoers files on my machine and the cloud machine are *identical*23:27
azbyindoes it differentiate between 'system' users with uid < 1000 and non-system users where uid >= 1000 ?23:28
azbyinif yes, i do not see this in the manpage23:28
azbyinso, i assume it must be because of the differing group membership and consequently differing rule being applied23:29
azbyinbut the rule by itself does not indicate that passwd will not be authenticated! so i do not see how this behaviour occurs23:29
bmoylesyes, they're identical, but look in /etc/sudoers.d23:30
azbyinah indeed23:32
azbyini was accidentally looking at the wrong tab23:32
azbyini.e. i looked at /etc/sudoers.d/ on my own workstation *twice* and saw only the README file23:32
azbyingrr..23:32
azbyinthanks for the clarification. and i'm happy that i could understand the manpage by myself :)23:33
sarnoldbmoyles: nice :)23:33
bmoylesthe sudoers manpage has to rank as one of the more confusing manpages to ever exist :)23:35
azbyinwell, it does list the BNF grammar23:36
azbyinbut compensates by giving very clear examples at the end23:36
=== evilshadeslayer is now known as shadeslayer
sarnoldthe downside is that the examples it gives are for a very comprehensive deployment suitable at a business or something23:53
sarnoldit needs a "dumb guy" section for simple uses for people like me :)23:54

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