[00:00] oh [00:00] this client sucks [00:18] <[Solars]> heh i hate bx [00:58] I reading an article in Linux Magazine all about webmin and how it works with ubuntu, if anyone was interested. [00:58] !webmin [00:58] webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. See !ebox instead. [00:58] !otherwaystoinstallcrapthatdontinvolveapt [00:58] Error: I am only a bot, please don't think I'm intelligent :) [01:03] <[Solars]> stainer lies! [01:06] sometimes you can get stuff that aren't even .deb packages. Some people will even install stuff from source, and God forbid, compile their own kernel. [01:06] hippies [01:08] <-- free thinker [01:08] <[Solars]> i remember when you were forced to compile my own kernels [01:08] on a 386 with 640k of ram [01:09] and it took several hours [01:10] or at least it seemed [01:10] :) [01:13] <[Solars]> erm mine always took over an hour [01:13] <[Solars]> i still have a set of old slackware floppies [01:14] hahaha [01:20] kirkland: just wanted to say you do good work... don't let the haters get you down :) [02:38] If a folder has the owner "root" and the group "escrow". Both have Read and Write access. If user "steven" is in group escrow, should he be able to edit that folder and the stuff inside or would I also need to have "root" added to the group "escrow"? === PrivateVoid_ is now known as PrivateVoid === mathiaz_ is now known as mathiaz [03:37] Redixe_home: Your question still current? [03:37] Redixe_home: Yes, if the user is in the group and the group has permissions, then the user will have those permissions [03:37] The user root can do anything anyway, so root doesn't need to be in the escrow group [03:38] I wish I could be root in real life. Then i wouldn't have to pay bills and stuff. rm -rf /bills === Jell1 is now known as crus1 === crus1 is now known as javaperl [05:25] hello i want to know how i can add another internal hardrive to my ubuntu server setup [05:33] Leading question Thorsten11, you mean how to physically add it or how to get it detected in the OS? === freaky[t] is now known as fReAkY[t] [05:51] how to get it detected in the OS [05:53] i am wanted to add more hd space on my server without swapping hd's. I just wanted to know if add a second hd was easy enough === fReAkY[t] is now known as freaky[t] [05:58] wow my english was horrible there and its my primary tongue lol [05:58] =p [06:03] Thorsten11: shove it in there and it'll be detected [06:03] (within reason) [06:04] lol great thanks [06:04] thats all i needed to hear [06:05] Heh, Ubuntu recognises my 320GB drive but my BIOS says it's 137GB... given, it's a P3 [06:19] kirkland: I've tested the latest -server iso. virtio network seems to work [06:19] mathiaz: thanks, i did too [06:19] mathiaz: i was going to confirm what i saw with you tomorrow [06:20] kirkland: there has been another commit to the kernel tree [06:20] mathiaz: there's still an issue with the disk [06:20] mathiaz: we're closer [06:20] mathiaz: there's something wrong with the bootloader installation [06:20] kirkland: that has not been uploaded yet [06:20] kirkland: so we should plan to test again this configuration once the kernel is updated or d-i [06:20] mathiaz: k [06:20] kirkland: right - I haven't looked into the virtio block issue [06:21] kirkland: it's not a regression though. [06:21] mathiaz: i'm going to spend more time on it tomorrow [06:21] kirkland: I don't think installing on a virtio block device was supported/working in hardy [06:21] mathiaz: the installer was in bad shape until about 4pm today, though [06:21] mathiaz: no, i saw the same problem there [06:21] mathiaz: i think we're pretty close [06:22] kirkland: hm - ok. I'll ask about the commit to the kernel tree in the release meeting tomorrow [07:21] `find -name .svn -prune -o -type f` includes directories called .svn but not files within them [07:21] Is there any way to have the .svn directories not listed? [07:23] Aha, got it, they appear in the output, but they're not parsed by -exec :) [07:40] * delcoyote hi [07:46] how to find the performance checking for dns dhcp [07:54] please anyone tell me to do performance checking for DNS [07:54] or proxy server [08:06] sommer: happy b-day! :D [08:08] * Koon hugs sommer too (hi, ivoks !) [08:31] Koon: hi ;) === Doc|Serverising is now known as Indoctrine [09:02] how to move mysql data dir to another dir? [09:02] shutdown mysql [09:02] when i move it to /home/mysql [09:02] mv /var/lib/mysql /home/mysql [09:03] mysqld cannot start [09:03] you have to configure it [09:03] i already configured [09:03] my.cnf [09:03] ln -s /home/mysql /var/lib/mysql [09:03] don't touch that file [09:03] start mysql [09:03] there's conf.d directory in ubuntu [09:04] create /etc/mysql/conf.d/my_custom_conf.cf [09:04] and add: [09:04] [mysqld] [09:04] datadir = /home/mysql [09:04] in my.cnf or conf.d? [09:05] 10:04 < ivoks> create /etc/mysql/conf.d/my_custom_conf.cf [09:05] 10:04 < ivoks> and add: [09:05] i am trying [09:05] if you create a symlink to the database dir you don't need to change the config [09:06] much easier imho [09:06] .cf?or.cnf? [09:06] cnf, right [09:07] any file name? [09:07] correct [09:09] just this step? [09:10] just those two lines, yes [09:11] it would be wise to change log-bin too [09:11] since, bin-logs can be big if you have lots of SQL queries [09:11] should i change my.cnf? [09:11] i think i already said not to touch that file [09:12] i just add this file,but still not working [09:12] then something else is wrong, not datadir [09:13] 081010 16:12:41 InnoDB: Operating system error number 13 in a file operation. [09:13] InnoDB: The error means mysqld does not have the access rights to [09:13] to...? [09:14] InnoDB: the directory. [09:14] InnoDB: File name ./ibdata1 [09:14] InnoDB: File operation call: 'open'. [09:14] InnoDB: Cannot continue operation. [09:14] i get this message [09:15] well, does mysql user has access to your datadir? [09:15] i give the rights to mysql user [09:15] what's in /home/mysql directory? [09:15] use chown [09:15] the fils which copy from /var/lib/mysql [09:16] and what does 'ls -dl /home/mysql' looks like? [09:17] drwxrwxrwx 3 mysql mysql 4096 2008-10-10 15:38 mysql [09:17] i give all rights to this directory [09:19] and files in it? [09:20] are they owned by root or mysql? [09:20] mysql [09:20] not root [09:24] ps ax | grep mysql [09:24] maybe you have mysql running already? [09:24] not running [09:26] i use 64bits server edition [09:26] lsof /var/lib/mysql/ibdata1 [09:26] lsof /home/mysql/ibdata1 [09:27] both shouldn't retrun anything [09:27] errr [09:27] Operating system error number 13 [09:27] that's permission denied [09:27] how do you start mysql? [09:27] both nothing returned [09:28] sudo /etc/init.d/mysql start [09:28] and 'ls -dl /home/mysql/ibdata1'? [09:28] -rwxrwxrwx 1 mysql mysql 10485760 2008-10-10 15:38 /home/mysql/ibdata1 [09:29] why is everything 777? [09:30] i did this. because 666 not work,so i try 777 [09:30] all files in /home/mysql should be 660 [09:30] find /home/mysql -type f -exec chmod 660 {}\; [09:30] find /home/mysql -type d -exec chmod 700 {}\; [09:30] chmod 755 /home/mysql/mysql [09:31] chmod 600 /home/mysql/mysql_upgrade_info [09:31] chmod 644 /home/mysql/debian-5.0.flag [09:31] chmod 755 /home/mysql [09:32] chown -R mysql:mysql /home/mysql [09:33] ok,i did all above.but still not working [09:33] same error message [09:34] what's in your /etc/mysql/conf.d/custom_file.cnf? [09:34] or however you called it [09:34] [mysqld] [09:34] datadir=/home/mysql [09:34] just this two lines [09:34] do you still have a backup of original /etc/mysql/my.cnf? [09:35] i have restored this file [09:36] and still doesn't start? [09:36] when i delete custom_file.cnf [09:36] mysqld can started [09:36] yes [09:36] but then it uses /var/lib/mysql [09:37] yes [09:37] when it uses /var/lib/mysql,it's ok [09:38] stop mysql [09:39] rm -rf /home/mysql [09:39] stoped [09:39] cp -a /var/lib/mysql /home/mysql [09:40] finish [09:40] then? [09:41] still not working [09:46] just a sec... phone [09:48] create a file /etc/mysql/conf.d/custom.cnf [09:48] add in it: [09:48] [mysqld] [09:48] datadir = /home/mysql [09:48] save it and start mysql [09:50] no use.. [09:50] i try this in cent os, it works,but in ubuntu it fails [09:52] hm [09:53] is ubutun security policy too stricted? [09:53] /home is on separate partition? [09:53] aaaaaaaaaaaaaaaa [09:53] yes [09:53] correct [09:53] apparmor :D [09:53] separate partition [09:54] apparmor profile doesn't allow mysql user to read outside /var/lib/mysql [09:54] run this: [09:54] aa-complain /etc/apparmor.d/usr.sbin.mysqld [09:54] how to fix it? [09:54] then? [09:54] and then start it [09:55] ok.it works [09:55] but this is just a workaround [09:55] cause, your logs will get filled with apparmor complaints [09:56] ....... [09:56] how to fix it? [09:56] open file /etc/apparmor.d/usr.sbin.mysqld [09:56] at the end, before last }, add this: [09:56] /home/mysql/ r, [09:57] /home/mysql/** rwk, [09:58] just this two lines? [09:58] yes [09:58] then stop mysql [09:58] run: [09:58] aa-enforce /etc/apparmor.d/usr.sbin.mysqld [09:58] and then start mysql [10:00] during distribution upgrades, keep an eye on that file [10:00] ok [10:00] thanks [10:00] you'll have to add these two lines [10:00] if you would like to try... [10:00] do you have 2-3 minutes to test something? [10:00] yes [10:01] could you delete those lines you added and create new file? [10:01] /etc/apparmor.d/custom.mysqld [10:01] with these lines: [10:01] ok,i try [10:01] /usr/sbin/mysqld { [10:02] #include [10:02] /home/mysql r, [10:02] /home/mysql/** rwk, [10:02] } [10:03] hm... i'm not sure that will work, but let's try [10:03] finished [10:03] then? [10:03] aa-enforce /etc/apparmor.d/* [10:04] any errors? [10:04] no [10:04] try restarting mysql [10:05] does it work? [10:05] failed... [10:05] bah... [10:05] try removing that line #include from your new file [10:06] then run /etc/init.d/apparmor restart [10:06] and then start mysql [10:06] need aa-enforce [10:06] ? [10:06] no [10:06] it's already enforced [10:07] still failed [10:07] bah... [10:08] ok, than just delete that file and add those two files to /etc/apparmor.d/usr.sbin.mysqld [10:08] hah,give up~ [10:09] right... we'll have to work on how to configure 'custom' apparmor profiles as an addition to already existing profiles [10:09] however strange that sounds :D [10:10] ha [10:10] for those that still don't know: [10:10] Wikipedia moves to Ubuntu Server :D [10:13] thanks a lot. [10:13] np [10:14] aa just use in ubuntu? [10:16] and suse, iirc [10:16] apparmor is a security shield [10:17] this armor is too strong~ [10:17] :) [10:20] i am going to offline,Bye~ [10:22] New bug: #281188 in samba (main) "Panic or segfault in Samba" [Undecided,New] https://launchpad.net/bugs/281188 [10:30] please anyone tell me to do performance checking for DNS [11:07] gane: I have used a tool called dnsperf a while back, don't think it is packaged though, so you might have to take the source and compile them. === freaky[t] is now known as freaky_t === freaky_t is now known as freaky[t] === freaky[t] is now known as freaky_t [11:32] what is the version of used for server [11:32] what is the version of kernel used for server [11:32] 2.6.24-19-server [11:51] shall i get this config file [11:51] Deeps, what sre the features support by this kernel [12:04] Deeps, i want to enable smp in the kernel source how to [12:07] <_ruben> smp is enabled by default [13:31] ivoks, Koon thanks :) === freaky_t is now known as fReAkY[t] === fReAkY[t] is now known as freaky_t [13:54] Koon: have you tested the pam_smblogin fix? [13:55] zul: yes, quite extensively [13:55] cool so I can upload the debdiff for you as soon as I finished this SRU [13:56] I was worried that the samba-common upgrade would find /var/lib/samba "owned" by samba but that's not the case [13:56] ok sounds good thanks for the fix === freaky_t is now known as fReAky_t [14:36] is there any way to let sshd chmod files after they got transferred via scp (on the incoming side)? or let sshd automatically trigger a hookscript after file transfers ? === fReAky_t is now known as freaky_t === benkant_ is now known as benkant [14:56] Koon: uploaded it by the way [14:56] yay ! [15:03] lamont: hi! ping re feedback on bug #279316 [15:03] Launchpad bug 279316 in bind9 "SRU for bind9 to 9.4.2.dfsg.P2 on hardy" [Undecided,New] https://launchpad.net/bugs/279316 [15:04] this channel is now support :O [15:05] meh [15:06] jdstrand: looking now - anything particular you want on there that isn't asked for in the text already? [15:06] FAT32 is so useless [15:07] steven in group escrow can't edit a directory that is owned by root in group escrow with read and write... what would cause this? [15:08] I'm guessing something with sshfs is causing the problem :( [15:08] lamont: just a comment on regression potential [15:10] lamont: also, can you briefly comment on: [15:10] 16:28 < jdstrand> lamont: also, I noticed that we went from libisc32 to libisc35, but it looks like only bind9 binaries use it-- can you mention that as well? [15:12] RediXe: the directory *containing* the directory which should be edited is group escrow & has g+w ? [15:14] phaidros: On the server, the directory is /home/common /home is rwxr-xr-x root root ... while common is rwxrws--- root escrow ..... would the /home be ruining it? [15:14] jdstrand: done and done. [15:15] RediXe: yes. the user shoud have correct permission for the file/dir *and* the dir containing that one. [15:17] phaidros: So should I move the common folder to /common (because I also have /home/finance that is not part of group escrow but group finance)? [15:17] lamont: \o/ thanks man! [15:19] jdstrand: sorry for taking so long. :-( [15:19] RediXe: yes, this could be a solution. [15:20] lamont: I had my part in that as well... [15:20] phaidros: Is it a smart solution or is there a more "preferred" way? Just curious, trying to make sure I don't start doing something the "wrong" way that I might regret later on. [15:21] jdstrand: win! [15:23] RediXe: you only have two choices for that case: either move the content, which users should be able to edit somewhere else, *or* give the users the permissions where the files actually are. smart is, to never give users to many permissions, so they cannot mess, even accidentially. [15:23] RediXe: how you approach this is just a matter of taste. [15:26] phaidros: So lets say I put /companyname/common and /companyname/finance ....companyname is rwxrwx--- root companyname ---- /companyname/common is rwxrw---- root escrow ----/companyname/finance is rwxrw---- root finance ..... As long as the user was in group companyname and escrow that user could access /companyname/common but not /companyname/finance right? [15:26] RediXe: correct. [15:26] phaidros: I think I'm starting to get this stuff down finally :P [15:27] RediXe: until you understand it fully, it seems illogical. once u got it, you even stumble here and there over nice things, like sticky bits and such .. happens to everybudy :) [15:28] phaidros: Well I thank you a ton for your help! [15:28] np [15:35] phaidros: sudo usermod -g rps steven would add steven to the group rps right? .. ran id steven and I see I'm part of that group but I don't see a group I was in like it replaced it somehow [15:39] greetings,postfix : i was told for ubuntu there were some changes in the location of the smtpd.conf , what would be the right location today ? /etc/postfix/sasl/smtpd ? [15:40] ATA_Dark_Shadow: Yes as for many, many years before today. [15:44] kirkland: so I think that installation with virtio network interface is working for intrepid, correct? [15:44] mathiaz: i need to test today's iso, but i think so [15:44] where should i start to look for an error why postfix possibly does not "contact" saslauth? (i have saslauthd with -d started ; testsaslauth works fine and shows infos) [15:44] nijaba: wrt to JeOS. There are two things that need to be done - -virtual installation + install the minimal seed instead of the standard? [15:45] Is there a limit on how many groups a user can belong too? [15:45] mathiaz: yes [15:45] mathiaz: did you solve the -virtual kernel issue? [15:45] nijaba: yes - it's all good [15:45] \o/ [15:46] nijaba: I can just seed them on the -server iso [15:46] nijaba: we just need to fix the installer options and figure out the seed installation. [15:46] mathiaz: that would be GREAT [15:46] nijaba: could this be considered as a regression from hardy? [15:46] ATA_Dark_Shadow: It's probably because postfix by default is in a chroot. If you look in the Ubuntu Server Guide it'll tell you how to get it to talk to saslauthd. [15:47] mathiaz: certainly, as we decided to pull the JeOS iso, considering that this feature would be present [15:47] nijaba: ok. [15:48] ScottK ill check that guide again, maybe i missed something the first time, my postfix isnt in the chroot anymore though, ill have another look [15:48] kirkland: how does the openiscsi look? [15:48] mathiaz: delayed to next week; Etienne is going to recreate his test environment and we're goign to work on it then [15:49] ATA_Dark_Shadow: Then if it's not in the chroot, then guide will be wrong. [15:50] kirkland: ok. === ivoks2 is now known as ivoks [16:14] sommer: bday happy a you wish i :) [16:16] * nxvl waves on sommer [16:16] ScottK still no use, i changed back to chroot enviroment, i still can "login" with testsaslauth and recieve mails without "saslauth" but when i try to send a mail through a mail client i get authentification failed, and saslauthd seems not to get called [16:16] soren: happy birthday (if you aren't drunk already/still) [16:19] errr [16:19] sommer: ^^^ [16:19] soren: sorry [16:19] :P [16:20] ivoks: heh, thanks [16:20] nxvl: nope not yet, still at work :) [16:21] *yet* is the key [16:21] :D [16:22] on my birthdays i always end when a friend driving mi gf home and another driving me [16:22] :D [16:23] lol, as long as you get there it's all good [16:23] ATA_Dark_Shadow: I don't have time for detailed assistance right now. My advice is very carefully go over the recommendations in the server guide and see where yours is different. Sorry. [16:24] can anyone suggest a good tool for keeping track of a couple dozen ubuntu servers? I'd like to use ssh-agent and pubkey auth to handle the auth part, but I'm still stuck on some sort of way to show me a list of these servers. [16:25] spowers: you could put them into an ldap directory [16:25] working on that, actually [16:25] the main reason i'm using ubuntu-server instead of debian is the easy availability of likewise-open [16:26] which is fantastic in our windows environment [16:26] the list part is really more about how do i put the list in front of my eyeballs [16:26] most of my experience dealing with linux has been about how to safely cram as much stuff onto a single machine as possible, and the whole thing is inverting right now [16:26] ldapsearch would work to list them... but also maybe just list them in a wiki page? [16:27] hmm [16:27] i wonder if i can tell firefox to handle ssh:// or something [16:28] not sure about that, I was just thinking about my server specific documentation is in a wiki [16:29] you might be able to do something creative like place links to a shell script that will open a terminal and start a telnet session :) [16:32] how do i increase the size of my lvm root partition? [16:44] with live cd [16:44] ivoks: bacula uploaded to hardy-proposed againa now we wait :) [16:45] Does anyone use an off site Back-up service that they would recommend? === trashguy_ is now known as TrashGUY [17:06] ivoks: live cd doesn't seems to have lvm enabled.. [17:09] acemo: did you lvchange -ay ? [17:10] or lvm vgchange -ay [17:10] or lvm vgchange -ay [17:11] thefish: i tried to use lvextend but that gives a bunch of errors [17:11] acemo: what kind of errors? [17:11] did lvdisplay show the volumes? [17:12] lvdisplay also gives a few errors.. but does show the volumes.. lemme type over the errors [17:14] "/proc/misc: No entry for device-mapper found. Is device-mapper driver missing from kernel? Failure to communicate with kernel device-mapper driver." that 2 times and then "Incompatible libdevmapper 1.02.27 (2008-06-25)(compat) and kernel driver" [17:16] mm ok maybe the livecd kernel doesnt have lvm :/ [17:17] nah first time i tried lvm it said its not installed yet and have to type this to install so i installed hoping it would work [17:18] is it posisble with the ubuntu-server cd to setup a lvm that uses 4 disks during the installation? [17:18] acemo: what does vgchange -a y do? [17:19] it used to be that only the alternative cd supports lvm, so i always take that one [17:19] dont know if its still the case [17:19] same errors + the text 0 logical volume(s) in volume group "Leonardo" now active [17:20] mm wierd it sees the group [17:20] live cd and server cd do support creating a lvm during the install [17:20] thefish: I think that is still the case [17:20] acemo: really, I didn't know the desktop cd could. [17:20] acemo: ^ there you go then [17:20] dont have a knoppix disk about do you? [17:21] or the alternate cd [17:21] thefish: i got an empty cd and a dvd burner :) [17:21] ;) [17:22] with all due respect to the various *buntu CDs, i still keep a knoppix disk about [17:22] im sure theres a *buntu cd with similar powers somewhere [17:23] if you have a server disk about you could try that as well [17:24] acemo: actuall... still there? [17:24] modprobe dm-mod [17:24] still here [17:24] sudo modprobe dm-mod [17:25] try that ^ [17:25] it gives no output [17:25] shouldnt [17:25] alright [17:26] not try vgdisplay etc [17:26] now [17:26] it shows the volume group now, with no errors [17:27] huzzah! [17:27] so now you should be able to proceed as before [17:27] lvextend -L+1G /dev/Leonardo/root seems to work.. [17:27] kees: jdstrand: how can I run the test-dovecot.py from qa-regression-testing? [17:27] atleast it says its resized successfully [17:27] acemo: dont forget you also need to resize the fs [17:28] kees: jdstrand: I've tried sudo ./test-dovecot.py -v but all the tests fails [17:28] is there a command to let it extend to the max w.o having to calculate how big this will be? [17:28] mathiaz: it should have instructions at the top [17:28] acemo: whats the fs? [17:28] ext3 [17:28] mathiaz: did you install all the packages? [17:28] jdstrand: I think so [17:29] acemo: ext2resize does it online iirc [17:29] i ment extending the lvm volume [17:29] and no need to do calcs, it just voodoos it [17:29] mathiaz: apt-get -y install python-unit dovecot-imapd dovecot-pop3d procmail [17:29] i like voodoo [17:29] jdstrand: yes - I've installed the packages in my vm [17:29] jdstrand: I'm not running them from a schroot [17:29] mathiaz: it may need to be updated for intrepid-- I'm trying now [17:30] mathiaz: I nearly always run from a vm [17:30] jdstrand: the assertion error is: assert (os.path.exists(pidfile)) [17:30] mathiaz: oh-- well, what is the pidfile for dovecot in intrepid? [17:30] acemo: i cont remember the exact commands, but im sure a bit of google etc should get you there [17:30] mathiaz: I think it is using /var/run/dovecot/master.pid [17:30] (the script) [17:31] thefish: great, thanks a million [17:31] no worries :) [17:33] mathiaz: dovecot is working fine here (1:1.1.2-1ubuntu7) [17:34] mathiaz: are you testing 1.1.3? [17:34] jdstrand: testing 1.1.4 [17:34] mathiaz: did it move it's pidfile? [17:35] mathiaz: just look in /var/run/dovecot after you start it outside of test-dovecot.py [17:35] jdstrand: hm - I suck. only dovecot-imapd was installed. not dovecot-pop3d [17:36] mathiaz: heh-- that was *exactly* why I added the pidfile check yesterday :) [17:36] jdstrand: tests are working correctly now :D [17:36] mathiaz: I spent quite a bit of time on that with my ruby update (which uses testlib_dovecot.py) [17:36] mathiaz: \o/ [17:36] jdstrand: are the tests the same as the one that are added to the package? [17:37] zul: SInce you uploaded cyrus-sasl2, please do a no change upload of cyrus-sasl2-heimdal to unbreak it. [17:37] mathiaz: no-- these are package tests, not binary tests [17:37] mathiaz: eg, does the installed package still work [17:37] mathiaz: in various common (and not so common) configurations [17:37] jdstrand: right - the dovecot packages has tests/ directory [17:38] mathiaz: think of qa-regression-testing as Ubuntu specific tests [17:38] jdstrand: which is a delta we carry from debian. [17:38] mathiaz: it has nothing to do with build testing [17:38] mathiaz: oh, well, you could maybe integrate them into q-r-t to get rid of the delta :) [17:38] jdstrand: right - I think the tests/ directory is supposed to be package testing. [17:39] jdstrand: looking at debian/tests/ it seems that testlib_dovecot.py is a similar to the one in qa-regression-testing [17:39] mathiaz: I see-- I don't know the dovecot packaging well. usually tests/ comes from upstream and is used by 'make check', etc. I assumed that's what it was [17:40] mathiaz: I'm curious who put it there :) [17:40] jdstrand: IIRC it's ian jackson work on using automated tests. [17:40] jdstrand: he was working on something like that last year. [17:41] mathiaz: well, normally the scripts in q-r-t are destructive, which may not always be appropriate in a build... [17:41] but I don't know anything about the dovecot tests/ [17:41] jdstrand: right - debian/tests/ are not run during the build process. [17:41] mathiaz: I'd yank them out and integrate it into q-r-t then [17:42] jdstrand: could you have a quick look at the files - they seems to be really close to what's in q-r-t [17:43] mathiaz: so you are creating a 1.1.4 package based on 1.1.2-1ubuntu7? [17:43] mathiaz: and deciding if this can be dropped? [17:44] hi all! [17:44] jdstrand: yes [17:44] mathiaz: the testlib_dovecot.py stuff was copied out of q-r-t by pitti when ian jackson was working on build-test-automation [17:44] is anybody here that can help with nfs4, ldap and kerberos? [17:44] what's in q-r-t is better/newer [17:45] kees: ok - that's what I thought - I guess I can drop this delta then. [17:45] mathiaz: yank it-- q-r-t is more up to date [17:45] thefish: this is great.. the program ext2resize is currently not installed you can install it by typing bla bla.. dun works.. ~kicks crappy live cd~ [17:45] 5 [17:45] I've just asked on #ubuntu-devel if autopkgtest is still run [17:45] o/ [17:45] acemo: :( [17:46] acemo: what about resize2fs? [17:46] * jdstrand is disappointed by the split high-five... [17:46] thefish: ext2resize should work when booting from hard disk? [17:46] acemo: you want resize2fs instead of ext2resize, IIRC [17:46] ^ [17:46] resize2fs works but no magic voodoo :( [17:46] magic voodoo? [17:47] i have definately used it for shrinkage, and it was all automagic [17:47] kees: i'll take two [17:47] mneptok: coming right up! [17:47] * mneptok puts on his bib and grabs his pincushion doll [17:48] acemo: you can give it a Gb value though no? like 200G [17:48] ext2resize is not supported by the ext3 developers -- resize2fs is the right tool to use. [17:48] in fact, ext2resize has been removed from the repository recently because it's so broken. [17:48] thefish: it says Usage: resize2fs [-d debug_flags] [-f] [-F] [-M] [-M] [-P] [-p] device [new_size] [17:49] `and size can be in s,K,M or G :) [17:49] yep, just found that on manpage [17:50] wierd, doesnt seem to be a "expand as much as you can please" option [17:50] indeed [17:50] theres a shrink version of that though [17:50] http://manpages.ubuntu.com/manpages/intrepid/en/man8/resize2fs.html [17:50] thefish: resize2fs shrinks. [17:50] resize2fs expands too [17:51] thefish: to match the size of the underlying device on expand, just leave off the "new_size" (that's why it's in []'s) [17:51] cool [17:51] acemo: there you go then ^ [17:52] nice one kees :) [17:52] aweasome kees [17:52] theres so much love here right now [17:52] now to find what the device is and am done :) [17:52] * thefish feels all warm inside [17:53] hehe [17:53] * kees hugs everyone [17:54] how i love installing programs on a live cd :) [17:55] greetings, how do i check if postfix is "cotacting" saslauthd? for mail-sending, if i start saslauthd with debug i do not get any information on a connect, setup is chroot, /etc/default/saslauthd is changed, where could i look for errors? [17:58] sda, sda1, sda2, sda5, sdb, sdb1, sdc, sdc1, sdd, sdd1, sr0.. thats my disk devices.. im guessing sr0 would be my lvm? [17:58] or no.. that would be the ram drive? [17:58] acemo: you dont have it under /dev/mapper/... [17:58] ? [17:59] thefish: depends what the installer put it under [17:59] ls /dev/mapper should answer that quickly though ;) [17:59] only a yellow control [18:01] what about /dev/{volgroupname} ? [18:01] vgdisplay will give you the group name [18:01] ScottK: done [18:01] zul: Thanks. [18:02] thefish: Leonardo is the VG Name, but there is no /dev/Leonardo [18:05] thefish: okay lets install gparted and see if that will make it clear where the device is hiding itself [18:07] thefish: gparted shows /dev/sda1 (wich is /boot), /dev/sda2 with a sub partition /dev/sda5 (wich is flagged lvm) [18:07] thefish: conclusion: gparted does not likes lvm. [18:10] /dev/sda5 is the partition holding all lvm i think [18:10] acemo: did you `sudo ls /dev/Leonardo` ? [18:10] cos its prolly owned by root [18:11] did sudo -i so am in root all the time [18:11] ah [18:11] in gparted there also is /dev/sdb1 and sdc1 and sdd1 all got a warning sign (the sda5 had too) [18:12] ye dont change those [18:12] did you install lvm2? [18:12] and going on forums made me notice that gparted can't handle lvm atm [18:12] yup [18:12] kay [18:12] what does lvdisplay show? [18:13] /dev/Leonardo/root and /dev/Leonardo/swap_1 [18:13] wierd, and you cant ls /dev/Leonardo/root ? [18:14] cos thats what i would use as the target for resize2fs [18:14] nope, no such file or directory [18:14] bugger [18:14] sorry, gotta go - good luck :) [18:14] thanks =) [18:14] ill let u know if i see u ever again how and if i fixed it [18:17] thefish: resize2fs /dev/Leonardo/root seems to work when booting from hard disk. [18:21] Performing an on-line resize of /dev/Leonardo/root to 486959104 (4k) blocks. seems good to me.. [18:21] does anybody know if I can use nfs4 with ldap if they are setup in different machines? [18:21] ls [18:21] woops [18:25] jdstrand: Were the Ruby issues fixed in Intrepid too or is that still oustanding? [18:30] acemo: nice to hear :) that sounds promising then [18:32] nijaba: should usplash be part of JeOS? [19:23] Hi [19:24] I can't create new entries with phpLDAPadmin on Hardy. /var/log/apache2/error.log says: File does not exist: /var/www/htdocs, referer: http://real.localhost/phpldapadmin/cmd.php?cmd=template_engine&server_id=0&container=cn%3Dadmin%2Cdc%3Drainct%2Cdc%3Dhomelinux%2Cdc%3Dnet Any idea? [19:28] RainCT: did you use the config file from the package? [19:29] sommer: I guess so, I haven't changed anything of phpLDAPadmin's config. [19:31] RainCT: mmm... well from the error it seems to be looking for /var/www/htdocs instead of /usr/share/phpldapadmin/htdocs/ [19:32] Right, but I haven't changed anything and grepping in /etc/phpldapadmin and /etc/apache2 I can't find anything either :/ [19:33] that is strange... you might double check the phpldapadmin apache config file though [19:38] sommer: http://paste.ubuntu.com/56075/plain/ (the VirtualHost is new, I added to try if using it might magically fixed the problem but it doesn't work neither there -although it doesn't log an error message, neither-) [19:39] you're loading php4? [19:40] sommer: no, it isn't even installed [19:40] RainCT: ya, just noticed that it's in the default config [19:41] RainCT: the vhost is enabled right? [19:41] and you restarted apaache [19:42] Yes. (What do you mean with "enabled"? It works :P) [19:42] wait... phpldapadmin is working? [19:43] sommer: Yes. But if I go to "Create new entry here" and set up a new User the form doesn't do anything on submit [19:43] oooohhh, sorry I misunderstood... sounds like a phpldapadmin bug then [19:44] well, but I guess someone would have already screamed if this 'little' feature doesn't work :P [19:44] but I couldn't find anything about it on Google [19:44] you might also double check the acls on the ldap directory [19:47] I've never touched ACLs so they should be the default ones.. [19:48] and modifying the cn=admin account works [19:49] does the account configured in phpldapadmin have rights? [19:49] it's the admin one, yes [19:51] RainCT: you might try enabling phpldapadmin's debug options [19:51] RainCT: and try running slapd with -d -1 [19:53] (I've just tried creating a user through "import" and writing it in LDIF and from there it works) [19:54] zul: It looks like it's more complicated than I'd guess on cyrus. Someone merged a new heimdal version from Debian, but not the regular one so they're hopelessly out of sync now. [19:54] Argh. [19:56] sommer: argh!! [19:56] sommer: it failed because I had no group and it seems like the form requires one but doesn't warn about it -.- [19:56] sommer: thanks for your time, though :) [19:57] Sounds like something for documentation ... [19:57] RainCT: np [19:58] ScottK: maybe, also could be a phpldapadmin upstream bug :) [20:01] * RainCT is filling one :) [20:02] ScottK: eh? [20:03] zul: the -heimdal packages has a very tight depends relationship with the main package so it has to have an identical version-revision to work. I assumed they were in sync which is why I said you needed to bump -heimdal [20:03] zul: Bad assumption on my part. [20:03] ScottK: k ill have a look [20:03] So now -heimdal is broken until the main package gets up to the same version. [20:03] zul: Thanks. [20:03] or if you want to have a look be my guest [20:08] ScottK: ruby1.8 it already fixed in intrepid. I am testing if a sync from debian is viable for ruby1.9 [20:08] * ScottK is quite tied up in $WORK and stuff. [20:08] jdstrand: Thanks. [20:13] I could ping from ubuntu server to windows server. I could NOT ping from windows server to ubuntu server. I added the win server to my ubuntu hosts file and then pinging worked in all directions - what's that about? [20:13] I think I'm missing something fundamental here (it wasn't even the hostname I couldn't ping - it was the IP too) === TrashGUY is now known as trashguy [20:46] Is it safe/secure to put your www file in /home/user file or should i keep it in /var? [20:46] (apache2 web server) [21:04] Delvien: I don't see how it would make any difference, as long as the files have proper permissions [21:06] New bug: #281423 in dovecot (main) "[ffe] update dovecot to 1.1.4" [Undecided,New] https://launchpad.net/bugs/281423 [21:08] RainCT im not sure what permission i need, never dealt with apache or web hosting [21:10] Delvien: well, chmod 650 for directories and 640 for files, and owner .www-data should work [21:10] ty [21:10] *750 for directories === erichammond1 is now known as erichammond [21:43] mathiaz: No upstart IIRC [21:44] nijaba: usplash [21:44] nijaba: it's the graphical booting process [21:50] nijaba: so the plan now is to add another option to the boot menu named "Install a minimal virtual machine" [21:50] nijaba: and that would create a system with a -virtual kernel and only ubuntu-minimal installed. [21:51] mathiaz: yes, upsplash is what I meant, I do hope we have upstart, sorry [21:51] sounds like a good plan. Thanks for doing it [21:52] nijaba: ok - IIRC jeos would install system that has graphical booting system [21:52] nijaba: to make it prettier [21:52] nijaba: if we can drop that, the better [21:52] nijaba: because usplash was on the jeos iso, whereas it's not on the server iso [21:52] mathiaz: I am pretty sure we never had usplash [21:53] nijaba: ok - so problem solved :D [21:53] mathiaz: or at least it was not displayed on boot [21:53] * nijaba takes off to a concert [22:29] meh. how doez I bludgeon cups into printing (via smb) to a PCL-only-configured HP2605 ? [23:02] mathiaz: ping [23:06] mathiaz: i'm going to close 279754, and open a new one for the bootloader issue [23:06] bug 279754 [23:06] Launchpad bug 279754 in kvm "Intrepid fails to install onto virtio block device" [High,Triaged] https://launchpad.net/bugs/279754 [23:06] mathiaz: the error reported in that bug is now fixed [23:06] mathiaz: but, the installed system will not boot [23:06] mathiaz: it's a bootloader problem [23:06] mathiaz: i'm hacking on grub now [23:07] kirkland: right - seems like a good plan. Or you could open a task for grub. [23:08] kirkland: does grub-install recognize /dev/vd* devices? [23:08] mathiaz: given my experience with the raid threads, i'd like to keep bugs focused on a precise issue [23:08] mathiaz: yeah, but it doesn't install correctly [23:08] mathiaz: i installed two systems with the same iso [23:08] kirkland: IIRC that's what had to be fixed in hardy to make the installer recognize virtio block devices. [23:08] mathiaz: one using a standard kvm disk, the other using virtio [23:08] mathiaz: both installs completed without error [23:09] Just installed ubuntu 8.04.1 fresh and installed xen and when I boot into xen kernel network gets dhcp but can't ping lan or wan. If I boot back into normal kernel network works fine. can someone give me a hand? [23:09] mathiaz: but the virtio installed system wouldn't boot [23:09] mathiaz: after install, i switched the machine i installed with a standard disk to a virtio disk [23:09] mathiaz: it booted and ran fine, with virtio disk [23:09] mathiaz: then, on that system with a virtio disk, i tried re-running grub-install [23:09] mathiaz: updated /boot/grub/devices.map accordingly [23:09] kirkland: ok -that seems like an issue with grub-install [23:10] mathiaz: that bombed the machine, made it unbootable [23:10] kirkland, please use pastebin as to not flood the channel [23:10] joebob777as7: huh? [23:11] whois joebob777as7 [23:11] http://pastebin.ca so you don't paste a bunch in the channel [23:11] whois RainCT [23:12] joebob777as7: dude, take it easy ... all of that is hand-typed conversation, intended for mathiaz. pastebin's are used for pointing to giant blobs of generated data or files [23:13] good riddance [23:13] kirkland: could it be possible that the geometry of the virtio block device is wrongly reported? [23:14] mathiaz: that is very, very possible [23:14] heh. /me is off for today, good night [23:14] night [23:14] * mathiaz waves at RainCT [23:14] mathiaz: i've been digging, but i haven't seen anything obvious [23:14] mathiaz: there's regex support in the script for *vd* disks [23:15] mathiaz: so someone has *tried* to get it to work [23:15] mathiaz: i was actually about to try it with fedora [23:15] and see how we differ [23:15] kirkland: does grub report an error while booting? [23:15] kirkland: like an error code? or is it just stuck on loading grub? [23:15] mathiaz: its more like "no bootable disks" [23:16] mathiaz: i can screen capture it, one sec [23:17] kirkland: hm - may the partition table is borked by grub-install. Could you try to a vm with the broken disk attach as a secondary hd and look at the partition table? [23:17] mathiaz: http://people.ubuntu.com/~kirkland/Screenshot.png [23:18] kirkland: ah ok - so grub is found [23:18] mathiaz: hmm, i don't think that's necessarily the case [23:18] mathiaz: doesn't even look like it hit stage 1 [23:18] kirkland: right - but the BIOS detected a bootable drive [23:19] kirkland: what you could try to do is to look at the first bytes of both drives you have [23:19] kirkland: the MBR should be the same [23:20] kirkland: you may be able to figure out if there is an issue there. [23:21] kirkland: IIRC soren had to do something special in ubuntu-vm-builder to make the generated images boot correctly. [23:25] mathiaz: hrm, okay [23:25] mathiaz: https://bugs.edge.launchpad.net/ubuntu/+source/grub/+bug/281492 [23:25] Launchpad bug 281492 in grub "grub does not install properly to a virtio disk in kvm (/dev/vda)" [Undecided,New] [23:27] .away [23:30] mathiaz: what's the current source package for ubuntu-vm-builder? [23:30] mathiaz: i'll check that source [23:30] kirkland: well - it has been renamed to vm-builder in intrepid [23:31] kirkland: so you may wanna look at ubuntu-vm-builder in hardy [23:31] kirkland: soren rewrote the whole thing in python for intrepid. [23:31] mathiaz: hmm, is it fixed or broken in intrepid? [23:31] kirkland: I don't know. [23:31] hmm, okay [23:32] kirkland: I'm not sure it's the exact the issue you've reported. [23:32] kirkland: it may be related [23:46] mathiaz: i've grepped my irc logs and can't find it... [23:46] mathiaz: what was the kvm virtio invocation you gave me for virtio networking? [23:46] mathiaz: i want to make sure i'm testing it the same way you did [23:48] heya [23:48] emgent: hi [23:48] hey kirkland :) [23:49] sommer: btw, thanks for the words of encouragement ;-) [23:49] sommer: i made a mental note to thank you when i got a chance ;-) [23:50] kirkland: -net nic,macaddr=52:54:00:25:84:fa,vlan=0,model=virtio [23:50] mathiaz: \o/ thx. [23:53] mathiaz: do i need to do anything on my host? [23:53] mathiaz: and that mac address can be anything, really, right? [23:56] kirkland: the mac adress can be anything - usually I use 52:54:00 as this is the vendor id for qemu [23:57] kirkland: as for the host side I'm not sure [23:57] kirkland: I'm using libvirt to handle all of this