[00:39] keescook, hi [00:40] 2 of my ubuntu servers with completely different hardware have network speed problems *slow* === jjesse_ is now known as jjesse === `6og is now known as Kamping_Kaiser === \sh is now known as \sh_away [02:48] can ipv6 enabled nics have an effect on samba within a network ? [04:30] amavisd-new MIR (and the 8 dependencies) are finally done! [04:42] ScottK: party! [08:07] hey there everybody [08:08] I gotto question: in mysql replication, how can I ignore replication of a table in the replication db? [08:10] replicate-ignore-table=db.table3 statement in the slave mysql my.cnf doesn't help [08:14] moin === ewook_ is now known as ewook [09:21] can someone look at http://wiki.ubuntu.com/CalendarServer and add a way for it to start and stop with the system? [09:24] cron it? [09:25] or, simply just add it to init.d [09:35] XiXaQ: init script? [09:35] yes. [09:35] I have no idea how to do that. [09:36] XiXaQ: You can base it off of /etc/init.d/skeleton [09:40] should it be sufficient to change the values between PATH and SCRIPTNAME, or must I do more? [09:42] I can't read sh-script very well. [10:21] XiXaQ: That depends on what you're trying to do.. I haven't read the wiki page. [10:21] would you mind doing so? I think it'd be useful for many people. [10:38] hello, I'm having some trouble with software raid...the raid works and all, no failed disks or anything, but at every reboot, it says the raid is not clean, and then gives me "could not bd_claim sde1" [10:38] and starts rebuilding the array [10:39] I have a 6-disk raid 5, and my dmesg output is here: http://qman.strangled.net:8080/dmesg.txt [10:42] are you sure there are no failed disks? [10:42] it looks to me that it's doing syncing [10:42] check cat /proc/mdstat [10:43] it says it's about 40% through a resync [10:43] but the thing is, all the disks are new, and I just installed the system and created it [10:43] and it does this on every reboot [10:43] well, syncing does need time [10:43] sometimes more than 24 hours [10:44] hmm [10:44] attach output of 'cat /proc/mdstat' somewhere [10:44] ok [10:46] http://paste.ubuntu-nl.org/52122/ [10:47] well, wait for 3 hours and it will finish [10:47] I suppose I probably didn't let it finish building, thanks for the help [10:47] <_ruben> qman: have you ever actually seen the raid finish building? [10:47] * _ruben guesses not [10:47] <_ruben> indeed [10:48] I just assumed that since the md process was no longer in top that it was done [10:48] but obviously not [10:48] you can use your disk [10:48] but it still isn't synced on all disks [10:48] yeah, it kept data after the reboot even [10:48] it's mounted at /home [10:48] so not much on it yet [10:56] soren: i need your opinion [10:56] Chocolate is good. [10:56] soren: we agreed on doing MIR for drbd, but it requires kernel module [10:57] there is a upstream progress on integrating drbd into vanilla kernel [10:57] and, reading lklm, there are some issues which need to be resolved before merging [10:58] so, i'm not quite sure it's a good idea to merge it into hardy, since there is a big chance code will have partial rewrites [10:58] what do you think? [11:03] Well, if it's being actively rejected by upstream, I'm not very keen on it. [11:06] i can't say actively... [11:06] http://lkml.org/lkml/2007/7/21/255 [11:06] but it's quite long discussion :) [11:07] I'm on my way to (a long) lunch. We can talk later? [11:08] enjoy your meal ;) === \sh_away is now known as \sh [12:47] how come the stupid NIS slave server decides to serve it's own /etc/hosts?? === \sh is now known as \sh_away [16:14] mathiaz_, soren: I just migrated a gutsy server to using raid1 on /, and discovered that the server kernel's initramfs does not have raid1 included. [16:14] I had to add to /etc/initramfs-tools/modules 'md' and 'raid1' [16:14] is there a reason this isn't included? [16:18] jdstrand: no reason... You should file a bug. [16:18] mathiaz_: will do [16:26] hi all. does someone know of a online, web based addressbook or contact management software, preferably also with a calendar. one with a good quality? [16:42] hey guys, I notice ubuntu does not include apt-spy in gibbon and netselect-apt doesn't seem to work for me (complains about my firewall when I disable akk iptables rules and host is not behind router) [16:42] is there another package I should be aware of to help me find the fastest mirrors? [16:43] or... even better, I just want to simply test a host on a 100Mbps net connection, can anyone tell me anywhere where I can download a file and see close to that speed [16:46] <_ruben> hmm .. ages ago i read about a method to install debian using a tool similar to (or possibly even using that) debootstrap .. cant seem to find it, nor any docs on an easy 'migration' from one linux distro to ubuntu [16:47] <_ruben> ah .. found the debian docs .. lets see if i can make this work for ubuntu [16:47] _ruben: did you try apt-cache search debootstrap because I found it pretty quickly on 7.10 server [16:48] and then of coarse you just run apt-get install debootstrap [16:48] <_ruben> jetole: well .. the goal is to migrate one of my fileservers from opensuse to ubuntu [16:48] might want to comment out the CD in /etc/apt/sources.list [16:49] _ruben: well good luck with that buddy but all IO would expect is to move the files and then see what software configuration files are still compatible, most should be [16:49] but don't expect it to run out of the box [16:49] you are trying to upgrade from one OS to another in a system that was never designed to do so [16:50] Particularly from an RPM based distro to a Debian based distro. [16:50] <_ruben> the idea is to end up with a clean ubuntu install .. without me getting out the screwdrivers and all to hook up a cdrom drive ;) [16:50] ubuntu uses all different package names and neither OS is aware of the other ones packages [16:50] <_ruben> so 'upgrade' or 'migrate' is a bit of an overstatement [16:50] was SuSE rpm? [16:50] SuSE uses rpm. [16:51] _ruben: if you want a clean install that would be a little more likely and easier [16:51] ScottK: huh, been a decade since I used it [16:51] didn't like it then and don't want to try it now *cough*microsoft*cough* [16:51] There are scripts for booting off of a USB thumb drive out there. You could perhaps boot Ubuntu off one of those and install from there. [16:52] * ScottK used opensuse, but got tired of being thought of as an SLES beta tester who had no right to expect stuff to actually work. [16:52] <_ruben> ScottK: hmm .. hadnt thought about that one [16:52] "linux violates our patents, we won't tell you which ones but if you use our linux, I mean our partners linux, then we won't sue you for the patents that we won't admit exist" [16:52] <_ruben> we actually are using SLES9 at work currently .. working on migrating to ubuntu tho ;) [16:53] :-) [16:53] _ruben: you can install a clean system from debootstrap, I have done it but it has been a few years since so I don't remember how [16:53] look up a readme on it [16:53] a clean system from deboo strap can usually be done, although you may require a temp partition [16:53] or you can just install the CD Rom, that takes about 5 minutes [16:54] <_ruben> got a nice start here: http://www.underhanded.org/papers/debian-conversion/remotedeb.html .. tho i dont think i have any spare diskspace to repartition :/ [16:54] <_ruben> jetole: the install would go fast .. that is, once i have the cd in cdrom drive in that box .. it currently doesnt have one :) [16:55] _ruben: right but installing a CD ROM drive has always seemed easy to me [16:56] then again, I am the only person I know who was building computers as an after school job when I was a freshmen in high school [16:57] <_ruben> jetole: hehe .. the technical part aint the problem .. its the physical side of it .. its currently tucked in nicely between some other boxes with a fair share of cables hooked up and stuff like that :) [16:59] <_ruben> huh .. since when is /dev/sda1 etc used for pata disks? just upgraded another fileserver from feisty to gutsy [17:05] _ruben: for a while now [17:06] I had a feisty box with a jmicron sata controller and HDD hooked up to both sata and pata and a sata cdrom, all disks appeared as sdX [17:06] <_ruben> hmm .. hadnt noticed before .. then again, havent been playing with ubuntu much, yet [17:07] _ruben: Since the kernel migrated to the new PATA code that's in the same module as the SATA/SCSI code. [17:08] It's a general kernel change, not Ubuntu specific. [17:08] that sounds about right, I remember reading up on it some time ago [17:08] anyways, hd = sd; [17:09] <_ruben> ScottK: figured as much .. suse being quite behind on the kernels, makes it a new feature for me on ubuntu ;) [17:09] what is the package I need to install to provide server kernel headers? === mathiaz_ is now known as mathiaz [17:09] <_ruben> hmm .. 50 euros for external dvd writer .. not too bad [17:10] <_ruben> jetole: the source is the same for server and desktop [17:10] I never said it wasn't [17:10] I asked what is the package I have to install to provide server kernel headers [17:11] since the two use different kernels and there is a header specific package [17:11] <_ruben> hmm .. actually .. i might be wrong [17:11] <_ruben> indeed .. never looked at it this closely .. the source package is identical from what i gathered .. so i guess it'd be linux-headers-server [17:11] _ruben: I can type apt-get source to install source for any package on either version of the distro [17:12] I am trying to install vmware-server atm and it is looking for /usr/src/linux/include [17:13] now I know that is the default spot for the raw kernel source headers but like I have done this a while ago and there is another package that provides this I beleive [17:13] unless... [17:14] * jetole tries sym linking /usr/include/linux to /usr/src/linux/include [17:14] <_ruben> that usualy doesnt work .. tho that's gathered from the few times i tried to get away easily that way .. havent tried on ubuntu tho [17:16] <_ruben> linux-headers-server it seems to be .. would require a symlink to /usr/src/linux tho .. or tell vmware to use the actual path [17:17] actually neither one has worked so far, I am installing linux-source atm [17:18] back to my initial question, does anyone in here know what tool would work well for finding the fastest mirrir to you? [17:18] <_ruben> that was actually my first thought, since vmware probably requires more than just the headers [17:19] <_ruben> i recall a tool from that from my own early debian days, but have no clue about the name .. for me any dutch (me being dutch) mirror gives me full speed [17:19] _ruben: not in this case but the kernel headers that are provided here are different then the ones it needs [17:19] <_ruben> odd [17:19] jetole: netselect (but haven't used it on ubuntu) [17:20] _ruben: apt-spy worked for me in the past and ended at feisty, netselect-apt seems not to work at all with firewall issues being reported when it is not behind a firewall, iptables, or a router [17:20] _ruben: there are actually two sets of kernel headers but I forget the difference [17:20] one is used for the active kernel and one is used to compile programs against it [17:21] <_ruben> ic [17:29] <_ruben> hmm .. my gutsy machine just kernel panic'ed on sudo halt [17:30] <_ruben> most of the backtrace scrolled away .. power down it is [17:50] Where is IMAP mail stored? I've been trying to reset a users mailbox structure, but I can't find out why. [17:50] how. :) [17:56] XiXaQ: IMAP servers look wherever you tell them to look. IMAP doesn't dictate the structure. [17:57] well.. I deleted the user on both the server and the client, then recreated it. Still, the folder structure is fubar [19:05] lamont: Got a moment to discuss Postfix plans? [19:05] ScottK: if it's convenient, about 2.5 hours from now would work better. [19:06] lamont: Should be fine. [19:06] poke me then pls [19:06] lamont: Will do. [19:07] i working on Bug 183485 , and found it will be nice to have wiki hovto debug kvm/qemu guest. any ideas about it? [19:07] Launchpad bug 183485 in linux-source-2.6.22 "kvm can not boot Ubuntu guest" [Undecided,Confirmed] https://launchpad.net/bugs/183485 [19:11] ubuntu-bugs team think it will be better organice it with ubuntu-server [19:13] dendrobates: any thouts? [19:14] fishor: we do handle kvm [19:15] fishor: will it not install, or not boot after install. [19:15] see Bug 183485 [19:15] Launchpad bug 183485 in linux-source-2.6.22 "kvm can not boot Ubuntu guest" [Undecided,Confirmed] https://launchpad.net/bugs/183485 [19:16] i mean it will be usefoul to have wike: debugging kvm/qemu guest [19:17] dendrobates: wiki.. [19:19] fishor: It's not much different than debugging a real machine. Do you think so? [19:21] olmost... jast explaine: "kvm -serial file:debug.log" and in grub "earlycon=ttyS0 console=ttyS0" and any other usefoul info if guest can't start [19:22] espessially for kvm. kvm on intel will give mach more not complete reports [19:22] I usually don't find that any easier than just using the console that kvm usually pops up. [19:24] fishor: Feel free to start a wiki page though. If you find it useful, there's a good chance someone else will, too. [19:26] ok. i'll try [20:29] Server team meeting is in ~30 minutes, right? [20:29] That's the idea, yes. [20:29] It doesn't appear to be on the schedule for #ubuntu-meeting. [20:30] Nothing conflicts, so I guess we can just be squatters, but thought I'd mention it. [20:33] I think mathiaz has already told the relevant people to update the schedule? [20:44] ScottK, soren: I've sent an email to fridge-devel about the new schedule [20:45] OK. I just wanted to make sure we could have a peaceful meeting in ~15 minutes or so... [20:46] ScottK: Oh, sure. We sorted out any conflicts before we proposed the new time. [20:46] soren: please can ypu check it, https://wiki.ubuntu.com/DebuggingKVM [20:46] Glad to hear it. [20:47] soren: any other suggestions ? [20:47] fishor: About what? [20:48] any other info, we need from user with crashed gust or kvm host [20:49] Than what? I don't know what you have now? [20:52] i initiated wiki about debugging KVM/qemu guest or host [20:53] Where? [20:53] wee tolked about this to for 1 hour .. https://wiki.ubuntu.com/DebuggingKVM [20:54] PING: Question for anyone working for Canonical (ubuntu) on the server team... [20:57] Is the position for the Server Dev still open? [20:58] InSearchOf, I believe so, yes. [20:58] somerville32, Alright, thanks! [20:59] why not send your resume and see? [21:00] somerville32: You don't work for Canonical do you? [21:00] ScottK, unfortunately not [21:01] ScottK, you're welcome to recommend me for employment though :] [21:02] somerville32: He specifically asked the question to someone working for Canonical. That's why I mention it. [21:05] ToyKeeper: #ubuntu-meeting.... [21:39] lamont: Ping [21:39] ackage [21:40] I sent you a patch in Debian BTS today. [21:40] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=345047 [21:40] Debian bug 345047 in postfix "Please add [::1]/128 to the Default mynetworks setting" [Normal,Open] [21:40] I was thinking about how to handle the Postfix 2.4/5 transition in Ubuntu and I have a proposal for you. [21:41] On postfix-users, Weitse said today that he's going to release another update to 2.4 right after he releases 2.5. [21:41] ok [21:41] I was thinking it would be nice to get that last 2.4 bugfix uploaded to Hardy (either directly or via Debian) and then backported to Dapper/Edgy/Feisty/Gutsy and then we upgrade Hardy to 2.5 [21:42] So that the 2.4 backports are as bug fixed as we can get them. [21:43] Also, about the patch, I think Hardy is going to be around long enough that it'd be really good if it assumed IPv6 exists. [21:43] I tested that on my test server and I think there's no performance downside to it. [21:43] What do you think? [21:43] I like that idea [21:44] in the case where v6 doesn't exist (which means you can't install in debian-installer yet...), it already strips v6 - we _should_ make sure that it installs on a machine with no v6 [21:44] and yet assuming that v6 is there when it is is certainly the way to go [21:44] is 8.04 going to be LTS? I had heard that, but I wasn't sure what the schedule was [21:45] Yes. It's planned for LTS [21:45] 8.04 is LTS for ubuntu et al [21:45] ok that's what i was told, thanks [21:45] kubuntu is planning to not do LTS with 8.04 [21:45] * ScottK grumbles. [21:45] kubuntu is jerks [21:46] lamont: My Ubuntu boxen have IPv6 enabled, but no IPv6 connectivity. Having the added values in mynetworks seems to cause no hard in that case. [21:46] mitchp: Please don't make me argue with you. I'm a Kubuntu developer too. [21:46] lol sry [21:46] No problem. [21:47] ScottK: mv ipv6.ko out of lib/modules/$(uname -r) for the acid test [21:47] OK. I'll do that. [21:47] This would be why I have a test server, right? [21:48] and you'll need to reboot... [21:48] Yah. [21:48] Only 15 sets of modules on that machine ... [21:50] Rebooting. [21:55] zul, I did... a week or so back :-) === leonel_ is now known as leonel [21:59] lamont: If there's a difference it's in milliseconds. [21:59] ScottK: it was more a question of 'does it bitch' :- [21:59] ifconfig or ip addr show lists no v6 addrs? [21:59] * ScottK looks [22:00] lamont: ifconfig shows no ipv6 addresses [22:00] woot! [22:01] * ScottK puts the kernel module back now... [22:01] lamont: Anything else before I reboot again? [22:01] I'll get the change tested again today/tomorrow (I'm assuming we want this in 2.5 certainly, what about 2.4)? [22:02] I think it matters most for 2.5. [22:02] Up to you where you drop it in. [22:02] given that we want to backport 2.4 to warty-1, I'm inclined to make the change start with 2.5.... [22:02] From an Ubuntu perspective and future 2.4 work would be very unlikely to see this code in a new install. [22:03] Dapper actually, but yes. [22:03] Agreed. [22:03] * ScottK reboots again. [22:04] lamont: Did you see my core-dev application? [22:04] saw the mail, haven't actually gotten that deep in my mailbox today (or was it yesterday...) [22:04] Please send a nice vote of confidence real soon.... [22:04] It was yesterday. [22:05] will do today. [22:05] Just think, if I'm approved, I won't need to bug you for the source backport for Dapper [22:05] Thanks. [22:05] * lamont needs to run into town and do a couple things, then will be working through the email glut [22:06] See you later. [22:21] lamont: After rebooting, I confirmed that ifconfig does in fact show an IPv6 loopback, so that we for sure had no IPv6 when it said we had no IPv6. === Bambi_BOFH is now known as kgoetz [22:52] !ebox [22:52] ebox is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See the plans for Hardy at https://wiki.ubuntu.com/EboxSpec [22:52] !webmin [22:52] 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. [22:52] :-) [23:45] any issues moving from feisty to gutsy?