[00:04] hey guys, is there a way to find out if the ubuntu version im running supports ipv6 or not? [00:05] smokie, all supported ubuntu versions have ipv6 enabled by default [00:06] I would say all versions, but I'm not sure about really old ones, before dapper [00:06] qman__, im trying to run HE ipv6 tunnel on my ubuntu, but its always giving me destination unreachable when i try to ping6.. from what i read it has something to do with protocol41.. do you know anything about it? [00:06] but dapper and all since do [00:07] sorry, I don't [00:07] I know you can verify that ipv6 is enabled with a simple ifconfig [00:07] your interface should have an ipv6 address [00:08] but I'm not familiar with configuring tunneling [00:08] it is showing up there [00:08] and i can ping6 the ipv6 localy fine [00:08] but cant ping6 anything outside my network [00:09] <_Techie_> smokie, does your ISP support IPv6? [00:09] _Techie_, yes it does === dendrobates is now known as dendro-afk [00:17] i got it working.. user error [00:17] i had the ips switched [00:22] hola [00:22] ayudenme con esto [00:22] estoy convenciendo a unas personas para que se cambien de win a linux [00:22] entonces quiero preparles una esposicion del porque pasarse a linux [00:22] ademas [00:22] quiero mostrarles unos detalles sobre la inseguridad que tiene win [00:22] y es lo que estoy buscando ahora [00:22] me pueden ayudar? [00:23] nadie en espaƱol? [00:23] one last thing, might b unrelated, if i have a shell script to run, is there a way to track what it does? [00:26] fundacion: the official language here is English. You can try #ubuntu-es [00:27] ok thanks [00:33] smokie: -x will give you verbose output, at least for bourne shell derivs. sh -x ./script.sh [00:33] Can anyone explain to me how to install a SSL Cert? I have already requested it and have the cert, just not sure how to import it [00:50] Crankygeek, it's in the server guide, allow me to link you more directly [00:53] Hello, having some trouble with networking.. sudo ifup eth0 -> SIOCSIFADDR: No such device. eth0: ERROR while getting interface flages : No such device. Bind socket to interface: No such device [00:53] Crankygeek, for apache, see the section "HTTPS Configuration" here: https://help.ubuntu.com/10.04/serverguide/C/httpd.html [00:53] Datz, looks like you don't have an eth0 [00:54] did you recently replace a network card? [00:54] qman__: well, using vmware, other VM's work. I just moved this VM after an install. [00:54] should have told ya, sorry [00:54] it probably just changed MACs [00:54] you have to clear the udev persistent rules for the new one to get eth0 [00:54] humm, I see [00:55] although, I'm not sure how to go about it. [00:55] /etc/udev/rules.d/70-persistent-net [00:55] /etc/udev/rules.d/70-persistent-net.rules contains the information [00:55] ok thanks, taking a look now [00:55] you can delete it or modify it [00:55] then reboot [00:56] so I can delete 70-persistent-net.rules and reboot? [00:56] yes [00:56] it will regenerate [00:56] awesome, trying now. thanks :) [00:56] with the first attached NIC getting eth0, etc [00:57] ah, great :) [00:58] haha.. working. I didn't think it would be that easy. thanks a bunch qman__! === Wandrewvious is now known as WALoeIII === _Techie_ is now known as _TechAway_ [01:26] hello [01:27] i cant found corect code for install ubuntu desktop === _TechAway_ is now known as _Techie_ === dendro-afk is now known as dendrobates === _Techie_ is now known as _TechAway_ [03:29] hello everyone, i just setup a samba server for file shareing a specific directory, and i am able to access on one machine, but as for wireless access im not able to access the samba shared directory " Windows cannot access \\media-server\movie library" How can i trouble shoot this ? [03:33] Four2zero, Check the settings in your smb.conf to see if the IP address of the wireless network is allowed access mate. [03:34] YankDownUnder what am i looking for in smb.conf ? [03:35] interfaces = XXXXXXXXX blah blah blah [03:35] Four2zero, It's directly under the ## NETWORKING ## header [03:35] YankDownUnder i will check [03:38] YankDownUnder the comment is still enabled, " interfaces = 127.0.0.0/8 eth0 [03:38] should i uncomment that section [03:39] Four2zero, On my server here, and on all the servers I setup, I just comment them completely out so that there is access on every interface. [03:39] so leave as default with ; [03:40] so already comment them out [03:40] Yes. [03:40] still same issue windows cant access [03:40] Ditto with "bind interfaces only = yes" [03:40] after i did reloads samba [03:40] Um...if you make a change, you have to 1.) Restart samba, 2.) Reboot the Windows workstation. [03:41] Is the security set to user or share? [03:41] okay, yeah, i restarted the user [03:41] security set for user [03:42] bind interfaces is commented out as well. [03:42] rebooting the machine. [03:42] Four2zero, Here is mine: http://pastebin.ca/1941056 [03:46] New bug: #638585 in bacula (main) "package bacula-director-mysql 5.0.1-1ubuntu1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,Confirmed] https://launchpad.net/bugs/638585 [03:55] Daviey: for when you wake up: (1) obviously, power is back ;-); (2) so far... never been better, ~1,000 instances ran, no failures; (3) heading to bed [03:56] Daviey: oh, yes, still ~1,000 instances to go. [03:58] YankDownUnder, thank you for your smb.conf but still did not fix the issue. [03:59] is there a way to prevent VMWare server from grinding the host OS to a halt? [03:59] i used exact same settings you provided. === _TechAway_ is now known as _Techie_ [04:00] could be a permission issue ? [04:00] could it be a permission issue ? [04:01] Four2zero, what's the issue? [04:01] Jeeves_ samba shared directory. [04:01] i'm able to access the shared directory on my desktop running windows7 [04:01] Four2zero, and you're having issues with people connecting? [04:02] you need to make sure the users are in the samaba group [04:02] but my other devicese such as netbooks under user accounts cant access the shared directory and they too are windows 7 [04:03] hummm, max # of clients set? [04:03] same user/pass? [04:03] in windows 7 netbook it says: The network path was not found [04:04] can you ping the server? see the share with "net use"? [04:04] from the netbook [04:05] the one that's having issues connecting [04:06] ping from netbook is successful [04:06] sent 4 rec 4 [04:06] 0 lost [04:07] thing is when trying to connect to the netbook does not get no prompt for username or password to be entered [04:08] when trying to connect it gets a pop up, Windows cannot access \\media\movie library [04:13] Four2zero Was out [04:13] np [04:13] Four2zero, Windows what? [04:13] windows still giving me same error [04:14] Um...have you possibly overlooked the obvious? [04:14] windows cannot access \\MEDIA-SERVER\Movie Library [04:14] Rename the share to something simple. "Movie Library" is NOT simple. No capitals, no quotes, nada. [04:15] okay. [04:15] Four2zero, So if you rename the share in your smb.conf to a simpler name, restart samba, that should resolve that issue. [04:16] Four2zero, Just as an example, on a server I admin, I called the movie share simply "video" => also, I have "documents" and "pictures" => get the picture? [04:17] got the picture [04:17] Windows cannot access \\MEDIA-SERVER\videos [04:18] Samba does not like to utilise complex names - that's for all-around compatibility. However, if your clients were something other than Windows, you'd have a bit more success...but that not being the case, you have to live with a kludge. [04:18] Four2zero, Check/edit the share information in your smb.conf, then also DOUBLE CHECK the actual pathname in linux. [04:20] Four2zero, Just to be more clear, if you've got a linux directory called "Movie Library" - before you're sharing it, rename it to possibly Movie-Library => then in your path to the share in the smb.conf, you can do /blahl/blah/Movie-library => restart samba. [04:21] okay [04:33] hahaha [04:33] i figured it out....finally [04:33] what it was [04:33] that was preventing the access [04:34] sudo umount -t ntfs /dev/sdb1 /media/storage -o uid=1000,gid=100,utf8,dmask=027,fmask-137 [04:34] minus the u [04:34] in "umount" [04:35] Right oh. [04:36] i went ahead and umounted the drive as so... without the "-o uid=1000,gid=100,utf8,dmask=027,fmask-137 [04:36] and was able to access with netbook. [04:37] so i would like to understand why i had to edit the line in the fstab, so it can be accessable ? [04:39] In all reality, sharing from a "removable" device is a bit silly, but hey, to each their own. [04:40] I notice with "writeable = yes" set, Anyone can read/write to it....what is the option to set to read only for different users ? [04:41] i dont want all family members to delete any files within that specific directory. [04:41] is this possible [04:43] <_Techie_> Four2zero, yes, permissions [04:44] _Techie_ is what i need to modify ? [04:44] File creation mask is set to 0700 for security reasons. If you want to [04:44] # create files with group=rw permissions, set next parameter to 0775. [04:44] <_Techie_> not quite [04:45] _Techie_ what did you mean by "permissions" ? [04:45] <_Techie_> you want your read only users to be connected via guest, and you want to login with a user that is the owner of the files [04:46] <_Techie_> so you will want the permissions to be 774 [04:46] right. i want the rest of the house hold to only have read-only [04:47] <_Techie_> that way only the owner and the group of the file will be able to write and execute, everyone else will be read only [04:47] okay, so i need to add this to the same area as my [Shared] section [04:47] <_Techie_> yep [04:47] gotcha, will test out right now. [04:48] 774 or 0774 [04:48] <_Techie_> 0774 for good luck [04:51] lol i first tryed 774, no luck, it was 0774....lol [04:51] thanks [05:06] New bug: #638641 in ntp (main) "package ntp 1:4.2.4p8+dfsg-1ubuntu2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/638641 [05:32] New bug: #638565 in php5 (main) "php5 crashed with SIGSEGV in strcasecmp()" [Undecided,New] https://launchpad.net/bugs/638565 [05:36] New bug: #638568 in apache2 (main) "apache2 crashed with SIGSEGV in strcasecmp()" [Undecided,New] https://launchpad.net/bugs/638568 [06:11] New bug: #638663 in elinks (universe) "Startup Error $HOME directory" [Undecided,New] https://launchpad.net/bugs/638663 [06:22] hey guys - I'm looking to do virtualization on a server: what's my best option? [06:22] (ubuntu server) [06:22] (lucid) [06:22] MrPPS: what is the intended purpose of virtualization? [06:22] SpamapS: to run 2 to 3 small virtual root servers for others to use [06:23] MrPPS: do you want to be able to setup/teardown nodes easily, or just want to run 3 or 4 things on one big box isolated from one another. [06:23] SpamapS: being able to setup and tear down the VM's easily would be nice, but not essential: i really just want to isolate the systems [06:23] MrPPS: if you use UEC they'll be able to create and destroy VMs themselves. [06:24] MrPPS: otherwise just setup kvm.. seems to have better support than xen. [06:24] SpamapS: how does UEC compare to Xen or KVM, both in support, and ease of setup? [06:24] MrPPS: UEC uses KVM [06:24] MrPPS: it attaches a frontend to it, and uses the EC2 API, so people can use tools built for managing EC2 [06:25] ah ok - are there any good setup guides for getting KVM up and running, and how does KVM's performance compare to Xen? [06:26] or could you suggest any guides that work for Xen on Lucid? [06:29] SpamapS: well, thanks for the suggestions :) === joschi___ is now known as joschi [06:37] MrPPS: https://help.ubuntu.com/community/KVM :) [06:37] !kvm [06:37] kvm is the preferred virtualization approach in Ubuntu. For more information see https://help.ubuntu.com/community/KVM [06:48] SpamapS: thanks, I'm on that page now - i think i won't have any trouble with that, but my main trouble will come from networking [06:48] i have a dedicated box, which has 1 primary IP, and 3 additional IP's available on the 1 interface [06:48] figuring out how to pass them onto a xen machine as a dedicated IP so they can be reached from the internet may be difficult? [06:48] MrPPS: bridged should work fine with that. [06:49] MrPPS: you'd just take them off the host OS's eth0, and put them onto the eth0's of the guests. [06:49] I think [06:49] I admit, I've never done a kvm setup.. just Xen. [06:49] i would really like Xen, but it doesn't appear to be ubuntu friendly at the moment [06:50] and i dont really want to go to debian or opensuse if i can avoid it [07:18] Configure database for phpmyadmin with dbconfig-common, i said yes to this is not recommned, or can i revert back to say No [07:18] by just reinstalling phpmyadmin ? [07:44] Four2zero: sudo dpkg-reconfigure phpmyadmin [07:45] SpamapS i figured it out thanks [07:45] however can you explain why i get this: http://pastebin.ca/1941153 [07:46] Four2zero: no clue, what is printing all of that stuff out? [07:46] Four2zero: FYI, I do not recommend using phpmyadmin [07:46] phpmyadmin [07:46] Four2zero: its worth your time to learn the mysql command line. [07:47] really. [07:47] so i should remove it [07:51] SpamapS do i want to pur the database for phpmyadmin ? since im uninstalling it ? [07:51] Four2zero: I don't know. I'd say no if you're not 100% sure. [07:53] well purge the database for phpmyadmin? if its, not installed then I should purge it if im correct [07:54] okay i removed phpmyadmin [07:55] SpamapS can you take a look at this link please ? [07:55] can i pm u the link [07:57] that's a no [08:00] anyone run squid? [08:01] whats squid ? [08:01] squid proxy [08:02] !anyone [08:02] A large amount of the first questions asked in this channel start with "Does anyone/anybody..." Why not ask your next question (the real one) and find out? [08:04] well, I am specifily looking for the rough successes rate of the cache function. looking to implement it and see if i could make out a rough "reduction of bandwidth" percentage to write on my proposal [08:05] that depends a whole lot on what you do with the internet, and what settings you use [08:05] it's all customizable [08:05] I use squid to cache updates [08:05] windows, or just ubuntu [08:06] ubuntu, as windows update is proxy unfriendly [08:06] if you're looking to consolidate windows updates specifically, look into WSUS [08:07] or opendns [08:07] opendns is not a proxy [08:07] i know. [08:07] well, the enviroment that this would be deployed in is a student dormitory, so basicly, just looking to cache websites [08:07] HTTtrack [08:08] well, it'll reduce the number of times the facebook logo gets downloaded, but stuff that only gets looked at once, or updates constantly like twitter or news articles, it won't help with [08:08] this would have to be transparent to the users, so squid in a transparent proxy form would be the role [08:09] it will reduce your bandwidth, but depending on usage it could be minor, a few percent [08:09] are we talking like 10-15? [08:09] and to be effective you need a few hundred gigs of proxy data [08:10] if they're heavy youtube users, 1% [08:10] well, I was planing on giveing it a ~4tb storage pool to use [08:11] and a hefty amount of RAM [08:11] good choice [08:11] with that much space you could cache much bigger files [08:11] so it really depends on how many files get requested repeatedly [08:11] there looks to be a kinda beta for youtube cacheing on squid [08:12] if they visit the same sites and watch the same videos and such, it'll be a significant reduction [08:12] they are college students, facebook is like 40% of the hits on the gateway some days [08:12] try to track where they go and parse for repeat requests [08:13] then I think you'd see notable improvement, lower bound 3%, upper bound 20% [08:13] sounds about right [08:14] but yeah, unfortunately, windows update doesn't work with a caching proxy [08:14] wouldent expect it to [08:14] it does some nasty things that break it [08:15] basicly, all http will be ran threw the proxy [08:15] that should not fuss to much with WU right? [08:15] no, it'll still work, it just won't cache the updates [08:15] that was not a goal anyway, so that is not a problem [08:16] you're going to want to mark some sites (such as blackboard) as never cache, or with short cache times like 5 minutes [08:17] but most sites will work fine with a 12-24 hour cache cycle [08:17] good point, putting that in my notes [08:17] SpamapS: the graph is still broken. More subtly, but still :) [08:17] is there any HA guide for squid aswell, or will that have to be implemented in the roughter [08:17] err, router [08:17] I don't know about squid's abilities in that regard [08:17] but I know you could do some routing tricks to have a backup server [08:18] or to just let traffic through normally if the server goes down [08:18] well [08:18] yea, but, the students are kinda, well, not the most well manerd users [08:19] so if it goes down, management would have my... head [08:19] at least until it came back up [08:19] right, you can configure a router to detect if it goes down and just switch over [08:19] and when it comes back, switch back [08:19] that was kinda the plan, as i did not see any HA docs on the squid wiki [08:20] not sure about doing that with ciscos or what have you, but with a linux or BSD router you could [08:20] sonicwall, but yea, it is a function [08:20] script a check to make sure it's up [08:22] you think a dual 3.0 xeon with 4tb raid5 and 8gb ram would do the trick, becase that is what I had spec'd out [08:22] that should be fine for at least a few hundred users [08:23] think it is 300-400 depending on the time of year [08:23] might want a little more RAM if it's slow [08:23] but that's a good choice [08:23] yea, 16? [08:23] yeah [08:24] gigabit backbone? [08:25] fiber to each floor, then 100mb to each room [08:25] yeah, that should work fine [08:26] just want to make sure the link between the router and the proxy server is fast and clear [08:26] New bug: #638708 in tomcat6 (main) "Cannot start as logged in user" [Undecided,New] https://launchpad.net/bugs/638708 [08:26] to work effectively, it's going to need testing and tweaking [08:26] of cache sizes and parameters [08:27] that is what I plan to spend the most time on [08:27] hopefully only about a half day of no internet, then ~4 days of tweeking [08:28] should only take maybe ten minutes of no internet at all [08:28] then a few days tweaking [08:28] well, I always quote more, just in case [08:28] yeah [08:28] plus it makes you look better :) [08:28] and during the testing you'd have lots of momentary downtimes [08:29] of course that's a good time to set up and test the automatic failover [08:29] if it doesn't break the internet when you restart squid, you'll be set there [08:29] testing will be done on a sperate network, then once all is solid on the network end, we will move the servers over [08:30] good show [08:30] my school didn't have the budget for that [08:30] but definitely the right way to do it [08:30] heh, we are a private compnay [08:31] we own a few buildings and all of them are student housing [08:31] ah [08:31] so we have kids that go to a few differnat colleges [08:32] well, done right, it should definitely improve performance and reduce overall bandwidth used [08:32] being done right the key qualifyer [08:34] thanks for your help qman [08:34] no problem, good luck [08:47] New bug: #459048 in php5 (main) "php5 crashed with SIGSEGV in _Unwind_ForcedUnwind() (dup-of: 343870)" [Medium,New] https://launchpad.net/bugs/459048 [08:59] is webmin good to use on ubuntu-server, or should there be another recommended panel [08:59] what panel should i use to administer the my ubunut-server ? [09:10] !webmin | Four2zero [09:10] Four2zero: 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. [09:11] !ebox | Four2zero [09:11] Four2zero: 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 [09:11] thank you qman__ [09:11] of course I wouldn't recommend using a web GUI at all, but that's my preference [09:11] qman__ why is your opinion ? [09:12] why, what is your opinion on that [09:12] more trouble than they're worth, I'd rather just use SSH [09:13] when you use one of them, you're limited to what they offer [09:13] if you need configuration outside of that, they cause trouble rather than make things easier [09:14] that's food for thought there ! [09:14] I will keep that in mind. [09:26] what's the equivalent of the "tunctl" command in ubuntu? [09:27] New bug: #638738 in libcommons-fileupload-java (main) "Enable portlet API in commons-fileupload" [Wishlist,Triaged] https://launchpad.net/bugs/638738 [09:32] JamesPage: nice and quick work on those libraries ! [09:43] MrPPS: tunctl. [09:43] MrPPS: It's in uml-utilities. [09:44] MrPPS: Next time, if you want the equivalent of something rather than the thing itself, it might be helpful to specify what you want to do. [09:47] soren: sorry, i just wanted to find the tunctl util [09:47] to i did an aptitude search [09:47] couldn't find it in there, and figured you used a different utility [09:48] thanks for letting me know [09:48] MrPPS: That's what I'm saying. If we didn't have tunctl, you should specify what you want to do so that we can point you at the right tool. [09:49] soren: ok, thanks for the help [09:51] hi guy's is there a debootstrap script for lucid? [09:53] of course. [09:55] soren:I am running an old edgy box running xensource so I can use apt to retrrieve the lastest debootstraps ... [09:55] donttrustem: Then you're screwed. [09:55] In so many ways, it's not even funny. [09:55] Really ... [09:56] edgy went out of support /years/ ago. [09:56] only 2 years ago [09:56] I don't even want to think how many high-risk kernel vulnerabilities that thing must have. [09:56] OK .... point taken [09:56] I will upgrade the system [09:57] soren: I ran 3 RH 8.0 boxes up until May of this year.. ;) [09:58] that one, old, proprietary, statically linked, PoS app that wouldn't work on any other version.. just refused to die.. ;) [09:58] soren: Can I download the lucid script seperatley [09:58] SpamapS: When did that go out of support? I don't follow RedHat's release cycle at all :) [09:58] donttrustem: You can try. [09:58] donttrustem: It's in the debootstrap package. [09:59] soren: hah.. 2004. ;) [09:59] :( [09:59] soren: I just want to build the domU ... How do I extract the .deb [09:59] Just install it? [09:59] OK... [09:59] Or extract it with dpkg-deb if you insist. [09:59] * SpamapS has just realized that he completely forgot to go to bed [09:59] * SpamapS passes out [09:59] donttrustem: Oh, by the way.. [09:59] I will upgrade the server next week [10:00] donttrustem: We don't support Xen dom0 anymore. [10:00] I am moving to xenserver [10:00] soren: Citrix version [10:00] I'm not sure what that means in this context. [10:00] Do they provide the dom0, then? [10:01] soren:http://www.citrix.com/English/ps2/products/product.asp?contentID=683148 [10:02] soren: the do it a different way [10:02] I'm not in the mood to read marketing material. [10:02] Ever. [10:02] :) [10:02] OK [10:02] :) http://community.citrix.com/ [10:03] Doesn't look much better. [10:03] Do they provide the dom0 is all I'm asking. [10:03] Because Ubuntu doesn't. [10:04] Not that I really care. I'm just saying that Ubuntu doesn't, so if you rely on Ubuntu to provide a dom0, you're screwed. [10:04] dunno need to check but the DC I am using run this on all there ubuntu servers [10:05] so it must work [10:05] Ubuntu Edgy provided a dom0. [10:05] We grew up, got wiser, ditched Xen. [10:06] soren: virtualistion is the way to go ... [10:06] Yes. Xen isn't. [10:06] soren: what do you suggest [10:06] KVM. [10:06] that's not just what soren suggests, that's the thing we support. [10:07] post-edgy. [10:07] A surprising coincidence :) [10:07] yes, there must have been a reason for keeping those old insecure boxes around :) [10:07] donttrustem: However, if your hardware is from the Edgy era, you're probably screwed. [10:08] soren: poweredge 1950 [10:08] ttx: They work? [10:09] donttrustem: I don't speak Dell. [10:09] We're slowly moving our dapper machines up to lucid now. But then we got more time by chosing LTS releases. [10:09] LOL [10:09] soren: Doesn't Ever Last Long - yes. [10:10] maswan: *gasp* [10:11] soren: We're getting better at it, I think our last couple of breezy machines were about a year out of support by the time we finally ditched them. :) [10:11] I actually found a laptop on a shelf recently that still ran Breezy. [10:11] Aw, breezy. Great days. [10:12] crap ...I ran the dpkg and it wiped out my scripts from /usr/lib/debootstrap/scripts [10:12] * Daviey did the same with dapper... and couldn't remember by login details :) [10:12] my* [10:12] where does it install to please [10:12] jpds: depends on your definition of "work" [10:13] fpind them [10:13] jpds: i wouldn't want to trade security vulnerabilities in exchange of "don't touch it if it works" [10:14] ttx: Oh, I thought you meant old == age of the box, not the release. [10:14] All working ... :) [10:15] how can I change the screen resolution on the console in 10.04? [10:16] it's way too small :) [10:16] donttrustem: /usr/share/debootstrap where they've always belonged. [10:16] :( W: Failure trying to run: chroot /mnt/xen dpkg --force-depends --install /var/cache/apt/archives/base-files_5.0.0ubuntu20_amd64.deb /var/cache/apt/archives/base-passwd_3.5.22_amd64.deb [10:16] or - resolution is too large, that is [10:17] soren: any idea on that error? [10:17] no [10:22] does anyone know [10:23] yay, all server $bugs triaged [10:24] donttrustem: what are you trying to achieve? [10:24] twb: debootstrap lucid /mnt/xen http://gb.archive.ubuntu.com/ubuntu/ [10:24] donttrustem: pastebin the entire transcript [10:25] donttrustem: also, if you're doing that ON a lucid host, try cdebootstrap or multistrap [10:25] twb: the debootstrap script [10:26] donttrustem: run script -c "debootstrap lucid /mnt/xen http://gb.archive.ubuntu.com/ubuntu/" [10:26] donttrustem: it will create a file called "typescript". Pastebin the contents of "typescript" [10:28] SpamapS: working on EU tz ? [10:29] twb: http://pastebin.com/3tb8UncB [10:31] donttrustem: are you running this as root? [10:32] yeah [10:32] Daviey: is there any value in me installing a fresh UEC right now ? (i.e. do you have another upload pending ?) [10:32] * ttx needs to reproduce bug 628055 again [10:32] Launchpad bug 628055 in libvirt "Instances don't start correctly: Security Labeling error running aa_change_profile()" [High,Incomplete] https://launchpad.net/bugs/628055 [10:33] donttrustem: what does "grep /mnt /proc/mounts" say? [10:34] "/dev/vg1/cccam.root /mnt/xen xfs rw 0 0" [10:35] OK, run the same script command again, but pass --verbose to bootstrap [10:37] twb: I ran as --verbose but the output was the same [10:38] twb: script -c "debootstrap --verbose lucid /mnt/xen http://gb.archive.ubuntu.com/ubuntu/" [10:41] Does dmesg | tail give any hints? [10:41] Are you running this on a host that's *already* a domU or so, or are you running it on a "real" ubuntu machine? [10:42] I am running it on the Dom0 [10:42] ttx, Hmm [10:42] ttx, Well i do have an uploading pending... but the binary ~= what is in my PPA [10:42] twb: which is the real ubuntu machine [10:42] ttx, the change being fix to metadata service. [10:43] hm. I can install from ISO, upgrade to your PPA [10:43] twb: no entry in dmesg .... [10:43] ttx, So if you tested installation from current archive, then upgraded to my PPA you'd get a good test [10:43] ack [10:43] (nothing in regards to installation procedure has changed from my PPA) [10:44] ttx, Depending how much time you have spare.. testing lucid->maverick upgrade is also something that really needs to get done today aswell (on my hit list, but more eyes the better) [10:45] let me try another distro [10:45] twb: I have done this many times and it works .. but this is the first time with Lucid [10:45] Daviey: ew. maybe not time for that. will try [10:46] donttrustem: right; I don't know why it isn't working [10:46] twb: what was the distro before lucid [10:46] karmic [10:46] hardy [10:46] ttx, I'm expecting success - the last test was successful.. so don't need to panic about it. [10:47] (or seemed to be successful... but i guess today i want to add confidence testing) [10:48] hmm! same problem [10:49] maybe I can run that part of the script manually [10:51] twb: chroot /mnt/xen [10:51] groups: cannot find name for group ID 0 [10:51] I have no name!@xen:/# [10:56] twb: http://pastebin.com/fwjpiW3L [10:56] That's because /etc/passwd (provided by base-files) isn't unpacked [10:56] debootstrap *should* be unpacking a bunch of core packages before it does the first chroot [10:57] twb: is there a problem with the scripts [10:57] I don't know what is up with those EBADFD's [10:57] Try cdebootstrap [10:57] OK [10:57] It's faster anyway [10:58] twb: I don't have that installed [10:59] So install it [11:03] Man I going to get the server rebuilt .... [11:04] twb: can't install the package dependencies [11:39] an we block all traffic coming to my server on port 5060 udp and we can only allow some specific IP to come and register on 5060 udp port?? through IPtables [11:39] ? [11:42] I believe rule work off the first match, so as long as the address-specific rule comes first, should work as described [11:44] please can you guide me.. how to make iptables rules [11:46] oryxtec > http://www.google.dk/search?hl=&q=iptables+rule+generator [11:50] !ufw > oryxtec [11:50] oryxtec, please see my private message [11:51] oryxtec: oh, sorry, 5060 is SIP. SIP doesn't just use one port, it's tricky [11:51] oryxtec: maybe have a chat with #shorewall or #netfilter [12:05] twb: acutally there are some IP addreses [12:06] who are hitting my asterisk server .. trying to regiester with my server [12:06] i want to block those ip address [12:06] how can i do that? [12:07] oryxtec: man hosts.deny? [12:07] wht is that? [12:07] :) [12:08] oryxtec: sorry, I don't care to help you further. Wait patiently; maybe someone else will. [12:09] k. === _Techie_ is now known as _TechAway_ === dendrobates is now known as dendro-afk [13:15] Hey. Im trying to wrap my head around setting up a dev-server at home, behind my router, with several virtual hosts, and I want each virtual server, to have its own subdomain, on my domain. Should I do some DNS server, or reverse proxy? [13:15] virtual hosts = virtual machines [13:17] bobslaede: try dnsmasq [13:17] joschi: I have been looking into that. Will that enable be to map dev1.example.com to one VM, and dev2.example.com to a second one, while i only have one wan-ip [13:18] bobslaede: that's not possible at all [13:18] bobslaede: you'll need to do some portmapping/NAT/PAT [13:18] thats why i havent found any articles on it [13:18] joschi: i have been thinking about doing a reverse proxy with nginx [13:19] bobslaede: you could do something like port 10022 maps to vm1, port 22/tcp; 20022 maps to vm2, port 22/tcp etc. [13:19] bobslaede: if you use only http/https that will work [13:19] bobslaede: but not with much else [13:19] joschi: yeah, but then i would have to have dev1.example.com as example.com:1022 instead right? [13:19] bobslaede: yes [13:20] yeah, thats what im trying to avoid [13:20] I think it can be done with nginx, but then it would still only be for http/https [13:20] bobslaede: what exactly do you want to achieve? [13:21] site1.dev1.example.com, site2.dev1.example.com db1.example.com [13:21] as virtual machines [13:21] with one wan-ip [13:21] having *.example.com mapped to the wan ip [13:22] and routed to the server [13:22] if that makes any sense at all :) [13:25] bobslaede: ok. you can use a proxy for some services, e.g. http. but that only works because these protocols send the hostname in-band (HTTP/1.1 host header for example). [13:25] bobslaede: this won't work with other services, like MySQL or SMTP, so you'll need to use port-mapping for these [13:26] joschi: yeah, that was my worries, and why i was thinking about some dns setup. [13:26] bobslaede: which won't get you anywhere [13:26] joschi: nginx however, is able to translate for smtp, and couchdb at least [13:26] bobslaede: why do you need to make your development boxes externally available anyway? [13:27] joschi: well, mostly for fun, and to enable clients to the progress on websites [13:27] bobslaede: couchdb uses http as its application protocoal [13:27] s/protocoal/protocol/ [13:27] joschi: then, that explains that :) havent used it before, but wanna try it out [13:29] joschi: if clients wants to check out a website in progress, or bug fixing, its easier to go to dev1.example.com than example.com:1022 [13:29] bobslaede: and now tell me again why you need any other service than http available externally... [13:29] joschi: or if im running a 3rd party cms tool or something, that doesnt work well with ports [13:30] joschi: that parth would be the just for fun part of it. Just to try [13:30] and learn === elb0w`away is now known as elb0w [14:02] hggdh, Are you around? === ivoks-afk is now known as ivoks [14:17] Daviey: yes [14:18] hggdh, two mins, on a call. === luis__lopez is now known as luis_lopez [14:28] jdstrand: do you know why libvirt on lucid is failing to build (https://launchpad.net/~ubuntu-server-qa/+archive/regression-test/+build/1960449)? (well, it's bailing bc it finds no macvtap support in kernel headers, but...) [14:29] hallyn: I never looked into it. I brought it up with you and kirkland way back when I uploaded 0.8.3 and encouraged an interested party in fixing it :) [14:30] clearly, I wasn't that interested party [14:30] ;) [14:30] oh [14:30] jdstrand: you did? [14:30] hallyn: lucid [14:30] ah, I assumed maverick/powerpc [14:30] let me actually look at what you asked me to look at [14:34] hallyn: you are building maverick's package on lucid? interesting. uhm, yeah, as mentioned, this is the line: configure: error: Installed linux headers don't show support for macvtap device. [14:34] hallyn: looks like you need --without-macvtap for lucid [14:35] hallyn: look at DEB_CONFIGURE_EXTRA_FLAGS in debian/rules [14:35] Is this behaviour expected with vmbuilder? I created a virtual machine with vmbuilder and specified a destination directory with -d. Something in vmbuilder failed and recursively removed everything in the destination directory [14:36] pkhamre: Can I see the full command line? [14:36] I specified /var/lib/libvirt/images as destination, which also was home to some images for other virtual machines [14:36] pkhamre: I've also had the same problem. Shouldn't be expected behaviour, I think. [14:37] Luckily I was just setting up the server so I only deleted some fresh installations. [14:38] soren: yes, here you go: http://pastebin.com/ptsgzEtG [14:38] pkhamre: Then yes. [14:38] pkhamre: -o means overwrite. [14:39] overwrite == remove old stuff? [14:39] "Force overwrite of destination directory if it already exist." [14:40] it accidentally deleted my private email- and webserver, but fortunately i had offsite backup :) [14:42] Hi, I am trying to install phamm packages. But I get "Setting up phamm (0.5.15-1) ... dpkg: error processing phamm (--configure): ... subprocess installed post-installation script returned error exit status 10" I tried purging and reinstalling multible times without any success. Doing sudo dpkg-reconfigure phamm also gives "/usr/sbin/dpkg-reconfigure: phamm is broken or not fully installed" [14:42] jdstrand: i'll take a gander at fixing that (when i get a chance), thanks [14:43] jdstrand: Secret trick when uploading packages that are managed in bzr /and/ use a patch management system: Make sure you apply the patch and commit that to bzr (along with whatever extra files quilt puts in .pc) [14:43] jdstrand: Otherwise we end up with: [14:43] jdstrand: and, while i'm yappin' about libvirt, have you looked at any of the virt-manager segfaults (segfaulting in glib/libc)? [14:43] jdstrand: Uh... Well, try going into your libvirt bzr checkout and do an update or a pull, and you'll see. [14:44] hallyn: I haven't but mdeslaur has. iirc he said it was python that was the problem [14:44] soren: are you saying if i do a bzr co of lp:ubuntu/lucid/libvirt, i won't get the latest? [14:44] soren: I didn't commit anything to .pc or with quilt. my changes where all within debian/ [14:44] hallyn: No, that's not what I'm saying. [14:44] jdstrand: ok, so i'll leave those bugs alone. lately i can't reproduce anyway [14:45] jdstrand: Yes. [14:45] jdstrand: That's the problem :) [14:45] jdstrand: ...because when the importer thing comes along, it /does/ apply things and imports the package that way. [14:46] jdstrand: I made the exact same "mistake" with the previous upload. [14:46] hallyn, jdstrand: I uploaded a new gtk-vnc that should fix most of those virt-manager segfaults [14:46] jdstrand: I can fix it up so that you can just pull/update, but I just wanted to let you know. [14:46] mdeslaur: cool, that might explain why i can't reproduce [14:46] * jdstrand is a bit confused as the +N files in .pc were the 9025 files that soren added [14:47] hallyn, jdstrand: There is another fix in upstream git that may fix some more...but my virt-manager doesn't segfault anymore [14:47] mdeslaur: dude, that is awesome! :) [14:47] hallyn, jdstrand: if you guys see anymore segfault, let me know [14:47] mdeslaur: thx [14:47] jdstrand: I might now have taken appropriate care myself. I may have been assuming I'd be uploading it myself so I'd get to check up on this before uploading. [14:48] soren: I guess what you are saying is I need to be sure to 'quilt push -a' before committing [14:48] s/now/not/ === ivoks is now known as ivoks-afk [14:48] jdstrand: Yes, /and/ add the resulting files in .pc//. [14:48] hggdh, Sorry.. been on a call. Do you have plans today? [14:48] right [14:48] soren: ok. thanks for that, I've added it to my notes [14:48] soren: sorry I missed it [14:48] jdstrand: the package importer automatically moves our branch out of the way, puts the new one as lp:ubuntu/libvirt and proposes a merge. [14:49] jdstrand: I'm doing the merge now. [14:49] jdstrand: Hey, don't worry about it. It's hardly obvious. At all. It's really quite annoying. [14:49] shame how crappy quilt and bzr work together :( [14:49] makes merge proposals really ugly :( [14:49] anyone familiar with clonezilla? [14:50] soren: so did you say you are going to fix this up? [14:50] jdstrand: Yes. [14:50] soren: cool, thanks :) [14:50] Daviey: yes -- keep on euca tests [14:51] Daviey: I am looking at the long run I letf during the night: after about 1300 instances things went south === xfaf is now known as zul [14:51] hggdh, That is GREAT news [14:51] soren: jdstrand: that flies against how I'd understood it, so to be sure i grok it: you want the bzr working tree to have all quilt patches applied when checked into bzr? [14:51] hggdh, Do you have any capacity for me to throw stuff at you? :) [14:52] hallyn: Yup. [14:52] Daviey: yes, I am all euca now. [14:52] \o/ [14:53] Daviey: I am still to find out WTF happened after 1300 instances. Right now any euca-* command does not return [14:53] soren: jsut to help me remember which way is up - what's the good of that? [14:53] hallyn: Otherwise you end up with conflicts when the package importer does its thing. [14:53] hggdh, OK.. What would be *really* awesome if you could install lucid i386, run a couple of instances just to make the system dirty. Then upgrade to maverick.. keeping an eye out for bug 628055 [14:53] Launchpad bug 628055 in libvirt "Instances don't start correctly: Security Labeling error running aa_change_profile()" [High,Incomplete] https://launchpad.net/bugs/628055 [14:54] hggdh, Is the viable? [14:54] That way, i386 gets tested, upgrade and attempt to reproduce the bug :) [14:55] Would that be ok, or are you swamped already? [14:55] Daviey: yes, it is, I can take a few to do it. But we will have to get back to Maverick and Euca stuck [14:55] soren: what is the package importer? (is it the bzr equiv of 'uscan', or a lp tool, or something else?) [14:56] hallyn: It's the thing that takes package uploads and puts them into bzr branches. [14:57] soren: is that done automatically by lp? or is there a bzr command? [14:59] hallyn: You mentioned lp:ubuntu/lucid/libvirt yourself. [14:59] hallyn: That's created based on uploads. Automatically. by the package importer. [14:59] soren: i thought a person did that [14:59] hallyn: https://wiki.ubuntu.com/DistributedDevelopment [14:59] soren: cool, thanks, i'll look at that [15:01] hallyn: yw [15:02] soren / hallyn: Which causes headaches if you work on unreleased bzr branches as a team effort, and someone branches from the ~ubuntu-branches [15:02] hggdh, How many beers do i owe you now? [15:03] Daviey: Yeah, upstream bzr branches and package branches is a very unhappy love story. [15:03] Daviey: actually, none, but I will be glad to drink one with you on UDS :-) [15:03] \o/ === IdleOne is now known as IdleOp [15:04] Daviey: probably enough that hggdh doesnt need to think about an alcohol budget at UDS ;) [15:05] LOL [15:11] jdstrand: stupid question - i assume i should open a bug about the failure to build (if only to have something to point the fixed bzr tree at)? [15:11] (or would that be a faux pas?) [15:12] hallyn: I don't think it is required. just say something like 'fix FTBFS when building on lucid' or similar. if you want a tracking bug, feel free to file and assign to you [15:13] Daviey: ping [15:14] jdstrand: Fixed. [15:14] jdstrand: Do you expect to work more on libvirt before final freeze? [15:15] Daviey: I did install UEC, was fine. Then I upgraded to your packages, and then the CC doesn't start due to cluster cert mismatch [15:15] Daviey: so it looks like registration is still not idempotent [15:16] ttx, 2 mins - on call :( [15:16] Daviey: right -- would like your thoughts before opening a bug [15:17] jdstrand: ok, cool, if i can get away without filing a bug all the better, thanks [15:19] soren: no. the bug that seems most annoying is bug #628055, but contrary to the message, it doesn't seem related to apparmor [15:19] Launchpad bug 628055 in libvirt "Instances don't start correctly: Security Labeling error running aa_change_profile()" [High,Incomplete] https://launchpad.net/bugs/628055 [15:19] oddly, it is still assigned to me [15:19] (once I have all the requested info, I can look at it more) [15:20] is there a support channelf or Arch linux? [15:20] hallyn: is your change for maverick? [15:20] oh wait wrong channel, sorry [15:21] jdstrand: Cool. I'll do an upload this evening and that'll be it. [15:25] jdstrand: no, for lucid [15:25] hallyn: I'm confused. what do you plan to do with the maverick libvirt on lucid in the archive? [15:28] ttx, This is odd :/ [15:28] ttx, Can i see your registration.log? [15:28] sure [15:28] can't make the NC work either [15:29] jdstrand: what are you talking about? [15:29] 'the maverick libvirt on lucid' ? [15:29] Daviey: note that I re-rebooted and it didn't break my certs again [15:30] jdstrand: ok, seems i've confused myself [15:30] ttx, ... [15:30] hallyn: i386 build of libvirt 0.8.3-1ubuntu10+regressiontest20100915 in ubuntu lucid RELEASE [15:31] jdstrand: i was thinking that was the official lucid libvirt repo, but i see it's not [15:31] jdstrand: so nm all that [15:31] hallyn: ok, heh [15:31] back to real work [15:31] hallyn: I don't know what ubuntu-server-qa is doing, but you can at least pass them the --without-macvtap thingie [15:32] Daviey: http://pastebin.ubuntu.com/494217/ [15:32] the usual double stuff [15:32] + line 68 [15:32] jdstrand: right, but do i have upload rights to that ppa even? [15:33] Daviey: for some reason auca_conf --list-cluster didn't answer, so we reregistered, and it broke the cert [15:33] hallyn: I just meant tell whoever is responsible for that build about why it is failing and that they can use --without-macvtap to fix it [15:33] (on line 68) [15:33] hallyn: not that you should fix it [15:34] ttx, Hmm.. think the sleep needs increasing ? [15:34] jdstrand: oh, ok. it's be a trivial enough fix if i did ahve the perms :) but ok [15:34] Daviey: note the absence of the "cluster is already registered" message [15:34] hallyn: I'll leave that up to you :) [15:34] hm, strike that [15:35] Daviey: well, they should rather make registration really idempotent [15:35] Daviey: like it used to be [15:35] ttx, they have.. their QA proved it as such. [15:35] or fix euca_conf so that it returns correctly [15:36] Daviey: oh right, we must be getting something wrong. [15:38] Hey everyone! [15:38] I have a question about pure-ftpd [15:38] Is someone available to answer a quick question about it [15:39] soren: hggdh: is ubuntu-server-autotest a robot? [15:39] hallyn: Yes. [15:40] Daviey: I fixed my NC by reregistering the node. My CC certificate fixing must have borked it [15:40] I am trying to figure out how to create a conf file for pureftp and have it set a chown on uploads for the users. Has anyone had a setup like that? [15:40] Daviey: but we are left with a potential break-on-reboot thing [15:40] ttx, But need to find the underlying issue [15:40] Daviey: shoudl I reopen the bug ? [15:41] ttx, I guess so :( [15:41] ttx, somewhat suprised me and carlos haven't seen this [15:41] I think it was never fixed, just euca_conf was working sufficiently well so that we never re-registered [15:41] hi all, I'm having trouble accessing my apache2 server from other computers. I can access it on the localhost. Is their a default to block remote connections? [15:41] ttx, *sigh* [15:41] Daviey: if euca_conf works, you can't see it [15:41] Hi guys. I was installing ubuntu server from an installation cd, but after a restart it falls to (initramfs) saying "Gave up waiting for root device. Common problems: - Boot args (cat /proc/cmdline); - check rootdelay=(did the system wait long enough?)". Any suggestions, please? [15:42] soren: does he use a lp daily source build recipe to do uploads? where do i see the recipes? is there someone else i should be bugging instead of you? [15:42] fabianhoward: you sure it's not blocked by a firewall? [15:42] fabianhoward: I mean, the port 80 [15:42] progre55, is their a firewall on ubuntu-desktop by default? [15:43] fabianhoward: no, but maybe it's somewhere between your desktop and your other computer where you want to access from [15:43] maybe but i can ssh to it... [15:43] progre55, i think its unlikely === look is now known as k3rby [15:44] fabianhoward: try to telnet to port 80 =) [15:44] hallyn: Me or hggdh. I'm working on it right now, actually. [15:45] hallyn: And no, it doesn't use lp's recipe thign. [15:45] nothing [15:45] progre55, maybe their is a firewall by default [15:45] progre55, iptables is showing rules, I'll check [15:45] fabianhoward: yeah, I'm sure it's a firewall [15:46] I am trying to figure out how to create a conf file for pureftp and have it set a chown on uploads for the users. Has anyone had a setup like that? [15:49] soren: hggdh: is there some place i can see the source for ubuntu-server-autotest? and where he runs? [15:49] I created a new user on my server, I want them to be able to use gnu screen.... do I need to assign them to a special group or something? [15:49] also, I am getting this error => there is 1 zombie process....what does that mean? [15:50] hallyn: It runs on one of my servers. I think the code is somewhere on Launchpad. Let me check. [15:50] DigitalDeviant, zombie processes are processes that were created by a 'parent process' (this is called forking) and the parent has stopped. They have no parent... so they are a zombie. [15:50] I actually dont know how to detect which process is a zombie, and stop it [15:51] im assuming you can grep the ps command to find it [15:51] something like: [15:51] ps -ef | grep [ZOMBIE CRITERIA] [15:53] ha, it was screen [15:54] ps -ef| grep -i defunct [15:55] do I need to be an admin user to run screen? [15:55] No. [15:55] Pici, how do i make a user who is username:users be able to use screen [15:56] storrgie: You don't need to do anything special for a user to be able to use screen. [15:56] hallyn, soren: I have a branch on bzr+ssh://bazaar.launchpad.net/~hggdh2/autotest/automated-ubuntu-server-tests/ [15:57] since this is a two liner I'll put it here: [15:57] storrgie@minerva:/home/agd$ screen [15:57] Cannot open your terminal '/dev/pts/6' - please check. [15:57] screen -x [15:57] hallyn, soren: I think the original (from soren) is at ~soren/... [15:57] try that [15:59] its permissions to pts/6 === dendro-afk is now known as dendrobates [16:04] hallyn: It wasn't on launchpad. It will be in a few minutes. [16:04] hggdh: I've set up an ssh key for the user on the server. If you make changes on the server, just commit and push. [16:05] soren: cool! I intend to get back to it as soon as I am done on Eucalyptus (sigh, never?), and update all to Maverick [16:05] hggdh: I just did that. [16:06] soren: sir, you rock! [16:06] hggdh: I thought I did that a couple of days ago, but I didn't do it properly. Now I did. [16:06] Daviey, ttx: please see bug 639639 [16:06] Launchpad bug 639639 in eucalyptus "euca-* commands stopped responding" [Undecided,New] https://launchpad.net/bugs/639639 === rgreening_ is now known as rgreening [16:07] * ttx cries [16:08] ttx: believe me, you are not alone [16:09] ttx: I would say this is another barf on the jetty server (or related to it), so I am restarting the whole bloody thing with '--debug' [16:10] but I will only monitor it, I have to finish my remaining QA tasks [16:10] hggdh, that is good.. thans [16:10] hggdh, can you add your thoughts to the bug.. i've added an upstream task to it [16:10] Daviey: will do [16:11] hggdh, \o/ [16:11] hggdh, I don't want you to miss your other WI because of us :) [16:11] Daviey: should we mark it High? [16:13] smoser: a problem with cloud-init on a lucid instance, see http://pastebin.ubuntu.com/494234/ [16:13] smoser: any clue ? [16:13] hggdh: soren: thanks much [16:14] hggdh, We don't get to use Medium or Low very much these days :) [16:15] hggdh, This issue hasn't yet been reproduced.. but certainly looks valid. Thankfully, this is an issue that can perhaps be fixed post-release.. So it's not as alarming to me as some of the other issues we have been encountering [16:15] ttx, that fair to say? [16:15] ttx, i can't really look at it right now [16:15] Daviey: I agree with the reasoning [16:15] "consuming user data failed!" i'd start where that message comes from [16:17] * Daviey suspects it's related to the context issue [16:19] anybody know of a good text editor / ide that has bzr support built in? [16:21] ttx, which lucid image ? [16:21] the original releases of lucid image did not have as good of debugging as more recent ones [16:21] smoser: hm, looks like I'm still running a borken metadata service [16:21] investigating [16:21] yeah. that is what i suspected. but for some reason i thought you said ec2 [16:21] which, of course, never has metadata issues :) [16:25] ttx, i'll be probably disabling pacemaker support in RHCS later today :( [16:26] gosh [16:26] im getting confused on what irc session to use [16:26] lol [16:26] RoAkSoAx: I prefer it disabled before FinalFreeze, rather than maybe-fixed in two weeks. [16:27] ttx: yeah me too. I'll just do that now then [16:28] RoAkSoAx: +1 [16:29] ttx: https://code.launchpad.net/~clint-fewbar/ubuntu/maverick/libdbi/upstream-0.8.4/+merge/34365 [16:30] ttx: the libdbi issue is fixed in that merge proposal (actually its more like a sync proposal)... [16:32] ttx: once that is uploaded, the other bug tasks can all be fixed by rebuilding and retargeting to libdbi-dev [16:35] SpamapS: won't have time to look into that branch today -- maybe see with matiaz for review [16:35] mathiaz [16:35] SpamapS: the graph is still broken btw, in a subtle way now [16:35] ttx: will do [16:35] oh? [16:36] you have to have designed it to see the issue, I guess :) [16:36] the end of the dotted trends line [16:36] should match day-1 bar height [16:36] should be at 85, not at 77 [16:36] ~ [16:37] so it's probably missing some of the stacking [16:37] SpamapS: at first glance, I'd say the inprogress ones [16:40] ttx: ok === dendrobates is now known as dendro-afk === dendro-afk is now known as dendrobates === bkonkle_ is now known as bkonkle [16:53] how do i list configured vhost with apache on ubuntu server - lucid ? [16:54] tydeas: apache2ctl -S [17:02] ttx, done bug #635884 [17:02] Launchpad bug 635884 in redhat-cluster "redhat-cluster fails to build from source in maverick" [High,Confirmed] https://launchpad.net/bugs/635884 [17:04] andreserl: won't have time to upload it today, maybe ask zul or mathiaz for sponsoring [17:05] ttx, ok :) [17:05] zul, ping [17:07] andreserl: pong [17:07] zul, oculd you please sponsor bug #635884 [17:07] Launchpad bug 635884 in redhat-cluster "redhat-cluster fails to build from source in maverick" [High,Confirmed] https://launchpad.net/bugs/635884 [17:08] andreserl: ack...in a meeting right now will do it after lunch [17:08] zul, sure thanks :) [17:13] ttx: fixed the chart line, please check in 2 hours. ;) === luis__lopez is now known as luis_lopez [17:14] SpamapS: I will :P === dendrobates is now known as dendro-afk [17:22] New bug: #639712 in libvirt (main) "PCI Pass Through via libvirt cannot remap IRQ's" [Undecided,New] https://launchpad.net/bugs/639712 === ivoks-afk is now known as ivoks [17:38] hggdh, Your overnight testing, what package were you testing against? [17:38] Daviey: euca 2.0+bzr1239-0ubuntu3 [17:39] hggdh, sure? [17:39] hggdh, 3 is the version in the archive [17:39] Daviey: darn! no... 0ubuntu3.2 [17:39] hggdh, hmm... interesting... ttx is seeing a similar issue with 0ubuntu3.2 [17:40] yes, same version [17:40] * Daviey sobs [17:40] ... and walks away for a bit.. before his head explodes. [17:40] Daviey: and I have another one (just had a look at the test running): of 1400 instances started, only ~400 succeeded [17:41] hggdh, I suspect that is the same issue [17:41] Daviey: no, not quite: on my first run eventually *NO* euca-* command returns [17:41] is my maths right in saying a 1 in 3.5 chance of seeing the issue :) [17:42] aHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH [17:42] on this current run it is still responding [17:42] Daviey: so it sounds like Yet Another Issue [17:42] hggdh, when you get a monent, can you upload your logs including euca-get-console(s) ? [17:43] for the current run? [17:43] "1400 instances started, only ~400 succeeded" .. That one [17:43] yes, current. I will cancel the run, and upload [17:43] hggdh, thanks... [17:43] * hggdh thinks another bug is warranted [17:43] i'm going afk for a bit. [17:44] o/ [17:44] Anyone here use / have experience with Freenas? [17:46] KillMeNow: its been a while [17:46] i had some issues with user management [17:46] but i think i would do better now that i know more [17:46] the freenas channel is so quiet i can hear a pin drop [17:46] i would use it again [17:46] i'm having trouble getting it to install on a Dell Powervault 745n box [17:47] AH [17:47] i finally just started saving the config to floppy [17:47] and running it from CD [17:47] easy to upgrade [17:47] :) [17:47] yea, i tried to install it with the LiveCD and it crashes after the menu [17:47] the installation menu that is [17:48] wondering if i need to get an older version [17:48] i downloaded the new full version, but it's not in an ISO or some other format i don't recognize [17:48] i really didnt see any advantage to installing it [17:48] when i was testing [17:48] * pmatulis thought freenas was based on freebsd [17:49] yup [17:49] sooo [17:49] there was talking of moving it to linux [17:49] but i dont think they did [17:49] it is based on FreeBSD [17:49] right, so why ask about it in here? [17:49] not sure if the new version is LInux... there is talk about integrating to Ubuntu [17:49] i really havent needed it since i got more comfortable with setting up services in linux [17:50] 12:46 < KillMeNow> the freenas channel is so quiet i can hear a pin drop === roaksoax__ is now known as andreserl [17:50] so this is the catch-all channel? [17:51] sure, why not... since i'm usually in this channel anyways, figured i would ask if anyone in here had experience with it [17:51] ive been wanting to try http://www.openfiler.com/ [17:51] what difference does it make really? [17:51] yea i was looking at that as well [17:51] but Freenas has a few more features [17:52] KillMeNow: imagine what this place would be like if everyone used your logic? [17:53] yea, maybe more people would expand their experience... [17:53] KillMeNow: you feel like you cant do the job with a server install? [17:54] Ubuntu server? probably [17:55] yeah [17:55] but if someone has something out there that already does it.... besides, it makes me learn something new and different [17:55] but i'll stop discussing it in this channel [17:56] i like these too http://www.turnkeylinux.org/ [17:56] those are based on ubuntusever [17:56] i think some of them are still hardy [17:56] hadn't heard about that [17:57] andreserl: done [17:57] zul, thank yuo :) [17:58] holstein: wow. I suppose that's good for a complete linux nuub [17:59] i tried the drupal and joomla ones [17:59] stupid easy [18:00] actually, i think i was the wordpress and joomla appliances [18:00] but it's good to know about, had no idea that project existed [18:01] i did a drupal install on a PPC server [18:01] those were quite a bit easier ;) [18:05] I'm trying to create an upstart script but when I use respawn then script it will not respawn [18:05] any reason why? [18:07] actually my script isn't running at all [18:07] soren: hello, I'm the one from the mailling list having bridging issues... [18:08] soren: let me know if you have an ideas of things I can try to debug the bridge creation process during startup. [18:32] New bug: #639768 in samba (main) "Samba process gets hung on maverick update" [Undecided,New] https://launchpad.net/bugs/639768 [18:34] Hi, can anyone point me to some docs about installing Ubuntu Server 10.4 from usb. I've tried to use netbootin and Universal usb install, it says that it cant find the cdrom. [18:35] drean: https://help.ubuntu.com/10.04/serverguide/C/index.html [18:36] Daviey: bug 639781 (link to the logs there) [18:36] Launchpad bug 639781 in eucalyptus "[AbstractSystemAddressManager:New I/O client worker #2-10] Found 2 vms with the same address: ClusterAddressInfo 10.55.55.101 172.19.7.131 orphanCount=0 -> i-363305B7(PENDING) i-3DE70712(RUNNING)" [Undecided,New] https://launchpad.net/bugs/639781 [18:39] * SpamapS smells fried chicken.. wonders if the pigeons strattled the powerlines again.. [18:40] hggdh, thanks [18:40] SpamapS, sounds like your lunch is set for today :) [18:46] ttx, Do you still have your maverick i386 setup, or did you trash it? [18:46] ttx, bug 628055 [18:46] Launchpad bug 628055 in libvirt "Instances don't start correctly: Security Labeling error running aa_change_profile()" [High,Incomplete] https://launchpad.net/bugs/628055 [19:40] RoAkSoAx: hey - re bug 635884 [19:40] Launchpad bug 635884 in redhat-cluster "redhat-cluster fails to build from source in maverick" [High,Confirmed] https://launchpad.net/bugs/635884 [19:41] RoAkSoAx: so we'll drop pacemaker? [19:41] mathiaz: yes! [19:41] unfortunatelly [19:42] mathiaz: all because of cluster-glue and pacemaker [19:42] mathiaz: we'll provide packages in PPA though [19:43] for natty we'll split pacemaker's libraries on its own packages to get this over with, but I'm not sure about what's gonna happen with cluster-glue since ivoks is in charge of it [19:45] Daviey: trashed it, have an amd64 now [19:50] SpamapS: o/ [19:50] SpamapS: reviewing the libdbi bug === ivoks is now known as ivoks-afk [19:52] RoAkSoAx: how does bug 635884 apply to cluster-glue? [19:52] Launchpad bug 635884 in redhat-cluster "redhat-cluster fails to build from source in maverick" [High,Confirmed] https://launchpad.net/bugs/635884 [19:53] mathiaz: the reason of redhat-cluster's FTBFS was because cluster-glue and pacemaker are not in main [19:54] RoAkSoAx: right - however there is a task open for cluster-glue [19:54] RoAkSoAx: does the bug require any work on cluster-glue to get fixed? [19:55] RoAkSoAx: should the bug be marked as Invalid/Fix Released for cluster-glue? [19:56] mathiaz: well it does, to be able to get it into main cluster-glue needs to be "fixed". However, we just decided on dropping pacemaker support so I guess it will be Invalid [19:56] RoAkSoAx: ok - I'll mark the bug as invalid for cluster-glue then [19:56] \just did it :) [19:56] RoAkSoAx: thanks for taking the time to untangle this issue [19:56] RoAkSoAx: cool - thanks! [19:58] np :) Though I wish we could have had this cluster-glue issue fixed to have redhat-cluster with pacemaker-s support. [20:02] RoAkSoAx: that's ok - we'll target natty instead [20:02] RoAkSoAx: natty opens for business in less than a month [20:03] RoAkSoAx: if you work on providing a PPA with maverick packages, you may be able to push everything early on in natty [20:03] mathiaz: yes that's indeed the plan :) [20:04] RoAkSoAx: o^5 [20:05] mathiaz: great! thoughts? [20:05] SpamapS: well - I was about to ask about the status [20:06] SpamapS: I'm not sure I fully understand what needs to be done [20:06] SpamapS: could we get 0.8.4 in maverick? [20:07] mathiaz: the merge proposal does just that [20:07] mathiaz: not having 0.8.4 in maverick means potentially having very confusing error messages [20:08] SpamapS: once 0.8.4 is uploaded, does it require any other source upload? [20:08] SpamapS: or 0.8.4 actually reverts the ABI break from 0.8.3? [20:08] mathiaz: yes, all of its build deps. They're all marked as bug tasks on bug 625882 [20:08] Launchpad bug 625882 in rrdtool "libdbi0: ABI breakage without package name change" [High,In progress] https://launchpad.net/bugs/625882 [20:09] mathiaz: 0.8.4 bumps soname [20:10] SpamapS: IIUC it also changes the -dev package name? [20:10] mathiaz: the original packaging had some mistakes.. so libdbi0-dev becomes libdbi-dev... [20:10] SpamapS: thus requiring source changes to all dependencies? [20:10] mathiaz: yep [20:10] SpamapS: ok - exploring another option [20:10] SpamapS: what happens if we don't do anything? [20:10] mathiaz: I'd rather do that than have a confusing -dev package name (libdbi0-dev -> libdbi1??) [20:11] mathiaz: anybody who has built programs linked against libdbi0-dev and used the error enums gets weirdness. [20:11] SpamapS: I'm trying to figure out if we should proceed given where we are in the release [20:12] mathiaz: the archive will be fine, none of them use the broken parts of the ABI [20:12] mathiaz: I'm more concerned with developers [20:12] SpamapS: right. [20:12] people who are building something new that we don't package. [20:12] SpamapS: understood [20:12] mathiaz: another option is to just have the dev lib remain the same for maverick [20:12] SpamapS: is libdbi mainly used on server application? [20:13] mathiaz: its a database abstraction layer.. so most of its uses are in server apps yes. [20:13] tho gammu is a dev tool for talking to mobile phones [20:14] SpamapS: option "have the dev lib remain the same" would still require a rebuild of dependencies in the archive [20:14] SpamapS: however we wouldn't need to change their source [20:14] mathiaz: right, they'd be noop rebuilds at least. [20:20] smoser: what was that url with the list of AMI's for maverick that is easily greppable? [20:21] smokie: ping [20:21] smokie: sorry [20:21] smoser: ping [20:21] smoser: when is /etc/hostname created in ec2 images? === _TechAway_ is now known as _Techie_ [20:21] early [20:22] ttx, ok [20:22] SpamapS, [20:22] uec-query-builds latest-ec2 --stream=daily --suite=maverick --build-name=server [20:22] I need to compile nginx because I need a custom module. What would be the best way to make the installation process as Ubuntu-friendly as possible? [20:22] or, http://uec-images.ubuntu.com/query/ [20:22] (i think you wanted the ladder) [20:23] smoser: sweeeet [20:23] smoser: actually no I wanted the former. :) [20:23] kirkland, its created by cloud-init (/usr/bin/cloud-init) [20:23] smoser: okay, cool ... and at instantiation of the instance? [20:24] once per instance, yes. [20:24] been writing up a bunch of things to do automated builds/tests on ec2.. but I'm tired of changing the ami every day. ;) [20:25] smoser: perfect, thanks [20:26] thafreak: Hi. Still having problems? [20:27] is there a way to get logrotate to name the files by the current date/time or epoch time? [20:27] anyone about my question on compiling and installing nginx in a Ubuntu-friendly way? [20:29] SpamapS, I'm using a vanilla ubuntu ami and customizing it with puppet [20:29] thafreak: Ok, here's the thing: [20:30] thafreak: The way it works is that every time an interface comes up (meaning the the kernel discovers it), an event is generated. [20:30] pting: man logrotate [20:30] pting: if you use 'dateext and dateformat you can do what you want [20:30] thafreak: This event triggers a script that checks if the interface in question is listed in another interfaces bridge_ports. [20:31] thafreak: If it is, this other interface is brought up. [20:31] thafreak: So, in the words of the guy who implemented this: [20:31] 16:22 < slangasek> soren: for a bridge, I don't know that you need any 'iface eth0' stanza at all; the 'iface br0' + 'bridge_ports eth0' + 'auto br0' seems to be enough, unless your eth0 needs other configuration [20:32] thafreak: Can you make it so, check if it works, and if it doesn't I can tell you how it's meant to work, then perhaps you can instrument things with debug blah yourself. [20:35] SpamapS, thanks! i swear i was blind when i read the man pages the first time === _Techie_ is now known as _TechAway_ [20:37] pting: it happens. :) [20:40] soren: sorry, was afk....I ended up fresh-reinstalling my two machines side by side [20:41] soren: the only difference between the two it seems was that the one that didn't work, I manually partitioned the disk (and may have added extra mount options to some mount points) [20:42] soren: Do you think that if I added say nodev to /tmp or /usr that it may have caused those problems? I know I put noatime on everything [20:43] It's just odd that it works now, and litterally the only thing different was my partitioning scheme and some mount options [20:43] thafreak: Depends. [20:43] thafreak: Er... I mean: Yes. [20:44] ..since you asked whether it may have caused problems. It might. Or it could be something completely different. [20:44] Well, it had to have been related to mount options, as that's the only thing I did differently this time and it works now. [20:44] thafreak: How did you do the partitioning? All from the installer? [20:45] yeah, one machine I must have done manual partitioning (and probably also messed with mount options), the other I just did guided lvm to make it quick [20:45] Separate /var ? [20:46] yeah, when I do manual, I tend to do all the important ones separate [20:46] so /, /var, /usr, /home, /tmp [20:47] I may have done nodev on /var and noatime...and maybe nosuid [20:47] thafreak: Would you have /manually/ split out /var or done it from the partitioner in the installer? [20:47] but I remember going back and removing everything but noatime from most of them just in case... [20:48] If the former, I have a reasonably good idea what your problem was. [20:48] I would have used the installer, and put /var on lvm manually...why? [20:49] Sorry, i'm confused. [20:49] Did you use the installer to put /var on a separate partition or did you do it manually? [20:49] during the installer, I choose manual as the partioning scheme... [20:49] Did moving /var to its own partition (or lv or whatever) involve the instaler. [20:49] Ok. [20:49] then that's not it. [20:49] The installer does the right thing. [20:50] what could have happened? [20:50] Things get /really/ wonky if there isn't a /var/run at all times. [20:50] So, if you have a separate /var, you still need a /var/run on your root filesystem. [20:51] once /var gets mounted, it'll block access to the underlying /var/run, but for the time between / getting mounted and /var getting mounted, you still need a /var/run that you can mount a tmpfs on. [20:51] Or at least write state to. [20:51] ah, well I know there was one, cause /var/run/network/ifstate was there and had the bridge in it, even though the bridge wasn't getting brought up [20:51] Without it... Well, things get really wonky. [20:51] oh hmmm [20:52] so maybe that's why it worked fine with just a root fs and no /var... [20:52] So, for instance.. [20:52] Something may want to write to /var/run/network/ifstate. [20:53] But.. [20:53] Gah, I can't even wrap my head around it right now. [20:53] right...I think I get what you're getting at [20:53] You just get all sorts of weird problems where you half the time have stale old data and half the time everything's great. [20:54] But usually, the time with stale old data is gone when you've booted up far enough to log in and see what's going on. [20:54] so, maybe I just make /var/log a separate lv, and not all of /var :) [20:56] Having /var separately is fine. I do it all the time myself. [20:56] Well, I really appreciate your help, and the community as a whole. [20:59] hallyn, hggdh: Sorry, I completely forgot to post you this link: https://code.launchpad.net/~ubuntu-server-autotest/+junk/autobuilder [20:59] thafreak: Sure. I would have loved to find out what the heck was going on :) [20:59] soren: thx, noted [21:00] zul: https://bugs.edge.launchpad.net/ubuntu/+source/mysql-dfsg-5.1/+bug/627723 [21:00] Launchpad bug 627723 in mysql-dfsg-5.1 "package mysql-server-5.1 5.1.41-3ubuntu12.3 failed to install/upgrade: subprocess new pre-installation script returned error exit status 1" [Medium,Incomplete] [21:01] soren: me too...but at the end of the day, it's time to move on and actually get these machines doing real work :) [21:01] zul: i wonder if we should find a better way to resolve this than have everyone come up with their own dangerous dieas :) [21:01] ideas [21:08] hallyn: yeah i know :) [21:08] zul: but so far you don't know of any? [21:08] zul: duncan's solution was pretty brilliant! [21:09] hallyn: i would have to think hard though [21:09] don't start a fire [21:09] :P === _TechAway_ is now known as _Techie_ [21:14] i'm getting weird usb error's since I updated 10.04 last [21:14] the error is "usb 1-3: device descriptor read/64, error -110" [21:32] How can i config php5-mcrypt> [21:32] ?? [21:33] I have install mcrypt/libmcrypt/libmcrypt-dev and php5-mcrypt as well... [21:33] what should i add to the php.ini? [21:34] This is what i found in the php.ini http://www.pastebin.org/876170 [21:38] hola [22:06] New bug: #639946 in samba (main) "System crashes on samba update - unable to recover" [Undecided,New] https://launchpad.net/bugs/639946 [22:07] Why apache runs to many times? [22:07] http://www.pastebin.org/876495. [22:07] *http://www.pastebin.org/876495 [22:10] hello everyone, my network interfaces is setup for static ip and today when i tried to connect ftp it will not connect to same ip address, but when i switched monitor inputs, and logged in to the ubuntu-server, eth0 shows a different lan ip than what /etc/network/interfaces shows why ? [22:12] <_Techie_> Four2zero, can you please put yor /etc/network/intefaces on pastebin [22:12] <_Techie_> Four2zero, you should be used to this by now [22:13] http://pastebin.com/ZG5GQbf9 [22:14] i checked my /etc/hosts as well and it was set to 127.0.1.1 so i think that might have something to do with it. [22:14] <_Techie_> nah [22:15] <_Techie_> IIRC 127.0.1.1 is used to bind to local domain [22:15] <_Techie_> sudo ifdown eth0 && sudo ifup eth0 ? [22:19] okay, its back to eth0 192.168.0.100 === IdleOp is now known as IdleOne [23:02] Four2zero, the problem is that you have to run `sudo service networking restart` or reboot the computer after changing to a static IP, otherwise dhclient stays running in the background, and when your lease expires, it renews it, overriding the static assignment [23:15] qman__: Restarting networking will not help. [23:26] Don't know if this fits here or not, since I *am* running server, but I suppose it could be a more general question: Googling and reading docs seems to indicate to the negative, but it is possible, with LVM2, to take atomic snapshots of multiple LVs? [23:34] is there a way to tell rsync to not sync a specific folder and its subfolders, assuming I am doing a recursive rsync? [23:37] EvilPhoenix: --exclude=/path/to/folder [23:38] ah was looking for that [23:38] thanks [23:39] np [23:45] * EvilPhoenix now goes to rsync 500GB of data over wifi >.> [23:53] EvilPhoenix: No physical connection anywhere, eh? [23:53] pedahzur, dead ethernet nic card [23:53] pedahzur, laptop [23:53] ack. [23:53] ikr [23:54] i'll rsync it in 50GB parts [23:54] You can't remove the drive? External drive connections are only $20-$30. [23:54] throughout the next few hours [23:54] pedahzur, rsync to a VPS [23:54] so no [23:54] EvilPhoenix: Oh. [23:54] EvilPhoenix: You can mail drives to Amazon S3. :-D [23:54] s/VPS/server\ \/\ VPS/ [23:54] x] [23:55] i think i typed that correctly... hitting the spaces and special symbols... [23:55] :/