=== dendro-afk is now known as dendrobates === shade_ is now known as shade\ [00:51] on a fresh install I installed samba4 out of curiosity then uninstalled it. It's left some cruft that I think is interfearing with the samba3 install... doing a ps aux -e | grep samba & smbd & nmbd shows all are loaded.. #samba says that the samba process is related to version 4 and smbd & nmbd for version 3.. how do I erradicate samba4? Just delete all related files? [00:52] Scunizi: Better to reinstall and then purge instead of just remove. [00:52] Then if there's anything left and you remove it, the package manager won't remember you removed it by hand and reinstall it properly if you ever install the package again. [00:53] ScottK: I tried purge after remove.. of course it didn't work.. will installing samba4 eliminate samba3's installation? or replace my smb.conf file? [00:53] Scunizi: Purge only works if the package is installed. [00:54] It shouldn't, but if you're concerned, make a copy of it first. [00:54] did.. thanks.. [00:54] * ScottK isn't much of a samba person, so can't say for sure. [00:55] ScottK: just so I have the right purge command.. sudo apt-get purge samba4 ..? === oubiwann is now known as oubiwann-away [00:55] yes. [00:55] (assuming samba4 is the binary package name) [00:55] ok.. thanks [01:02] Also had to do a apt-get autoremove for some of the extra packages. [02:23] i have multiple subdomains on my home server [02:23] i used to keep the subdomain files in different folders in /var/web [02:24] is this a normal place to keep it? === unreal_ is now known as unreal === unreal_ is now known as unreal [03:22] I have a server using software raid in a esata enclosure. I physically moved the location of the server and the enclosure and now my raid array is resyncing am I loosing data? [03:39] http://pastebin.com/Z5DaVRvL <-- can't kill samba ... any ideas? [03:52] Italian_Plumber, kill -9 [03:53] if that doesn't work, the only way I know of is rebooting === luis_lopez is now known as ` === ` is now known as luis_lopez [03:55] :) thanks... I tried that and it seemst o have worked. [04:15] does anyone know how to connect wirelessly via CLI? [04:18] harrison, this is constantly asked here. I didn't find yet a wiki/tutorial/howto about that. If you find, please, tell me so I'll be able to answer other people. [04:25] http://blog.tplus1.com/index.php/2008/06/13/how-to-connect-to-a-wireless-network-from-the-ubuntu-command-line/ [04:25] there is the link [04:26] harrison, did it work for you? Can I recommend this link? [04:27] yes [04:30] What's some really amazing thing about Ubuntu that I could use in a report? [04:31] What he says is "Give an "Oh Wow" fact about the project." [04:32] MTecknology, the best features of ubuntu over other linux distros aren't "oh wow" things [04:32] over windows? [04:33] in general [04:33] in my opinion, Ubuntu's best feature is the simplified nature of the total package [04:34] it gets a lot of it from the debian upstream, but then ubuntu picks a well-thought out subset and puts it together nicely, very little work required [04:34] I was thinking something like, 50 billion people contributed to it [04:35] harrison, thanks [04:35] That is true though - it puts together an awesome default setup [04:35] I guess the crowning moment for me was when they added the LAMP tasksel [04:35] one check box for a working web server [04:36] and now tasksel is going away.... [04:36] :( [04:36] what's it being replaced with? [04:36] So far I've heard nothing [04:36] aptitude is being removed from the default install too [04:40] I always use apt-get anyway [04:40] I know aptitude offers more features but old habits die hard [04:40] it IS kind of silly to include both in the default install [04:41] but tasksel is one of the best features of ubuntu server [04:41] if they don't replace it with something they're seriously hurting it [04:42] qman__: indeed [04:50] qman__ what is tasksel? [04:51] harrison, `sudo tasksel` and see [04:51] okay [04:52] it also runs near the end of a ubuntu server install [04:54] That is something to remember [04:54] what is an edubuntu server? === Barre_ is now known as Barre [04:56] HI [04:57] hello [04:58] !edubuntu > ubottu [04:58] twb, please see my private message [04:58] Oops [04:58] !edubuntu > harrison [04:58] harrison, please see my private message [04:58] harrison: try "/msg ubottu foo" when you want to know "what is foo?" [04:58] how are you [04:59] fine [04:59] are you a system admin? [05:00] I will try that but what I wanted to know, does edubuntu have a server-client system? [05:00] Silentwhisper no I am not, someday maybe. [05:00] sorry, i have no idea [05:01] why you're here? [05:01] what is your careeer now? [05:02] Homeschool student. [05:02] This is a great start for you [05:02] i'm here to learn and shift career [05:02] in what way? [05:03] i hope i could do that a short span of time [05:03] What administer a server? [05:03] i would like to learn about server admin [05:04] I mostly browse the forum how-to's to learn how to install vsftpd and samba and the sort. [05:05] kylenet.dyndns.org is my server. [05:06] vsftpd to transfer files? [05:07] yes it's a ftp server. [05:09] samba to host files? [05:09] were you able to finish the installation [05:09] ? [05:09] what do you want to accomplish now? [05:09] yes [05:10] where are the files? [05:10] I have it setup on my parents computer, it runs win 7 and it works well. [05:12] I want to use postfix to, instead of having a unix account for each address, use LDAP to manage the users. Is there a guide for this? [05:12] What's postfix? [05:12] MTA. [05:14] It's 12:14 EST time for bed. (thats in Quebec, Canada) [05:15] what is the url for your files [05:31] where is your fileserver? [05:36] nowadays what are the skills i should learn a system admin? [05:36] as a system admin [06:01] What form does Postfix want LDAP users in? === mrjazzcat is now known as mrjazzcat-afk [06:33] No matter what form I put the password in, ldap stuff gives "ldap_bind: Invalid credentials (49)". [06:38] Nonpython: is the LDAP server slapd on lucid? [06:40] yes [06:40] On the server, does "slapcat" work? Does "ss -lp" report that slapd is listening to the appropriate port? [06:41] If so, then try to use "ldapsearch -x" on the server. [06:41] Probably ldapsearch -x -H ldap://127.0.0.1/ [06:43] Slapcat does not work. ss -lp reports that slapd is running on port 24726. [06:46] What does slapcat say? [06:46] Nothing. [06:47] I am just running "slapcat". No arguments. [06:47] What is its exit status? [06:50] 127. [06:50] That is "command not found" [06:50] You should be getting an error message from your shell -- you should investigate that sometime. [06:50] Are you running slapcat as root? [06:51] Lawl. That was from a typo. [06:51] Really it is 0. [06:53] OK; do you have any evidence that there are any objects in your LDAP database? [06:56] no [07:03] I think your LDAP is working but you don't have any objects [07:10] How do I get some? [07:16] ldapadd [08:37] I have 2 virtual hosts. One works, and one shows the same content as the other one, which is not right. What is wrong? [08:41] By "virtual hosts" do you mean apache vhosting? [08:50] yes [08:51] I don't do apache, sorry [09:01] twb: i got that error when having apache read the config files in the wrong order. so i renamed the other one to 00_name and that solved the problem. its about having a default one i guess. [09:33] hi all [09:33] i need help with shorewall. Is possible log mac address with masquerading? [09:46] hello! need help with UUIDs: i have 3 partitions on my disk, sda1, sda2 (primary) and sda5 (extended). now, for some unknown reason sda5 suddenly has no UUID and I can't mount it anymore. How to create a new UUID and attach it to sda5? [09:47] Depends on the filesystem [09:47] What does "sudo blkid /dev/sda5" say? [09:48] its ext4 [09:48] What's the whole line? [09:48] no result for blkid /dev/sda5 [09:49] That doesn't make sense [09:49] blkid gives me result only for sda1 and sda2 [09:49] What does "sudo file -s /dev/sda5" say? [09:50] "/dev/sda5: Minix filesystem [09:51] e2fsck -f says that filesystem is clean [09:53] You've fucked up your filesystem somehow [09:53] what I have already tried is tune2fs -U random /dev/sda5, but that didn't yield any result for me [09:54] What is the filesystem type for sda5 in the partition table? [09:54] IIRC it should be 83 for ext [09:54] but how? only users with limited right are working on the system, I didn't touch anything for at least a month or two :( [09:54] I don't know [09:55] how can I get this what you need? (83) [09:56] I would probably run "sudo cfdisk /dev/sda" [09:57] write new partition table? will all data then be lost? [09:58] That will *inspect* the partition table [09:58] Make a backup of it first if you're really paranoid [09:59] aha, no. I'm not that paranoid :) [10:00] well, i'm in cfdisk, all partitions are listed [10:00] And what type does it list for sda5? [10:01] sda1 (swap),sda2 (root, ext3) and sda5 (extended, ext3) (sry!, its ext3 after all) [10:02] sda5 is listed as logical ext3, linux [10:02] That's what it SHOULD say [10:02] I don't know what you've done to this box that it can't see it in file -s and blkid [10:04] this is computer of my aunt, she has 3 children and some of them must have done something. but they all have limited user accounts, what they can do?! [10:04] all is there except UUID [10:06] when I run 'ls -l /dev/disk/by-uuid' only sda1 and sda2 are listed [10:10] sir twb i was able to setup webserver and mail server quit and dirty [10:10] what else should i learn in maintaining server? [10:11] anyone can help me with shorewall? I need log macaddress of all my packet [10:16] someuser: /dev/disk/by-uuid won't see tune2fs changes until you restart udev [10:17] trinkity__: try #shorewall [10:17] i did shutdown -r now [10:18] but still no changes [10:42] twb: i did it on the old way, with /dev/sda5 instead of UUID in fstab. but still no explanation while UUID is suddenly missing. I will do the upgrade to 10.04 and hope that upgrade process will notice that UUID is missing and generate a new one for me. thank you for your help! [11:35] How do I run installation in text mode? i'm trying to install server on remote HP server using the ILO interface that I can access only via ssh, but as soon as installCD boots up it goes to graphics mode. Is there a way to circumvent that? [11:35] MmikeDOMA: boot with fb=false [11:36] Unless you're using 10.04, in which case you're fucked due to a bug in the server CD. [11:36] (Serial also works, as long as there's no graphics card installed.) [11:39] twb, i'm using 10.04. as soon as i get ISOLINUX booting message it goes to fb [11:39] so I'm fucked [11:40] Neat :) === someuser_ is now known as someuser [11:40] MmikeDOMA: you can either preseed past it and do the install over ssh, roll your own CD, or netboot [11:41] Or do a d-i-less install from an existing unix system [11:41] in try in channel #shorewall but nothing help for me [11:41] i need log mac address in shorewall log [11:42] twb: problem is that ILO remcons won't let me press any keys because it's in graphics mode. It tells me to use web graphics interface which I can't because ILO adapter is on private network, i used ssh port-forwarding to access it. And I don't have hands on site. [11:42] Oh, well. [11:42] MmikeDOMA: yeah, I hate that stuff with a passion [11:42] I'm begining to grow a passion for that hate these days, yes. [11:42] At least your LOM actually supported direct ssh [11:43] Most of the routers and loms I get thrown at ONLY support some goddamn activex-based web UI [11:44] Maybe there is a way to fool the java applet [11:44] MmikeDOMA: I assume you're nowhere near the machine? [11:45] twb, 2 hours drive. I'll be there tomorrow but I wanted to prepup from home. [11:45] And I'm not going there now. [11:45] Nod. [11:45] Can you boot the host off PXE? [11:47] Nope. [11:47] But, would the installCD recognize the serial console if attached? [11:47] It'll support it, but you'll need to pass parameters via the VGA console first :-( [11:48] (You need to do that for 8.04, too, it's just I assumed you could type that blind.) [11:49] Gnj! [11:50] So, the only thing I'm left to is making java applet to connect trough the ssh-forwarded connectio [11:50] Good luck [11:51] Why does a server install CD goes to framebuffer anyway? [11:51] Because of a fancier look? [11:51] MmikeDOMA: That's a very good question. [11:51] It broke stuff for much people [11:52] And has no use [11:53] MmikeDOMA: just to piss me off [11:53] that's the ONLY reason [11:53] The whole world is against me [11:58] Hi server guys. Remember that little django app for celebrating the 10.04.1 release I mentioned yesterday. Can anyone provide me a publically reachable VM to test it on (don't wanna waste time waiting for IS) [11:59] twb: Ehm, but we agree with you! :) [12:00] twb, i'm with you too ;) [12:00] The fact that plymouth is a required part of ubuntu-minimal, and that even with "splash" removed from the boot parameters, it's still *running*, is enough of a turn-off that everyone in my office is turning to one another and saying "so, remind me, why did we switch to Ubuntu again?" [12:01] Yes [12:01] implementing plymouth was stupid, and useless. [12:01] And they did it in an LTS release [12:01] which isn't supposed to be usefull at all! [12:02] And because i915 is enabled by default, my cheapo VGA-only LCD monitors chop off the leftmost column on the tty, so I have to spend five minutes every other day pissing about with their horizontal alignment. The old 80x25 default was big enough that it wasn't noticable. [12:03] The other one that has screwed me hard so far is the race conditions in upstart/mountall when booting from NFS [12:04] To "fix" it I had to change mountall-net.conf to just "killall -USR1 mountall" every tenth of a second, forever [12:05] Where does one report a bug for ubuntu-server? Same place as for desktop? [12:05] MmikeDOMA: launchpad somewhere [12:06] Try "BROWSER=w3m apport-bug" or so [12:06] https://help.launchpad.net/Bugs/EmailInterface is useful [12:13] wat is the recomended way to add and configure a bridge at boot? [12:14] add a br0 in /etc/network/interfaces/ [12:14] add a br0 in /etc/network/interfaces [12:14] Unless you're using libvirt, in which case it probably shits over that and does it internally [12:14] No it doesn't [12:14] I stand corrected [12:15] I just remember installing it somewhere and going "aah, wtf are you doing to my configuration?!" before going back to invoking qemu by hand. [12:16] Hmm, never had that [12:16] Jeeves_: will it also create the bridge and add the proper intefaces to it??? [12:17] a_ok: brctl's package has a supplementary manpage for interfaces(5) that makes it very clear [12:17] twb: by default libvirt should not do anything with bridging as it is optional [12:17] If you configure it correct, yes :) [12:17] !pastebin [12:17] For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://tinyurl.com/imagebin | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [12:17] twb: ah thanks! [12:17] a_ok: http://paste.ubuntu.com/470214/ [12:17] apt-get install bridge-utils; man 5 bridge-utils-interfaces [12:18] that should teach me to prefere google over manpages [12:30] Jeeves_: those are some sensable defaults, verry nice === oubiwann-away is now known as oubiwann [12:52] morning all [12:53] morning sommer :) [12:53] sommer: are you in EST? [12:54] yeppers :-) [12:54] same here, just got to work. [12:54] oh ya starting the day, heh [12:55] i'm gonna go find some coffee [13:00] twb, I did it! [13:00] twb, i needed to port-forward port 23 from my homebox to the ILO adapter [13:00] and now i have the neaty-neat graphical display of a ubuntu server install [13:01] stupid! [13:01] MmikeDOMA: you should post it on your blog or whatever [13:02] So that the next guy can google and work it out [13:02] Sure thing [13:05] Just to be clear: you're *reverse* forwarding 23 back to the ilo host? [13:06] twb: yes. I opened 23 on localhost and ssh-tunneled it to 23 on the ilo [13:09] does anybody know of an isp control panel that doesn't need postfix/apache2/bind to be locally installed? [13:10] <_ruben> directadmin [13:12] _ruben: it's for a charity so it needs to be free [13:22] morning [13:23] kirkland, here now if are [13:23] good morning mr zul [13:23] smoser: sup? [13:23] hey smoser [13:23] responding to ping about bug 610134 which i see you fix-commited. [13:23] Launchpad bug 610134 in byobu "'byobu -S _name_' is broken" [Low,Fix committed] https://launchpad.net/bugs/610134 [13:23] smoser: oh, i was just asking about your byobu bug [13:24] smoser: yeah, it's fixed, such that you can use byobu -S ... to name your sessions [13:24] smoser: however, byobu-select-session only looks for sessions that start with byobu* [13:24] i think thats reasonable [13:24] smoser: which should only affect you if you ssh into a system with multiple running byobu sessions, and you want to choose one [13:25] smoser: if you name your session byobu -S smoser, it won't appear in the list [13:25] yeah. hmm.., but one thing that would stnk about that. [13:25] if i had byobu by default enabled [13:26] and created a session with 'dev'. the select then wouldn't show it to me. [13:26] and may possibly not be able to attach at all when ssh'ing in [13:26] smoser: away for 5minutes, in a 1:1 [13:26] i guess i could work around with ssh -t bash, then byobu -d -r [13:47] is it possible for me to have 3 lvm volume groups on 1 physical volume? [13:54] AndyGraybeal, maybe, but why? you can just use one volume group and create as many volumes in it as you want [13:57] smoser: i could also force a byobu- in front of your "dev" when you do byobu -S dev [13:57] smoser: but i felt that might be rude [13:58] yeah, i was going to suggest that. [13:58] it may or may not be all that rude [13:58] screen -d -r 'dev' [13:58] would still attach [13:58] it does matching [13:59] zul: hey! [13:59] zul: what's the state of https://code.launchpad.net/~mathiaz/server-sru-tracker/visual-improvements/+merge/29329? [14:00] smoser: oh? cool.... [14:00] zul: mathiaz: morning guys [14:00] kirkland: o/ [14:01] kirkland: how is the east coast doing? [14:01] mathiaz: it's very, um, liberal :-) [14:01] kirkland: :) [14:01] mathiaz: :-) [14:01] hi kik [14:01] i'm wrong, kirkland [14:02] kirkland even [14:02] mathiaz: once i get it unbroken ill let yo know :) [14:02] it must match to the beginning only [14:03] which means you could change your selection to match at the end. [14:03] and always append .byobu [14:03] so that select would show all byobu sessions (searching by appending .byobu) [14:03] s/appending/ending/ ^ [14:04] qman__: thankj you for the response, i just came to the same conclusion a second ago after reading more. [14:04] bug 574910 is getting bad. [14:04] Launchpad bug 574910 in linux-ec2 "High load averages on Lucid while idling" [Undecided,In progress] https://launchpad.net/bugs/574910 [14:05] smoser: sweet, that sounds perfect [14:05] you should, of course, test my assertions. [14:05] :) [14:14] smoser, kirkland: about bug 574910, is it EC2-specific ? Or does it also affect UEC images / classic servers ? [14:14] Launchpad bug 574910 in linux-ec2 "High load averages on Lucid while idling" [Undecided,In progress] https://launchpad.net/bugs/574910 [14:15] i was just thinking that. [14:15] i have no idea on non-ec2 [14:15] but there is suggestion that it might be larger. i have no data at all about non-ec2. [14:15] ttx: unfortunately in our testing, we don't generally leave uec-instances run for very long [14:16] ttx: "does it respond on ssh? if yes: terminate; if no: mark test failed" [14:16] kirkland: could be good to compare some idle load [14:16] I'm pretty sure it's EC2/linux-ec2 specific [14:17] sounds like something we could integrate in our tests though [14:17] That last comment may be orthogonal to the issue, but talks about VirtualBox [14:17] what is good to use for an ftp server with ubuntu? [14:17] ttx, yeah, its becoming a metabug [14:18] Hey guys. Is there an easy way to figure out when last a disk was fsck'd? [14:18] smoser: the "virtualization sucks" metabug ? [14:18] but there is one other bug linked there about way-too-frequent wakeups that could cause such things. [14:18] more "performance on lucid sucks" metabug [14:18] jiboumans: agreed; we could run some long tests over the weekend, when the rig isn't doing anything else [14:18] i've just started testing something here. [14:19] I fired up A.) lucid image B.) lucid image with karmic kernel and ramdisk [14:19] on each, I am running [14:19] while : ; do read r < /proc/loadavg; echo "$(date): ${r}"; sleep 10 || break; done | tee log.txt [14:19] the goal of whic his to see if simply swapping the kernel makes the loadavg reporting go away. [14:21] What can I use for a ftp server, apparently filezilla doesn't have the server for ubuntu [14:21] I haven't seen clues to a general lucid issue. But it might indeed be a common issue that has more visible consequences in virtualized hosts [14:21] zatricky: tune2fs can set that time, so it probably has a way to read it as well [14:21] Jinxed-: vsftpd ? [14:21] so far, the lucid kernel is definitely showing spikes. [14:21] thanks, jiboumans. looking into that, will post result [14:22] ttx, will give it a try [14:22] thanks [14:23] SpamapS: Thanks for looking into the php5 patches for Kolab. Please let me know if there's anything I can do to help. [14:24] zatricky: yeah, tune2fs -l /dev/xxx [14:25] zatricky: Last checked: Sat May 15 00:54:20 2010 # on one of my boxes [14:25] cool. Just got tune2fs -l /dev/sda1 | grep Last\ checked, gives a similar result [14:25] thanks, jiboumans [14:25] my pleasure === mrjazzcat-afk is now known as mrjazzcat [14:39] what is a simple GUI based ftp server I could use [14:40] Jinxed-: gui???? [14:41] a server isn't gui based [14:41] vsftpd is a good ftp server, though, and quite simple to configure [14:41] !webmin [14:41] 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. [14:41] !ebox [14:41] ebox is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/eBox [14:42] ebox can probably (or possibly) help you configure vsftpd [14:42] So if I get rid of pure-ftp and pure-ftp admin and get wsftp with ebox [14:42] the pure-ftp admin only had configuration for users etc [14:42] I want to know how I can connect with a filezilla client [14:42] and it doesn't seem to work [14:43] vsftpd, not ws [14:43] ok [14:43] vs = very secure [sic] [14:43] it's probably the most used ftp server on the planet, so I guess it's rather secure [14:43] Jinxed-: just public ftp, or private users as well? [14:45] RoyK, I just want to have ftp on a private network that would allow for both public and private users [14:45] ok [14:46] out of curiosity, why use ftp on a private network? [14:46] eh... RoyK there are alot of different options for ebox [14:46] why not cifs or nfs? [14:46] Jinxed-: it's probably just as easy to configure vsftpd in the config file [14:46] 1.) I don't know what those are. 2.) I like ftp :) [14:46] Jinxed-: are the PCs on linux or windows? [14:47] or mac [14:47] linux/windows [14:47] for the windows machines, run samba - that'll work with linux as well [14:47] just install samba, configure the shares in /etc/samba/smb.conf and you're done [14:47] shares suck [14:47] why? [14:47] Im locked out 70% of the time [14:48] that shouldn't happen [14:48] we're running samba on a 200 user network - it works [14:48] for the linux machines, you might want NFS, but then, SMB/CIFS works with linux as well [14:49] yeah... it seems to work well when i use ip addresses [14:49] but not when I use the network name [14:49] do you have a windows server in there as well? [14:50] nope [14:50] Im just running [14:50] ubuntu (desktop edition) [14:50] ok, setup samba as a windows NT domain controller and it'll just work [14:51] ehhh how would i do that [14:51] :) [14:51] http://tinyurl.com/34fpthl [14:57] he he [14:57] not very nice [14:57] How do I install a printer via the command line? [14:57] :? [14:57] ask roy -> lmgtfy -> " How do I install a printer via the command line" [14:58] :) [14:58] harrison: I think that's covered in the CUPS docs, though I haven't done it for years [14:58] Where would I find the CUPS docs? [14:59] lmgtfy [14:59] not [15:00] zul: do you have any idea how to solve that thing with bacula? [15:01] zul: predepend is what we want, but not going to do it :) [15:01] ivoks: no idea but that predepends is what i really want to do :) [15:02] me too === mrjazzcat is now known as mrjazzcat-afk [15:02] sooooo.....whats stoping us? :) [15:03] i doubt that someone will like that solution :) [15:03] harrison: this was the first hit from google http://www.linuxquestions.org/linux/answers/Networking/Setting_Up_a_Network_Printer_using_CUPS [15:03] network printer? [15:04] yes [15:04] it's really easy [15:04] smb or ipp? [15:04] you need to change two lines in one file [15:04] The diference? [15:04] in /etc/cups/cupsd.conf [15:04] Listen localhost:631 [15:04] to: [15:04] Listen 631 [15:05] Browsing Off [15:05] to: [15:05] Browing On [15:05] restart cups and that's it [15:05] Browsing, obviously [15:06] on all other mac and linux machines, printer will automagicaly apear by it self [15:06] no drivers, no configuring, nothing [15:06] What is the diference betwean SMB (samba) and IPP? [15:06] on windows, you'll need to add them manualy [15:06] okay. [15:06] IPP is printing protocol, SMB is file sharing [15:06] and can be misused for printing :) [15:07] i.e exploited? [15:07] but whoever tells you that SMB printer is on server, he doesn't understand meaning of the expression printer server [15:08] harrison: no, windows use SMB for printing [15:08] harrison: not really, it works that way too. also, samba can be used to distribute printer drivers [15:08] If I have Samba installed do I need anything else? [15:08] ivoks: a windoze or samba server can be setup as a spooler, but not really a print server [15:08] RoyK: right, clients do the processing [15:08] yeah [15:09] harrison: just setup a cups printer and configure samba to use that, sharing it with a samba share === mrjazzcat-afk is now known as mrjazzcat [15:09] he can use ipp on windows too [15:09] no need for samba [15:09] and to complicated [15:09] too [15:09] imho it's easier for windoze users to browse printers with smb [15:10] I don't think you can add IPP printers from a login script either [15:10] when in rome ... [15:11] ccheney, ping [15:11] or maybe kirkland or Daviey can help. i think i need to install a maverick uec. [15:11] and am interested in the automation goodness. [15:12] zul: the only problem is... [15:12] zul: bacula doesn't really depend on local mysql/pgsql server :D [15:13] yeah i know.. [15:13] stupd dbconfig [15:13] it's not dbconfig's fault [15:13] it's apt [15:13] in ubuntu desktop, I have to go to system -> admin -> printing and configure the printer there, in ubuntu server is it audo detected? [15:14] harrison: no, you'll need to add it manually to cups [15:14] harrison: it's http://localhost:631 [15:14] how? [15:14] Whats at http://localhost:631? [15:14] cups config [15:14] Ahh [15:15] if my server is at the url kylenet.dyndns.org should I subsitute localhost for the url? [15:15] if you replaces localhost:631 with 631 in cupsd.conf, then yes [15:15] replaced [15:16] i did [15:16] and restarted cups? [15:16] well, if you are accessing from non-localhost machines [15:16] you'll have to allow access [15:16] *slaps head* I forgot to restart CUPS! [15:17] How do I do that? [15:17] service cups restart [15:17] okay [15:18] just to tell you add sudo to that command. [15:18] still didn't work? [15:18] *. [15:20] of course you'll restart services as root [15:21] i don't find it neccessery to tell you that [15:21] good thing I read error meseges. [15:21] I just tried 'allow from 81.191.180.x' (my IP) where order deny allow was set, still can't access it..... [15:21] How do I allow access? [15:22] What part of /etc/cups/cupsd.conf do I add my desktop to? [15:22] RoyK: try Deny from none [15:25] harrison: just add ServerAlias * [15:25] harrison: at the end of the file [15:26] it did something now I get the error 403 fobidin. [15:26] *somewhere [15:26] add Allow from all [15:27] Where? [15:27] to each [15:27] there are only 3 of them [15:27] do I deleat the deny part? [15:28] you have a deny part? [15:28] Order allow,deny [15:28] === MmikeMRMA is now known as MmikeDOMA [15:28] that is a copy past [15:28] *paste [15:29] that's ordering [15:29] of /etc/cups/cupsd.conf [15:29] don't delete it [15:29] have you ever configured apache? [15:29] Not realy. [15:29] order allow,deny shouldn't deny anything access unless it's explicitly denied [15:29] RoyK: i'm not sure what defaults are [15:30] ivoks: order allow,deny will allow anything that's not explicitly denied [15:30] RoyK: maybe, if unspecified, deny is always from all [15:30] order deny,allow is used when you want to restrict access [15:30] RoyK: and allow is from localhost [15:30] I've only used Apache for 10 years or so, but I'm quite positive order allow,deny allows everything [15:31] seems they have done something 'smart' in there [15:31] RoyK: true, but for apache, if i don't specify anything, i want it to be available to everyone [15:32] RoyK: in cups, if i don't specify anything, i want it to be forbidden for everyone :) [15:32] sure, but then, order deny,allow should be the one used [15:32] RoyK: order allow,deny is just ordering [15:32] RoyK: what's important is what's allowed and what's denied [15:32] RoyK: and this could be different, by default, for cups [15:32] that's what I'm trying to say... [15:32] RoyK: in apache, default deny is none, and allow is all [15:33] RoyK: it could be that for cups, default is deny all, allow localhost [15:33] but i'm just guessing [15:33] that doesn't make sense - there's no default deny or allow AFAIK, just the order [15:34] order deny,allow says 'everything is denied unless explicitly allowed' [15:36] I still want to know where to put allow from all. ;) [15:39] under location [15:39] i've told you that already [15:39] don't delete anything, just add [15:39] above the alow deny? [15:39] or below? [15:39] does it matter? [15:39] no [15:39] okay [15:40] IT WORKS! [15:44] Thank you ivoks and RoyK that helped a lot. [15:48] i'm trying to create an lvm-snapshot. my machine just gets stuck there. here's the kern.log http://pastebin.com/KUzVGZSQ [15:58] trapmax: looks bad [15:58] like a deadlock between lvm and ext4? is that possible? [16:04] Daviey: are you on top of hggdh-detected UEC maverick regressions ? [16:06] ccheney, ping [16:08] mathiaz: can you upload the fix for openldap to maverick as well? === dendrobates is now known as dendro-afk [16:21] zul: I was looking at creating an apport hook for spamassassin, and the spec mentions to collect email headers... just wondering how to go about that? [16:21] sommer: im not sure...i didnt get that far :) [16:22] okay, I'll do some figurin, heh === dendro-afk is now known as dendrobates [16:33] zul: the openldap fix is not needed in maverick [16:33] mathiaz: k thanks [16:35] Daviey: around ? [16:47] ivoks: howdy [16:48] ivoks: are you around? [16:49] kirkland: yes [16:49] kirkland: howdy :) [16:54] smoser, hello whats up? [16:55] smoser, i'm off today sick, fever keeps rising and is 102f so far, going in to the doctor once my father in law can take me [16:55] was wanting maverick uec install scripts [16:55] smoser, its in the archive [16:55] rock on. [16:56] ivoks: hey, i think mrjazzcat and hallyn [16:56] so apt-get install uec-provisioning-* [16:56] ? [16:56] yes [16:56] smoser, the dhcp part isn't done yet so do the dd-wrt method [16:57] kirkland: mrjazzcat left couple of minutes ago [16:57] kirkland: did you mean to finish that sentence? [16:57] :) [16:57] i'm here fwiw [16:58] ivoks: hallyn: sorry ... we wanted to get a chat together :-) let me track down mrjazzcat === oubiwann is now known as oubiwann-away [17:08] kirkland, ccheney so is that availalbe anywhere for lucid ? [17:08] i've 3 boxes : stable lucid, 2 for cloud. would like to run install server on lucid. [17:10] in fstab, the last two numbers in the lines, what does this mean? i seem them being 0, 1, and 2. === oubiwann-away is now known as oubiwann [17:11] i found it on the web, nevermind [17:18] smoser, not yet, i think it will probably work if force installed on lucid [17:19] smoser, or locally rebuild with s/maverick/lucid/ [17:19] gracias [17:19] * ccheney bbl, going to try to make it to the doctor now [17:21] interesting my fever has dropped a full degree in the past hour [17:21] * ccheney wonders wtf is going on [17:28] you forgot to stick the thermometer back in the boiling water :) === mconigliaro_ is now known as mconigliaro [17:50] hey i had a question about the enterprise cloud... is it possible to install Windows OS's as part of the cloud or are you limited to Linux Distros? [17:50] windows installs fine [17:50] at least windows xp [17:52] picard1421: most windows OSes will install nicely [17:52] even win311 on DOS should work [17:53] the other question i had about cloud computing in general.. ... How "realisitic is it" i mean that in the sense.. if i have thinclients around the office.. how responsive will Installs be ETC? [17:53] picard1421 what kind of thin clients [17:54] picard1421 I have had people RDP into a virtual windows xp machine with no problem...they have no clue it is running a virtual machine [17:55] picard1421: on a gigabit or even 100mbps lan, you won't notice much [17:56] well it would be over LAN.. in the sense people would obviously on that note not know what is going on etc... [17:56] but what about (in theory) gaming.. [17:57] i know the server could handle hte graphics and the processing.. but how about the "screen" transfer over the LAN [17:57] i have also seen products like DVI over Ethernet [17:57] I don't think any good games work over any thin clients [17:57] are those even worth it.. or is the LAN enough? [17:58] gaming from a thin client isn't what you want [17:58] office apps will work well [17:59] you will never get the same performance on video with cloud/thinclient then you will directly connected to the monitor [18:17] I have a small home server mainly for file storage, should I keep the ubuntu os on a separate drive? === trappist_ is now known as trappist [18:18] I have a rack app listening on port 8000. after running for some time and handling a lot of traffic with no problem, suddenly port 8000 was responsive only to a small percentage of traffic... [18:19] even nmap showed it as filtered, and even from the local machine. restarting the app didn't help, but rebooting the server did, so I'm thinking something in the kernel [18:19] there were no iptables rules - what else could it be? [18:19] I have 7 other servers in the same situation, and I'd rather figure out the actual problem than just reboot them all and wait for it to happen again [18:25] sommer: i queued up your apport hook for openldap the next upload [18:27] hallyn, ping [18:33] zul: nice, thanks dude [18:33] sommer: i just changed the refrences from ssh to slapd :) [18:44] oh, woops missed that one, heh [18:55] Daviey: feeling adventurous? [19:08] hi guys, i'm trying to use apache with chroot, but i haven't get it working in hours.. [19:08] i'm on ubuntu 10.04, so, installed libapache2-mod-chroot, enable, restarted apache, and still on vhosts it says that the directory not exists, i've set relative the chroot..is it right? [19:14] alex88: have you enabled the module? [19:14] oh, you did [19:15] alex88: did you set ChrootDir in config? [19:15] ivoks: yes, and on start it says that document root not exist, using absolute path outside chroot works (just the start) [19:15] ivoks: yes, just under the pidfile, also linked [19:15] how can i make 'pgrep apache2' return only 1 pid? instead of all of them? [19:16] alex88: 'also linked'? [19:16] ivoks: yup, i'm searching the how-to i've followed [19:16] alex88: what's your chrootdir? [19:18] /var/www [19:18] and DocumentRoot? [19:19] well, i want something like /var/www/domain.com /var/www/domain2.com etc [19:19] so in virtualhost config i've set domain.com [19:19] /domain.com [19:19] yes sorry [19:19] ok [19:19] domain.com looks in /etc/apache [19:20] set it as /domain.com [19:20] it is [19:20] let me see, un second [19:20] you've linked pid file and logs directory? [19:20] i'll check again [19:21] i've set logs inside user dir [19:21] k [19:21] well, is /domain.com/www the documentroot and /domain.com/log the logs [19:22] ttx, ping [19:23] ivoks: so, i've restored backup, now it has default virtualhost with fcgid for php [19:23] http://www.alexnetwork.it/ [19:24] now, installing libapache2-mod-chroot [19:24] 2.6.18-028stab067.4-ent [19:24] ? [19:24] that's not ubuntu :D [19:25] vps :) [19:25] ubuntu 10.04, trust me [19:25] i know, i know [19:25] btw, installed, enabled [19:26] restart fine, site still works [19:26] add chrootdir [19:26] ChrootDir /var/www [19:27] stop apache [19:27] edit /etc/apache2/apache2.conf under PidFile ${APACHE_PID_FILE} added that line [19:27] fix pid file [19:29] mkdir -p /var/www/var/run chown -R root.root /var/www/var/run ln -s /var/www/var/run/apache2.pid /var/run/apache2.pid ? [19:29] yes [19:29] done, start? [19:29] yes [19:30] done, start with no errors (strange, it has to say /var/www/ not exists) [19:30] http://www.alexnetwork.it/ connection closed [19:30] btw i've done chown -R root:root not root.root [19:31] and what's in error.log? [19:31] now i have to set document root of 000-default as / right? [19:31] yes [19:33] http://pastebin.com/tWkKu8G3 something like this would be fine? [19:33] yes [19:33] but your apache isn't working [19:34] at all [19:34] check error.log [19:35] oh..errors on fcgid [19:35] i'll dismod for the moment [19:35] does anyone know what service takes care of logs? [19:35] because I have no message logs or boot logs [19:35] memoen: syslog [19:36] thanks alex88 :) [19:36] memoen: np [19:36] ivoks: (2)No such file or directory: could not open mime types config file /etc/mime.types [19:36] No such file or directory: Couldn't bind unix domain socket /var/run/apache2/cgisock.24464 [19:37] fcgi and chroot need additional fixing [19:37] paths are different [19:37] this is after disabling fcgid [19:37] disable fcgid [19:37] /var/run/apache2/cgisock.24464 [19:37] already done and restarted, those are message for next start [19:37] that stinks like fcgid [19:37] or cgi [19:38] removed cgid [19:38] http://pastebin.com/uCiagyaX [19:39] stop apache [19:39] check if apache process is running [19:40] ps -A | grep apache nothing, also netstat -tapn nothing [19:40] should /var/www be www-data owned? [19:41] nope, that's irrelevant [19:41] put LogLevel debug to config [19:43] http://pastebin.com/aGhkNyNa [19:43] I'm trying to set up pub key auth. I've done this hundreds of times before but never seen this. I created the pair (no password) w/ssh-keygen, then ssh-copy-id to the host i'm trying to connect to. all looks fine but when i try to use the identity file it asks for a password [19:43] anybody know what the deal is? [19:43] not much more [19:44] matt_keys: need to change sshd config to no accept password? [19:45] alex88: don't know... it should work [19:45] ivoks: damn.. [19:45] ivoks: thank you for your time [19:46] np [19:47] alex88: RSAAuthentication yes and PubkeyAuthentication yes, AuthorizedKeysFile is commented (defaults) [19:48] What's a reasonable amount of swap on a high-usage production server with 16GB RAM? [19:49] whee openldap is fun [19:49] aaditya: depends on what you're running on it [19:49] ivoks: i'm asking on https, hope thay can help [19:49] aaditya : oracle, for example, has requirements for swap sizing [19:53] aaditya : old rule of thumb was 2x RAM. with 16gb ram I wouldn't do that though... 2048mb tops [19:58] re [20:01] New bug: #611005 in antlr3 (main) "Merge antlr3 3.2-3 (main) from Debian unstable (main)" [Undecided,New] https://launchpad.net/bugs/611005 [20:03] ivoks: you there? [20:04] alex88: yes [20:04] ivoks: from 2.2.10 you don't need mod_chroot, everything works without mod, just add chrootdir to config.. -.-' [20:05] :) [20:05] New question on Shapado: What's the most well-integrated monitoring solution on Ubuntu: Zenoss, Zabbix or Cacti(+nagios)? [20:05] now i just need to found the libcc file cause it says "libgcc_s.so.1 must be installed for pthread_cancel to work" [20:05] http://ubuntu.shapado.com/questions/what-s-the-most-well-integrated-monitoring-solution-on-ubuntu-zenoss-zabbix-or-cacti-+nagios [20:06] how to find a file? O.o [20:07] apt-file search [20:07] it's libgcc1 [20:07] that should be installed by default :) [20:08] it is, but maybe under chroot it can't find it [20:08] or... more likly, it doesn't work with fcgi [20:08] it's disabled [20:09] now it's normal apache [20:09] mpm_worker [20:09] hm [20:09] try prefork [20:10] too slow for me, i'll try to leave it as is [20:10] now is time for suexec :) [20:10] mattt: fair enough. So I believe it's safe to say that swap should never exceed 2GB. [20:11] matt_keys is gone, nevermind. [20:15] alex88: so [20:15] alex88: if chroot is now supported by apache itself [20:15] alex88: you don't need to fix pid file [20:15] alex88: remove the link [20:19] hallyn: how long does it take to build the eucalyptus liveiso/ [20:19] ? [20:21] ivoks: wait, i've added /var/www/alexnetwork/www/ [20:22] DocumentRoot [/alexnetwork/www] does not exist [20:22] i've serious doubts about it, but apache will give errors if chrootdir option in config is rejected [20:22] don't know what to think [20:24] and now says "SuexecUserGroup directive requires SUEXEC wrapper." [20:24] that's worker, right? [20:25] ivoks: true [20:26] i think this is broken [20:27] okok, now it works..but it still says that my docroot isnt' there [20:28] now alexnetwork.it says forbidden [20:28] apache is running? [20:29] [Wed Jul 28 21:28:53 2010] [notice] child pid 23581 exit signal Aborted (6) [20:29] libgcc_s.so.1 must be installed for pthread_cancel to work [20:29] this is bad [20:29] yes it's running [20:30] mathiaz: ping nssov seems to be broken with a newer gcc [20:31] ivoks: mod_fcgid: call /alexnetwork/www/index.php with wrapper /var/www/alexnetwork/php-cgi [20:31] suexec policy violation: see suexec log for more details [20:33] Exec format error: exec failed (php-cgi) [20:33] i got it working with worker [20:34] plain, default apache install [20:34] and just this in /etc/apache2/conf.d/chroot: [20:34] LoadFile /lib/libgcc_s.so.1 [20:34] ChrootDir /var/www [20:34] edited /etc/apache2/sites-available/default [20:35] so that DocumentRoot points to / [20:35] that's it. [20:35] sure? let me try to add LoadFile /lib/libgcc_s.so.1 [20:35] i'm sure :) [20:36] kirkland: the eucalytpus liveiso took prolly 20-30 mins [20:36] i couldn't do it in tmpfs (at 4G it called itself full) [20:36] ok working... [20:36] which slowed it down :) [20:36] now, last thing i hope... [20:37] /bin/sh: Can't open php-cgi [20:37] the php-cgi is in the fcgid wrapper [20:38] that's something else now [20:38] you have working chrooted apache [20:38] you'll have a hard time making fcgi working with chrooted apache [20:38] maybe it's because /usr/bin/php-cgi is unaccessible inside the chroot [20:39] cause apache conects on request [20:39] exactlly [20:39] hallyn: what's the URL to that again? [20:39] but... that's it from me for today [20:39] good night [20:39] hallyn: i think smoser might want to take a look at it [20:40] ivoks: good night, thanks for your time [20:40] kirkland: people.canonical.com/~serge/binary.iso.lzma [20:40] smoser: ^ [20:40] what is this [20:40] really the lzma was worthless here :) shaved 8M off [20:40] smoser: maverick live ISO with euca-everything installed [20:40] hallyn: yeah, i meant to tell you that [20:41] well, id ont particularly want that. iw anted an insall to work :) [20:41] smoser: then stop breaking them [20:41] duh [20:41] smoser: will probably need daviey to help sort that out [20:41] How do I add a static route to something to an interface where vlan trunking is enabled [20:41] bloody wonderful. [20:42] is that a subtle dig at our brittish companion? [20:42] kirkland: have you run into any real problems yet? [20:43] hallyn: well, it's not working yet, but i don't think it's due to image creation [20:43] hallyn: it's probably just euca-2.0 borkage [20:43] hallyn: but i need Daviey to confirm or deny that [20:43] hallyn: as far as I'm concerned, it's looking okay; can you drop the live-helper command you used into the blueprint? [20:44] kirkland: yup [20:44] hallyn: thanks, i'm going to leave it for now; i've verified all the bits are there, and the cloud controller is listening [20:44] hallyn: its friends eucalyptus-* are not yet registered and working [20:45] hallyn: i'll drop an email to you and dave [20:45] kthx [20:46] hallyn: would you mind lzma -d'ing that image, and renaming it to: maverick-uec-amd64.iso [20:46] hallyn: just post me the updated url when you're done [20:51] hello im having problems with CUPS Server, i cant auth and print, can someone help pme [20:51] ? [20:53] I changed the network card in my server and the new card isn't eth0, how do I fix it? [20:53] kirkland: http://people.canonical.com/~serge/maverick-uec-amd64.iso [20:53] hallyn: thanks! [20:57] how would you add a static route with virtual interfaces [20:57] Jinxed-: well, you can 'route add -dev' [20:58] 'route add -net 10.0.2.0/24 -dev veth3' might work [20:58] i have tried this [20:59] http://pastebin.ca/1910363 [20:59] i first added the up route add [20:59] that didn't work [20:59] then i tried making every virtual interfaces default gateway [21:00] what i wanted the static route to be [21:00] but that didn't work [21:01] zul: you may wanna check with upstream then [21:02] zul: file an ITS - they should be able to help out [21:07] does anyone use vmbuilder here? I create two instances then reboot the whole server and the first instance won't start. [21:08] so in theory i can install this iso ? [21:09] hallyn, kirkland [21:09] smoser: i suppose - i don't know if there is some 'installer' package i should ahve installed to do that [21:10] it's intended just as a live demo aiui [21:11] smoser: you should be able to just boot an ec2 node off of it though [21:11] presumably... [21:17] anyone with eucalyptus installed? If so, please confirm bug 610987 -- I will then set it to High [21:17] Launchpad bug 610987 in eucalyptus "euca-describe-availability-zones verbose incorrect output" [Undecided,New] https://launchpad.net/bugs/610987 [21:18] does eucalyptus do kvm failover? [21:18] dunno [21:18] no [21:18] failover: haha not directed to you [21:19] :) [21:23] anyone else running KVM? [21:31] papertigers I am [21:32] thesheff17: whats your setup like? [21:33] papertigers it isn't that advanced....I run standalone KVM machines [21:33] and slowely moving to eucalyptus [21:33] I want to set up eucalyptus too, I never used it before [21:33] it is pretty straight forward with ubuntu 10.04 [21:34] thesheff17: I would love to set it up [21:34] thesheff17: currently I have a quadcore box with 8g of ram running 3-5 vms [21:34] papertigers: I have the similar setup [21:35] papertigers: how do you install your virtual machines? virt-install? [21:36] papertigers: what I really want to be able to do is put a SAN behind everything and run eucalyptus and use vmbuilder [21:43] <_Techie_> i need help setting up postfix, port 25 is blocked by my ISP by default... i will request to have it unblocked at a later date... i just wish to be able to recieve mail at the moment [21:44] <_Techie_> i also have webmin as i am unable to get ssh out from my current location [21:49] Techie it sounds like most of the outgoing ports are blocked. [21:49] thesheff17: I usually install via virt-manager [21:50] papertigers: I have been using vmbuilder which can build virtual machines very quickly but I'm having some problems. [21:50] thesheff17: currently I have a 5.4TB raid6 shared over nfs that is trunk linked via gigabit to the kvm server, and I store all of their drives on there [21:51] papertigers: so are the mounted NFS actually running the virtual machines? [21:52] yeah basically, they all point to something like /mnt/vmimages/dnsserver.img [21:52] papertigers: have you tried virt-migrate? [21:54] yeah virsh migrate --live? [21:54] that one [21:54] <_Techie_> also, how can i redirect a virtualdomain in apache to another ip in my network, eg linksys.domain.com to get my main router and dynalink.domain.com to get my wireless repeater? [21:55] papertigers: does it work good? I have never got it to work...I think it is because I never had one central storage device. [21:55] thesheff17: yeah it works, you need a central storage like I have setup [21:55] kirkland: btw, did i ever mention that the ttylinux iso is under /srv in that liveimage? [21:55] (or, should be) [22:04] hallyn: cool, i didn't get that orginally, but thanks [22:17] kirkland: all right, lp:~serge-hallyn/live-helper/uec [22:26] _Techie_: you can put redirect directives inside the directive [22:27] <_Techie_> lowridah: can you please provide an example? [22:27] RedirectPermanent / routerIP (3 lines though) [22:31] <_Techie_> lowridah: and last of all, what config file is this in? [22:48] smoser: hey [22:48] smoser: reading your blog post about ssh keys on EC2 [22:48] smoser: is there a way to automate all of that? [22:48] smoser: could this part of the xc2-init project (whatever the name is)? [22:52] <_Techie_> lowridah: i couldnt seem to get it working, i added the lines into the default sites config, and then when i loaded it in webmin it reported an invalind address and wouldnt load the virtualserver [22:53] well then create a new blank site in apache [22:53] make sure you have mod_rewrite enabled [22:53] then create an .htaccess in each blank site you need to redirect to your router/etc [22:55] that should have been in your sites-available sites [22:55] the previous example [23:19] zul: the # of failed mysql install/upgrade bug reports is kind of impressive. Do you happen to know if they are always due to pre-existing configuration file being misplaced, or if there is a real bug? [23:19] (seems like every week i see at least one or two) === XyliosPC1 is now known as Xylios [23:47] hallyn: the main problem with mysql upgrade is that the init script uses a specific user (debian-sys-maintainer IIRC) to connect to the database and shut it down [23:48] hallyn: however if you take a snapshot of another mysql server and load it in your machine, either the debian-sys-maint doesn't exist [23:48] mathiaz: hm. the two bugs i was just seeing today were an apparently-known bug about old_passwords.cnf not existing [23:48] hallyn: or its password is different [23:48] hallyn: which leads to upgrade failure [23:48] all just seems so fragile :) [23:49] hallyn: yes - the mysql package is fragile on upgrade [23:49] hallyn: which leads to tons of bugs :/ [23:49] and there's nothing really to be done about it? [23:49] without fixing zul and a team of ninjas on it of course [23:49] hallyn: hm - not really [23:50] hallyn: https://wiki.ubuntu.com/DebuggingMySQL [23:50] hallyn: ^^ has some help in debugging things [23:50] thanks, i'll look at that [23:50] hallyn: mainly the standard reply that will ask for more information [23:50] (i imagine i'll be seeing more :) [23:50] hallyn: part of the reporter won't provide the information, so we'll close the bug in a few weeks [23:51] hallyn: and the rest will provide the daemon.log file that should have the actual error message [23:51] mathiaz: yeah, i've seen several bugs about non-standard install in the past weeks. not today though [23:51] hallyn: so in terms of first-list triager the standard reply is usually enough [23:51] anyway, figuring those out isn't particularly hard, i was just hoping we could do more than say 'yup, confirmed' [23:51] hallyn: starting from lucid there should be apport hook to provide that information though [23:52] hallyn: well - there can be multiple causes for an upgrade failure [23:52] heh, then maybe we can have a hook to not report a bug, but just auto-fix it :) [23:52] hallyn: :) [23:52] hallyn: usually the error message is about mysql not being able to be stopped or started at some point during the package upgrade [23:53] hallyn: however there can be multiple reasons for this failure [23:53] hallyn: so marking 'yup, confirmed' is not enough if there is only the dpkg terminal log [23:53] hallyn: looking at daemon.log is necessary to figure out the *actual* error message [23:55] mathiaz: right, the ones i saw in the past have *mostly* been bc of apparmor as i recall