=== _Techie_ is now known as _TechAway_ [00:06] can someone help me get remote access working on my server? [00:06] using ssh? [00:07] on the server, and putty on windows 7? [00:09] Just install openssh [00:09] already did [00:09] I don't know what to do from here [00:10] http://beyond-sight.com/ [00:10] router is forwarding correctly [00:10] Oki, on putty, just set your server ip, port 22, and press open [00:11] eriksson25: Connection refused. [00:11] eriksson25: my router is probabaly not forwarding that port, one moment [00:11] Exactly [00:11] =) [00:12] How can I set my ubuntu ethernet ipaddress? [00:12] 192.168.1.250 for example? [00:12] it's currently 192.168.1.9 [00:13] Depends on, you can set it from the router. Binding its mac to a ip. [00:13] You can set it grafical if you have it on the server. Or you can set it in commandline. [00:13] did ssh work now? [00:14] You shuld not have it on port 22. [00:14] havne't forwarded the port to the pc yet [00:14] verizon fios router [00:14] doesn't say anything about a mac address [00:14] before you do that, change the port. [00:15] change port how? [00:15] to a higher number like 5022 or something. [00:15] I don't know how to change the openssh port to something else [00:15] open file /etc/ssh/sshd_conf [00:15] 26254 in this case, is what I'd like. [00:15] okay [00:15] sudo nano [00:15] and there go down to port= and change it to whatever. [00:16] Then you need to run /etc/init.d/ssh restart [00:17] http://codesnippets.joyent.com/posts/show/319 [00:18] Thats how you set static ip to whatever you want on the server. [00:18] hm [00:18] ssh_restart [00:18] or ssh restart [00:18] no _ [00:18] nog being found as a command [00:19] ssh restart [00:19] Restarting SSHD? [00:19] I cd ed to the directory init.d [00:19] No need. [00:19] ssh restart isn't doing anything [00:19] Just type "/etc/init.d/ssh restar" [00:19] kirkland: hey, how come ubuntu-vm-builder doesn't accept maverick inside maverick? [00:19] sudo service ssh restart [00:19] And press enter. [00:20] VMBuilder.exception.VMBuilderUserError: Invalid suite: "maverick". Valid suites are: dapper gutsy hardy intrepid jaunty karmic lucid [00:20] ssh start/running, process 1631 [00:20] or that way as nichos says [00:20] Yes, now its using the new port. [00:20] awesome [00:20] Portforard it and you are good to go. [00:21] ZacharyNewb: Welcome to Ubuntu. [00:21] Hmm, wierd hehavior on iptables. I use iptables -t nat -A PREROUTING -i ppp0 -p tcp --dport 4000 -jDNAT --to-destination 82.211.214.96 This dont work if i try to connect to ip of the ppp0 interface, if i change ppp0 to eth1 and connect to that ip the packages are send of to 82.211.214.96 (are there some special thing with pptp connections) ? [00:22] eriksson25: those instructions on that link aren't working [00:22] no interfacing directory or file [00:22] wow, just found it nevermind [00:22] =) [00:23] Anyone familiar with using DRBD to create an active/active ISCSI target? [00:23] sudo nano interfaces [00:25] eriksson25: okay, I changed my ip address in the config file, and it matches that text except [00:26] he kept having "192.168.0.1" [00:26] I put "1.1" [00:26] Yes ofc, since that would be your router. [00:26] is my rendition correct? [00:27] ofc = ofcourse, ah. xD [00:27] okay [00:27] do I need to reset the ethernet connection somehow for the change to take effect? [00:27] eriksson25: [00:27] Yes, Dont remember how.... [00:28] sudo service networking restart [00:28] there you goo. [00:29] ah [00:29] eriksson25: nich0s Thanks, I also found "/etc/init.d/networking restart" [00:30] with yours, sudo service, IO'm getting "restart: unknown instance" [00:30] Yup, two ways of doing the same thing. [00:31] the one I found works [00:31] requires sudo though [00:31] also [00:31] sudo ifdown -a [00:32] to bring them down [00:32] sudo ifup -a [00:32] back up [00:32] to turn the network adapters on and off all-together [00:32] okay [00:32] eriksson25: Router is portforwarding 26254 [00:33] hm [00:33] connection refused [00:33] on the server, run ifconfig [00:34] And se if it says the correct ip [00:34] yes [00:34] correct ip [00:34] Is the server within the same lan? [00:34] yes [00:34] beyond-sight.com [00:35] Then you dont need portworading, thats only from outside. [00:35] is correctly being forwarded port 80 to the server, I get the web test page [00:35] true. [00:35] I travel sometimes as well [00:35] for now I'll test the network, you're right [00:35] Yes, but we get it to work internat first. [00:35] =) [00:36] In putty, did you sett the internal ip? like 192. and then the new port [00:36] putty 192.168.1.250 port 26254 [00:36] connection refused [00:36] however in the ethernet adapter settings you had me view, it's showing "bcast" (broadcast) as 192.168.0.255 [00:36] I say, reboot the server. Might me something with the ip numbers scruwing with you. [00:37] That is correct. [00:37] doesn't it need to be 0.1 ? [00:37] Or waait, now its noot. [00:37] like I changed all the others? [00:37] Shuld be 1.255 [00:37] -nods- I missed that then [00:38] everyting has to be 192.168.1.x [00:39] and "network" 192.168.0.0 should be 192.168.1.1 ? [00:40] network card reset [00:40] trying again [00:40] 1.0 [00:40] k [00:42] fixed again [00:42] trying putty again [00:42] dont know with win7, you might need to run it as admin. [00:42] I use xp as my windows comp. [00:43] connection refused still [00:43] using local lan address [00:43] I'm going to ping beyond-sight.com to see if I get response [00:43] and see if the web default page still pops up [00:43] web default page still up [00:44] do I need to configure open ssh further perhaps? [00:44] Reboot the server. [00:44] back in it's config file? [00:44] alright [00:44] do you know how I can do that from command line? [00:44] you shuldent need to. [00:44] sudo reboot now [00:44] thanks [00:44] hmm, sudo reboot lastnight :) [00:45] sudo reboot lastnight? sat mean? [00:45] He was funny [00:45] wishful thinking :) [00:45] I don't get it [00:45] =) [00:45] Zachary newb here doesn't get it. >.> [00:45] I didn't realize I typed "I don't get it" [00:45] that is more funny than my lame attempt :) [00:45] its ok. [00:46] LAte night humur. [00:46] it had better be [00:46] awesome server back up [00:46] Oki, try again and se what hapends. [00:46] connection refused again [00:46] Oki, try port 22 [00:47] just for the fact. [00:47] whoa [00:47] got something new [00:47] Wants to save a key? [00:47] "the server's host key is not cached in the registry" [00:47] yes [00:47] Save it [00:47] ah, it's looking for a date [00:48] ... [00:48] what it called? [00:48] deflection? [00:48] your attempting to move it to an alternate port? [00:48] Yes [00:48] yes, one previous attempt apparently failed [00:48] however changing the ip address did work [00:48] just edit /etc/ssh/sshd_config [00:48] That what I told him. [00:48] Port xxxx [00:49] sshd ? [00:49] there's also a ssh_config [00:49] service ssh restart [00:49] ssh_config is for the client, sshd_config is for the server [00:49] nor did I restart it before I restarted the server [00:49] so ssh_config == outgoing [00:49] sshd_config == incoming [00:49] open file /etc/ssh/sshd_conf [00:49] awesome [00:49] what I typed earlier [00:49] eriksson25, ya, I didn't scroll back that far :) [00:50] Its cool [00:50] * PatrickDK never runs ssh on alternate ports [00:50] oh my god [00:50] Thats why I asked him to change putty to 22, since I belived he would missed it. [00:50] so much easier typing it into putty here next to this window [00:50] =) [00:51] So putty up and running? [00:51] * PatrickDK has never used putty [00:51] =), well some of us plays computergames, and dont come dragging about wine. [00:52] I use linux, for Server, Router, HTPC [00:52] service ssh restart isn't working [00:52] heh, I use linux for everything [00:52] only use windows for vcenter [00:52] alright [00:52] nevermind, has to have sudo [00:52] =) [00:52] Yes, Putty works, now I've switched opensshd port to 26254 [00:53] including the client ssh port on the server too [00:53] Good boy [00:53] Now you are redy to do some real work. I am off to bed now. [00:53] YAY [00:53] man, I do like these new hp blades so much better than the old ones :) [00:53] Thank you, mother. >.< [00:54] PatrickDK: I'm on a netbook [00:54] No problem. [00:54] Have fun. [00:54] PatrickDK: The Asus 1201n Dual core 1.6ghz, hyperthreading/4 virtual cores, with ion graphics, including a GPU [00:54] and it's overclockable to 2 ghz [00:54] $500 [00:55] I can't use netbooks [00:55] this one is really nice [00:55] 3-4 hours regular use battery life, nicely sized keyboard, 11 inch screen [00:55] Full HD video [00:56] heh, I'm all about 6hours battery [00:56] I can run two 1080p monitors on this, dual screen, one on the VGA port and another on hdmi [00:56] using a t61p currently [00:56] This thing is wayyy too powerful to be compared to netbooks [00:56] I'm lacking hdmi on this one, my last one had it, loved it [00:56] lately I've been lazy [00:56] Actually, They compare my netbook to fully priced laptops [00:56] and just use my droid2 for everything [00:57] I'm jealous [00:57] I have a samsung propel [00:57] anywho [00:57] nothing is better I think than ssh+rsync on my phone :) [00:57] I need to configure the web server somehow [00:59] where to start? [01:00] PatrickDK: apache is installed [01:02] My website is residing on the secondary hard drive (500 GB) [01:02] formatted probabaly as NTFS [01:04] sudo mkdir /media/windows [01:04] sudo mount /dev/hda1 /media/windows/ -t ntfs -o nls=utf8,umask=0222 [01:35] clear [01:35] can someone help me? [01:38] I've just installed Ubuntu Server. With your peoples' extensive help, I've configured it's ip address, mounted an extra drive and I'm not remotely accessing it through putty [02:08] kirkland, regarding MIRs, if the -desktop is an official supported image, it can only include things in main [02:09] smoser: i think for Natty we should prove something unsupported, using Universe first [02:09] smoser: before we commit to stuff in Main [02:09] we basically have something unsupported [02:09] and have for 2 releases [02:28] smoser: right; i'm thinking we should get heavier feedback on that for a cycle before we start picking and choosing what needs to go into Main [02:28] thats fine. but for that we dont really need to do anything. [02:28] a couple cloud-init scripts that allow people to test [02:37] Do any of you know how to format a second hard drive? === jjohansen1 is now known as jj-afk === ZacharyNewb is now known as ZacJstSetUpHisSr [02:51] lol === ZacJstSetUpHisSr is now known as NewSrvrUsrRdyToH === NewSrvrUsrRdyToH is now known as SrvrUsrRdyToHelp [02:51] :D [03:05] Hello? [03:30] Are people here? [03:32] Hello? [03:32] hello? [03:32] hello? [03:32] LaND HO [03:42] SrvrUsrRdyToHelp: There are people here, but this channel is usually pretty quiet this time of day. [03:43] ScottK: I see [03:43] ScottK: Well, I'm slowing peicing together my ubuntu server [03:45] Great. [03:45] ScottK: remote access via putty with windows 7 netbook I have is working, ports, right now I just figured out how to format and mount a second hard drive, put it into the startup script to mount automatically, and now I'm thinking about how to configure apache server to serve my site, off of the second mounted hard drive [03:46] Sounds like you're making progress. [03:46] ScottK: Definitely. Manually set all ethernet settings, etc etc etc [03:46] ScottK: Hours of fun, Are you enjoying your day? [03:46] * ScottK has had a decent day, but is just about to head for bed. [03:47] * SrvrUsrRdyToHelp is sorry to hear that. :( [03:59] hi everyone [03:59] hi there [04:00] what's up? [04:00] b0gatyr: what's up? [04:00] not much ya know [04:00] how about you? [04:27] Anyone here used DRBD to create an ISCSI Target? [04:33] someone here? [04:33] hello? [04:41] sup [04:44] ..... [04:44] :-\ [05:17] hi [05:17] can someone please guide me about https://bugs.launchpad.net/ubuntu/+source/openssh/+bug/379329 ? [05:18] Launchpad bug 379329 in openssh "CVE-2008-5161: OpenSSH CBC plaintext recovery" [Low,Fix released] [05:38] hello people; [05:40] kaushal: what about it? [05:43] here's my scenario: i should make a web interface for iptables, (yes, it's a firewall web panel). problem is apache runs web page as "nobody" so i can't execute "iptables". question is "since there's a apache process run as root, is it possible to make apache use that process to run a web page?" [05:44] Dark-Sun: no [05:45] Dark-Sun: that process is only responsible for listening on port 80, which is restricted to root access (all ports below 1024 are) [05:46] Dark-Sun: You can, however, run iptables commands with sudo. However, I do not recommend this as this would mean anybody who can break into your web code can run any iptables command... [05:47] Dark-Sun: is there any reason you don't want to use one of the existing iptables frontends, like smoothwall for instance? [05:47] SpamapS: you mean that apache process can't do anything else but listening? [05:48] Dark-Sun: basically, yes [05:48] Dark-Sun: it just listens, and hands accepted connections to the child processes. [05:50] SpamapS: actually i was about to do some other stuff beside firewall, and i was curious how they do this [05:51] i got some work around clues, like jailing, a daemon running on the web server, but i don't know which one should i choose [05:57] SpamapS: hi [05:57] sorry was away [05:57] is there a fixed version available for 8.04 ? [05:57] I mean openssh-server [06:06] Dark-Sun: a daemon that is chrooted off and very simple would work, as long as it has a way to verify who is communicating with it. [06:07] kaushal: looks like https://bugs.launchpad.net/ubuntu/+source/openssh/+bug/379329/comments/3 has a workaround that will work for hardy. [06:07] Launchpad bug 379329 in openssh "CVE-2008-5161: OpenSSH CBC plaintext recovery" [Low,Fix released] [06:07] SpamapS: ok [06:08] so is it to add this Ciphers aes128-ctr,aes256-ctr,arcfour256,arcfour,aes128-cbc,aes256-cbc to to ssh_config and sshd_config and restart the daemon and restart openssh server ? [06:09] and then rescan the server again with OpenVAS Server ? [06:09] SpamapS: hummm... i should read more about it, thanks ;) [06:10] kaushal: also mdeslaur has replied in bug #651720 about why the patch was not pushed back to hardy. [06:10] Launchpad bug 651720 in openssh "Hardy OpenSSH version out-of-date - security risks" [Undecided,Invalid] https://launchpad.net/bugs/651720 [06:11] SpamapS: did you mean "executing daemon with a user with limited root access"? [06:12] SpamapS: ok [06:16] Dark-Sun: the daemon would have to either run as root, or use something like sudo to temporarily elevate its privileges for each command. [06:20] SpamapS: i see. thanks :) [07:24] New bug: #676304 in samba (main) "Samba fails with files having colon in file name" [Undecided,New] https://launchpad.net/bugs/676304 [08:30] is Empathy's MSN connecting broken for anyone else? the fix here: http://ubuntuforums.org/showpost.php?p=10003289&postcount=17 is already applied but it still doesn't want to connect [08:31] Callum__: that's probably offtopic for -server. [08:31] oh, oops [08:31] wrong channel [08:31] haha [08:32] SpamapS: I would do it by having the web page write to (say) /srv/www/rw/netfilter.rules, and have a separate root-owned cron job or inotify watcher that ran iptables-restore on it. [08:33] Although obviously web-based management of ANYTHING is WRONG. [09:14] m0n0wall? [09:16] alvin: what about it? [09:18] It's only web-based [09:19] Are you suggesting it *isn't* wrong? [09:24] Certainly. I prefer the command line, but m0n0wall is a very nice piece of software. [09:27] Maybe you just have lower standards than I do [09:28] You wouldn't say that if you had used m0n0wall, or something like it. [09:28] Using PHP in the boot sequence doesn't sound very clever to me. [09:29] I can get behind standardizing on XML as a configuration file format, though. [09:29] neither is using upstart [09:29] Oh, definitely. I hate upstart. [09:29] does someone know if drbd replication is only one way or is it possible to write in both machines? [09:32] alvin: the HTML they're using is pretty crap, at least. [09:33] I can see they're using tables for alignment and instead of , in the first thirty seconds of RTFSing [09:34] <_ruben> binBASH: master/master is supported [09:34] twb: You have m0n0wall running? Or do you mean their website itself? [09:34] <_ruben> does open up a nice can of worms though [09:34] alvin: I'm reading http://svn.m0n0.ch/wall/tags/release-1.32/webgui [09:35] alvin: obviously I'm not going to RUN anything PHP-based. [09:36] ah, ok. Could be. But I only used it as an example of a well working and stable web-based systems. of course, there are others. Managing ZFS on Solaris web-based will simply not work, SWAT (samba) isn't the same as smb.conf either, and eBox isn't there yet. I only wanted to say there are exceptions. [09:36] _ruben: say, you have a clue, and you know about drbd. [09:36] _ruben: I have a deployment case with side-by-side identical hardware, each with a two-disk RAID1. I want the backup unit's disks slaved to the production unit. I have been doing nightly rsyncs, and I started looking at instead adding the slave's disks directly to the RAID1 (with --write-mostly) over AoE. [09:36] _ruben: is drbd a better idea than aoe in that case? [09:37] alvin: "it works" isn't the same as the Right Thing. [09:38] alvin: as far as "it works for me" goes, I'm relatively impressed by OpenWRT's solution, which involves a YAML-esque configuration format for everything, which has both a nvram get/set-style CLI and a lua-based web UI. [09:39] twb: I agree. let's say: "it works very well and is stable". the Right Thing isn 't always the best solution. For example, Ubuntu removed libstdc++5 in Lucid. (Yes, the LTS). This broke many commercial applications and thus production systems. The right thing would have been to ask the commercial vendors to recompile their products, but you know you can't really do that. [09:39] Good to know :) I'll look at OpenWRT [09:40] alvin: products that you can't recompile are inherently not The Right Thing :-) [09:40] twb: No, but tell that to the management if your company really relies on it. [09:41] I could probably make a case that companies are the Wrong Thing, too -- but I'm not an economist :-) [09:41] * twb impotently shakes fist at the world [09:42] that brings a smile to my face === nijaba is now known as nijaba_afk [10:30] <_ruben> twb: afaik, drbd would be a nice a solution for that (dont have any hands-on experience with it yet though) [10:30] Nod [10:31] Now I have to convince to pay for another day of learning drbd (on top of aoe) [10:32] <_ruben> personally i'd prefer drbd over raid1+aoe [10:32] <_ruben> the latter sounds a tad hackish ;) [10:42] Nobody thought of drbd when aoe+md was proposed === nijaba_afk is now known as nijaba [10:58] <_ruben> hehe [11:01] The rsync technique we've been using for ages -- before USB keys were bootable, and we were doing it from a live cd. [11:02] New bug: #676418 in etckeeper (main) "not installing http[s]+webdav" [Undecided,New] https://launchpad.net/bugs/676418 [11:04] twb: csync maybe? [11:04] Though I agree with the +1 for drbd. [11:04] <_ruben> so they're "offline" rsyncs?? [11:04] <_ruben> as in: not from the running system itself [11:04] _ruben: no no [11:05] We have the slave run a live image, so to failover, all the monkey has to do is unplug the CD/USB key, and move the ethernet/power cables [11:05] Er, s/power//, but you would need to reboot the slave [11:06] <_ruben> ah [11:06] monkey-based failover was a core requirement [11:10] Consider a lucid server running about twenty lucid containers (broadly, one per service). End users do not have access to the dom0, and creating/reconfiguring/destroying containers is an infrequent activity. [11:10] Is there ANY benefit to lxc-behind-libvirt instead of direct lxc? [11:36] New bug: #676439 in freeradius (main) "missing manpages, listed in the freeradius manpage" [Undecided,New] https://launchpad.net/bugs/676439 [11:37] <_ruben> grr, $PS1 gets set/reset/overwritten in tons of places .. /etc/bash.bashrc and also a .bashrc in each user's homedir, and even within those files there's multiple blocks to set $PS1 [11:38] _ruben: why is that bad? === ihCiFriS is now known as SirFiChi [11:39] <_ruben> twb: no easy way to change the prompt globally [11:39] <_ruben> guess i should mess up /etc/skel/ prior to adding users [11:40] _ruben: why do you want to do that? [11:40] I mean, /etc/bash.bashrc would be the Right Way, but it sounds like you want to prevent users changing it back to whatever they prefer [11:40] <_ruben> twb: nah, i want to offer a saner default (fqdn instead of short name) [11:41] That's /etc/skel and/or /etc/bash.bashrc, then [11:41] And too bad for existing users [11:42] <_ruben> fresh servers and only handful of users, so it's not big a deal, is kinda annoying though .. the only diff between root's .bashrc and users' .bashrc is the bash autocompletion being disabled for root [11:42] <_ruben> a saner approach would've been to use /etc/skel/.bashrc as /etc/bash.bashrc and give root it's own .bashrc (which it already has) [11:42] <_ruben> current approach sems like a cludge to me [11:43] _ruben: I am *strongly* opposed to putting "clever" things in the default /etc/bash.bashrc [11:43] RHEL does that and it makes me livid [11:43] e.g. I ssh into a RHEL box and it changes my terminal title to gibberish [11:44] And if you, a user, want to *disable* such functionality, your .bashrc becomes horribly complicated by lines like "if I running on RHEL>=3 then unset magic_variable; magic_function() {:;}" [11:45] <_ruben> why would a .bashrc nee... oh, home-on-nfs for instance [11:45] Right [11:45] Or even if you store $HOME in a VCS [11:46] Even the bloody command-not-found stuff annoys me [11:46] unset command_not_found_handle # Ignore Ubuntu's attempt to slow exit(126) to a crawl. [11:47] <_ruben> if only i could trick bash to show the fqdn to show up for \h .. then again, even the supposed-to-work \H doesnt result in fqdn being displayed [11:48] Then there's http://paste.debian.net/100030/ for when you're sshing from Debian/Ubuntu to a server without locales [11:48] <_ruben> why not just install command-not-found? :P [11:48] _ruben: because I don't have root on all hosts [11:48] <_ruben> ah [11:48] And even if I did, other users might want that [11:48] (Remember those "server" things that have >>1 user? ;-P) [11:49] * twb rages on re-discovering you can't use kvm -curses with the lucid installer [11:50] <_ruben> i probably am the only user who actually cares about this stuff (users are sysadmins in my case, with only bare minimum linux knowledge) [11:52] Is there an option for df I'm not seeing? I want df to output in bytes instead of kilobytes. [11:53] alvin: df -B 1 ? [11:54] WFM [11:54] or df --blocksize=1 [11:54] Same thing [11:54] yup [11:54] twb: ok, a valid option I didn't see. Thanks. (It also appears that the manual is not entirely correct. 1k blocks are used by default while the manual says 512b) [11:55] ...and that's also in the man. I was looking at the man of another version... [11:55] <_ruben> Disk space is shown in 1K blocks by default, unless the environment variable POSIXLY_CORRECT is set, in which case 512-byte blocks are used. [11:55] Yes, just saw it :-) [11:55] where can I check out the installer's code ? [11:56] server installer [11:56] alcy: apt-get source debian-installer [11:58] jpds: doh. thanks. :) [11:58] One day I'll set POSIXLY_CORRECT just to laugh at how much breaks [11:59] <_ruben> ;) [12:01] Did we decide where -t cgroup should be mounted, since /cgroup is a FHS violation? [12:03] <_ruben> bah .. with linux you can create a tunnel using ipv6 endpoints and carrying both ipv4 and ivp6 inside (mode any), but you can't do so over ipv4 :( [12:03] <_ruben> oh well, enough reason to tunnel ipv4 over ipv4 and ipv6 over ipv6 [12:03] _ruben: just get native ipv6, man [12:03] Er, maaaaaan [12:04] <_ruben> twb: i do, but i don't have a vlan (yet) between the various pops [12:06] * twb squints [12:06] If you have a native ipv6 connection to the IPv6 internet, surely you don't need a vlan [12:07] unless you mean you want to encrypt ipv6 traffic passing across the ipv6 internet? === oubiwann is now known as oubiwann-away [12:08] * jpds hugs the native v6 internets === oubiwann-away is now known as oubiwann [12:08] hi, which is the better choise between maverick and luci for a KVM/corosync server? where i can get these kind of information, to know what is really changed about single kernel between these two distro? Thanks [12:11] ZMo: if you plan to use that system some time, use the LTS release (-> lucid) [12:11] ZMo: and for the kernel changes you'll need to check the kernel changelogs from kernel.org (or at least the short logs from kernelnewbies.org) [12:13] <_ruben> twb: dynamic routing between pops, that requires 'direct' connections between pops, so either vlans or tunnels [12:13] joschi, so is theren't a ubuntu kernel changelog? [12:14] ZMo: yes, but these only contain infos about the patches applied by debian/ubuntu to the respective kernel versions [12:15] ZMo: I'd also check Launchpad in order to know what bugs you will encounter in each version and whether they are important to you. [12:18] <_ruben> weird, openipmi package refers to /etc/sysconfig/ipmi .. didn't know debian/ubuntu even used to that dir [12:18] bad docs [12:18] bad docs? [12:29] <_ruben> ugh, more fail in the initscript [12:29] <_ruben> touches a file in /var/lock/subsys/, without checking if that dir even exists [12:35] Hello! [12:35] Boo! [12:35] don't suppose I could you trouble you guys for some support [12:35] Debian Policy requires that /var/run and /var/lock be supported as tmpfses [12:36] !ask | Ben604 [12:36] Ben604: Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) [12:36] we're running ubuntu server within hyper-v, behind an internet proxy server [12:36] we can ping to our internal network fine [12:36] but can't reach the big wide world [12:37] the proxy settings match our other, non-virtualised, linux boxes [12:37] which can connect to the outside world fine. Which leads me to believe it's a problem with hyper-v? [12:37] Ben604: so on the virtual Ubuntu server, you're running http_proxy=http://proxy.example.org:8080/ w3m http://example.net/ ? [12:37] yeah, that's correct [12:38] What error does it give? [12:38] Such issues are easiest to debug with "curl -vso/dev/null http://example.net/", but you probably don't have curl installed. [12:39] wget -vO/dev/null, I think, is the nearest equivalent [12:39] actually, that worked... [12:39] I'll be more specific [12:39] Ben604: OK, pastebin the output from curl [12:40] we definitely haven't got curl installed, I've just been reliably informed [12:41] Oh, you mean the *w3m* call worked? [12:41] it seems, that when I enter this command: http_proxy=http://proxy.example.org:8080/ w3m http://example.net/ ? it works fine, but it won't ping [12:42] ICMP won't work unless hyperv is bridging or natting traffic from the VM [12:43] If I were you, I'd ssh into your router(s) and start tcpdumping [12:43] we're trying to get some packages through apt-get, but that won't find anything, we've added the proxy settings to the etc/apt/apt.conf [12:44] (sorry about this, I'm sure you've noticed that I'm not exactly a long time Linux user!) [12:44] Ignorance is fine, as long as you're prepared to learn [12:45] can I use "ping" instead of "w3m" as a command? [12:46] Sure, but ping uses ICMP, not HTTP/TCP [12:46] <_ruben> Ben604: if w3m worked, try: sudo http_proxy=http://proxy.example.org:8080/ apt-get update [12:46] _ruben: he just set $http_proxy in /etc/apt/apt.conf, so he should be OK for apt-get runs from now on [12:47] <_ruben> twb: trying to rule out a wrong apt conf [12:47] OK [12:47] <_ruben> also, on proxied systems like this, it might be "useful" to just export http_proxy [12:48] <_ruben> any decent proxy-aware app will then use the proxy [12:48] right [12:48] sudo http_proxy=http://proxy.example.org:8080/ apt-get update worked [12:49] so it looks like the apt.conf file is the problem. [12:49] <_ruben> Ben604: so either fix the apt config, or add 'export http_proxy=http://proxy.example.org:8080/' to your login script [12:50] Preferably both [12:50] <_ruben> wouldn't harm indeed [12:51] $ ssh ueo cat /etc/apt/apt.conf [12:51] Acquire::http::Proxy "http://proxy:8080/"; [12:51] That's what is *should* look like. [12:51] guys, thank you very much! nfs-common is now installing [12:53] just checked apt.conf, didn't have "/" at the end, would that have been critical? [12:54] <_ruben> apparently ;) [12:55] aw, isn't the linux community ace [12:55] through irc too! It's almost as if it's 1997 again [12:56] <_ruben> aww bugger, ip's man page is horribly outdated [12:58] Need help, WARNING: missing /lib/modules/2.6.35-22-generic, I need to reinstall the kernel to /lib/modules [12:58] But how? [12:58] <_ruben> debian 504064 [12:58] Debian bug 504064 in iproute "ip: missing documentation for ip link add|del" [Normal,Fixed] http://bugs.debian.org/504064 [12:58] Ben604: What version of Ubuntu are you running as a virtual machine? [12:59] hi everyone [13:00] I try to install rabbitmq-server on my ubuntu 8.04 server using the rabbitmq apt repository, but it complains about an old erlang-base version. how could I update that? [13:06] eriksson25: apt-get install --reinstall linux-image-2.6.35-22-generic ? [13:06] eriksson25: you should probably investigate what caused the issue [13:07] _ruben: my ip manpage explains ip link add... (sid) [13:07] Noob behind the keybord. [13:08] Thanks [13:09] <_ruben> twb: ah, nice [13:09] WTF, man [13:09] apt-get install ubuntu-keyring wants to pull in imagemagick and xloadimage [13:10] <_ruben> 10.10 does too apparently (list link add) [13:10] <_ruben> twb: only depends on gnupg (on lucid) [13:13] How do I solve this? [13:13] ***WARNING:*** You do not have the full kernel sources installed. [13:13] This does not prevent you from building the v4l-dvb tree if you have the [13:13] kernel headers, but the full kernel source may be required in order to use [13:13] make menuconfig / xconfig / qconfig. [13:14] <_ruben> either install the full kernel source or ignore the warning [13:14] Why are you trying to compile anything from source? [13:16] I am trying to compile mantis driver for a tv card. [13:16] And I get that error, and then [13:16] make[1]: Entering directory `/home/eriksson25/s2-liplianin/v4l' [13:16] perl scripts/make_config_compat.pl /lib/modules/2.6.35-22-generic/build ./.myconfig ./config-compat.h [13:16] File not found: /lib/modules/2.6.35-22-generic/build/include/linux/netdevice.h at scripts/make_config_compat.pl line 15. [13:16] make[1]: *** [config-compat.h] Error 2 [13:17] It worked fine before, but then I needed to reinstall ubuntu after a hdd crash, and now.... [13:19] <_ruben> i'd start by installing just the kernel headers, might be sufficient [13:19] sudo apt-get install mercurial linux-headers-$(uname -r) build-essential [13:20] not helping [13:22] How do you write an upstart job that takes an argument? [13:22] I want to write a dozen /etc/init/lxc-foo.conf with identical contents except for the VM name they refer to ("foo"). [13:23] * _ruben makes an uneducated guess: not possible? :P [13:23] Goddamn upstart [13:24] They have a special case for mounting filesystems, and a special case for rc.conf, but for us end users have to do everything by hand. [13:28] _ruben: I reckon I know [13:29] _ruben: you have lxc.conf that says "start on filesystem" and it's *body* is "for CONTAINER in $CONTAINERS; do start lxc-container CONTAINER=$CONTAINER; done" [13:29] Then you have lxc-container.conf which has a body referring to $CONTAINER [13:39] twb: Yup, that's what I'd do. [13:40] twb: (remember the "instance $CONTAINER" line) [13:40] twb: Otherwise it won't know to key off that variable. [13:42] Thanks. [13:43] soren: do you know offhand how to match "all filesystems are up and all interfaces are up"? [13:44] twb: There is no such notion. [13:44] twb: There never was. [13:45] twb: How will anyone know if I'm going to stick another USB disk or NIC in the box in two minutes? [13:45] I don't want to say e.g. "start when br-lan is up", because some containers will need a different interface [13:46] soren: OK, the debian lxc LSB hader I'm translating has Required-Start and Required-Stop of $syslog $remote_fs [13:52] My gods. That actually seemed to work [13:58] soren: care to sanity-check it? http://paste.debian.net/100036/ === zul is now known as ep === ep is now known as zul [14:01] twb: Is the space before .conf here intentional: do start lxc-container CONTAINER="$(basename "$CONTAINER") .conf" [14:01] Yes; it changes /etc/lxc/foo.conf to just foo [14:02] Um... No. [14:02] If it were inside the parentheses, yes. [14:02] # basename /etc/lxc/foo.conf .conf foo [14:02] Oh, sorry, I se [14:02] *see [14:02] That's not what you're doing, though. [14:03] It's because upstart conf files aren't sh, so Emacs doesn't highlight the ""'s correctly and I don't spot the mistake :-/ [14:03] scratch that, I still would've missed it [14:03] I wouldn't have thought export was necessary. [14:03] ...but other than that, it looks ok. [14:03] Hmm.. [14:04] Well, yeah, it's fine. [14:04] I'm not sure if it'll shutdown cleanly [14:06] Seems to have gone tits-up during reboot [14:14] Now I'm confused [14:14] after a reboot, "status lxc-container CONTAINER=hera" says there's no such instance -- but lxc-ls and lxc-ps can see it. [14:16] If I manually run "start lxc", I can see the "lxc-container (hera) start/running" line on tty1 [14:16] But "status lxc-container CONTAINER=hera" still fails. [14:16] THEN, I manually run "start lxc-container CONTAINER=hera", and it *does* start, I can see it on tty1 and I can see its processes [14:26] Anyone that can help me, have been strugeling with this for 3h straight. [14:26] make -C /home/eriksson25/s2-liplianin/v4l [14:26] make[1]: Entering directory `/home/eriksson25/s2-liplianin/v4l' [14:26] perl scripts/make_config_compat.pl /lib/modules/2.6.35-22-generic/build ./.myconfig ./config-compat.h [14:26] File not found: /lib/modules/2.6.35-22-generic/build/include/linux/netdevice.h at scripts/make_config_compat.pl line 15. [14:26] make[1]: *** [config-compat.h] Error 2 [14:26] make[1]: Leaving directory `/home/eriksson25/s2-liplianin/v4l' [14:26] make: *** [all] Error 2 [14:27] twb: one thing that came out of UDS-n was that auto-starting of containers (and bridges for them) ought to be handled by libvirt. (which means we first need to write a driver to have libvirt use contaienrs, but...) [14:27] wait i thought there was one already/ [14:28] hallyn: I have yet to find a compelling reason to install libvirt [14:28] zul: (assuming that was to me) no, you can pretend to some degree, but the contaienrs are different [14:28] Note: libvirt already knows how to use LXC containers, directly, not using lxc-*(8). [14:29] mysql, after upgrade, fails to start, no socket created at /var/run/mysqld/. Error 2002 when trying ~$ mysql. my.cnf appears vanilla, has correct socket =. [14:29] the libvirt contaienrs are very limited (and shoudln't exist in that form) [14:29] How do I further trouble shoot mysql? [14:30] hallyn: OK, you've encouraged me not to use libvirt for containers in lucid :-) [14:30] twb: to have libvirt work with lxc containers I had to do http://s3hh.wordpress.com/2010/09/07/easier-creation-of-libvirt-lxc-containers/. [14:31] twb: my original thought had been lets have an upstart script for lxc. the consensus at the talk was 'that's what libvirt does' [14:31] hallyn: heh maybe i should bookmark that ;) [14:32] hallyn: but that won't happen until the next LTS for me, so basically I'll ignore it [14:33] hallyn: "(Mind you, shutdown from inside a container doesn’t work – you have to just kill init from the outside.)" [14:34] hallyn: did you solve that? [14:34] 08:32 < twb> hallyn: but that won't happen until the next LTS for me, so basically I'll ignore it [14:34] yup [14:34] I'm getting the same behaviour without libvirt [14:35] i.e. text-only dom0, text-only domu, run "lxc-start -n foo", log in and run "halt" -- it just sits there. [14:35] But "lxc-stop -n foo" works from a second shell on the dom0 [14:35] regarding shutdown, not solved for libvirt. lxc claims solve it somehow through the parent of the init [14:36] 'halt' works for me in lxc containers [14:36] oh, but im' on maverick [14:36] hallyn: inlxc 0.6 or 0.7? [14:36] yeah i don't really use 0.6 much [14:36] OK, I think that's something lxc fixed post-lucid. I'll try backporting maverick's lxc. [14:37] Hi, Anyway to remove linux headers, and source, and reinstall it all? How would I type then? [14:37] b0badabd2d3ec9c8506651bbb4900cc0ec3f8a16 [14:37] support shutdown/reboot with upstart within a system container [14:38] eriksson25: "aptitude reinstall "? [14:38] SpamapS: ping...lemme know when you are around [14:39] twb: when you get your upstart script tuned, can you send it to the lxc or contaienrs list? [14:40] hallyn: I might just give up and use the debian sysvinit script (which starts them asynchronously, and stops them synchronously). [14:40] Ping me about it next week, by which time I should have decided what to do [14:40] twb: will do, thanks [14:41] Where "next week" means, like, eight days from now, not next monday :0) [14:42] hallyn: btw, do you know how to define a container that has >1 virtual interface? [14:44] twb: pretty sure you just list two in your .conf file [14:45] How does it know where each stanzas ends? [14:45] never mind, I'll test it out sometime [14:45] the next 'lxc.network.type' i believe. [14:47] New bug: #676503 in openldap (main) "slapd 2.4.21 memory leak in syncprov" [Undecided,New] https://launchpad.net/bugs/676503 [14:47] Oooh. I had forgotten that "lxc-stop" is a hard-stop [14:49] I will need to fix the shutdown script do something like "timeout 2m lxc-execute -n $container telinit 0 || lxc-stop -n $container" [14:57] Is there a way to make cpufreq work on 10.04? [14:57] I seem to mis a few modules [15:00] a_ok: just set the scaling governor at boot time, then leave it alone [15:01] twb: how do I do that? [15:01] $ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor ==> ondemand [15:01] If you get that, you're done -- you have nothing to do [15:01] ah thanks! [15:02] twb: no such file... [15:03] the cpufreq dir does not even exist [15:03] OK, please pastebin the output of "lsmod" and "cat /proc/cpuinfo" [15:04] FWIW, I have a bog standard 10.04 install and it has picked the ondemand governor without any input from me. [15:05] If you're not using an Intel amd64 processor it might have a harder time... [15:11] twb: I don't have any cpufreq modules here (i mean not to be found anywhere) so that should explain it [15:12] "cpufreq" isn't a module [15:14] twb: on my 8.04 servers I have cpufreq_conservative.ko cpufreq_ondemand.ko cpufreq_powersave.ko cpufreq_stats.ko and cpufreq_userspace.ko [15:14] do you have those? [15:15] Well, sure. Those are the governors. [15:15] You also need a driver, like acpi-thingy or p4-clockmod [15:15] acpi-cpufreq [15:16] hi guys [15:16] I have an error with this script /usr/share/dtc/admin/cron.php [15:16] PHP Warning: Module imap already loaded in Unknown on line 0 [15:16] it is emailing me every 10 minutes, lol [15:16] I cant find the problem :( [15:17] * cwillu_at_work fails to refrain from pointing out that your problem is php [15:17] cwillu_at_work: probably, but I didnt install it, isnt dtc part of ubuntu? [15:18] !info dtc [15:18] Package dtc does not exist in maverick [15:18] cwillu_at_work: I have an infobot entry "most problems can be fixed by installing Debian". I should write another one: "most web problems can be fixed by removing PHP" [15:18] twb: http://pastebin.com/yVPVZVrD [15:18] "Most problems can be fixed by installing Debian. The remainder are fixed by removing PHP" [15:18] as you can see est is in there [15:19] cwillu_at_work: well, no. The implication of the former is that the other problems are the ones introduced by Debian [15:19] Like, say, "I can't play crysis anymore" [15:19] this is a problem? [15:19] twb: well as I don't have any goveners at all or acpi-cpufreq (p4-clockmod is presend but won't load manually) id say this kernel does not support it... [15:20] a_ok: "dmesg -c >/root/dmesg~; modprobe acpi_cpufreq; dmesg" [15:20] p4-clockmod is, unsurprisingly, for Pentium 4s [15:21] An E5520 is generationally about at acpi_cpufreq.ko [15:21] FATAL: Module acpi_cpufreq not found. [15:21] a_ok: well, WTF [15:21] a_ok: is this a stock 10.04 kernel? [15:21] twb: yep the server one that is [15:21] I give up, sorry [15:22] ill try to see if there is an update [15:22] 2.6.32-25-server x86_64 here, acpi_cpufreq exists and is in use [15:23] Sorry, I tell a lie. acpi_cpufreq doesn't exist, I *do* have working ondemand governor [15:24] acpi_cpufreq is apparently compiled in (i.e. =y) in Ubuntu kernels [15:24] a_ok: read through dmesg~, see if anything jumps out at you [15:25] twb: do I still need to install cpufreq-utils? [15:25] No [15:26] userspace packages have been obsolete for years [15:26] lol now they tell me [15:28] Sorry, I should've mentioned that at the start [15:29] Technically they're still useful if you want to switch from "ondemand" or "powersave" (on battery) to "performance" (on AC), but in practice you get that effect by just leaving it on "ondemand" [15:33] twb: yeah I want it on ondemand. I guess that 5520 is to new for the driver or something [15:33] dunno [15:33] i wouldn't have thought so [15:33] well nothing is loaded that is for sure [15:34] Hm, my Q9550 doesn't have it [15:34] ? [15:34] does not have what? [15:34] No /sys/class/devices/cpu [15:34] any ideas why the following iscsi tgt targets.conf will not work in an exported target? Setting it up per command line is fine: http://pastebin.ubuntu.com/533489/ [15:35] Oops, that's an 8.04+openvz kernel [15:35] twb: you mentioned another dir by the way [15:35] /sys/devices/system/cpu/cpu0 [15:35] In 8.04+openvz I can modprobe acpi_cpufreq and get a governor [15:35] a_ok: cpu0 is just the first cpu (or core) [15:36] yeah I can do the same on my 8.04 [15:36] twb: I mean you are looking at class now [15:36] a_ok: same shit [15:36] /sys is incestuous [15:36] ow [15:37] it's 2:40AM, I'm going home [15:37] thanks for the help [15:37] Ill file a repport or something [16:05] Hi [16:06] Are there certifications for Individuals on Ubuntu Server similar in the lines of RHCE ? [16:06] New bug: #676544 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,New] https://launchpad.net/bugs/676544 [16:17] running ubuntu server... i have three servers. server1 can ping server2 and server3 fine. the ping runs fast, and has low ping time. [16:18] server2 and server3, when they ping each other, or the server1, the ping time is low, but when i press enter to ping, it can take up to 3 seconds for the ping result to show. but it is still 0.1ms... [16:18] any ideas why the ping command would be so sluggish, but still return acceptable ping times? [16:19] talntid: sluggish name resolution [16:19] talntid: check the name servers in your /etc/resolv.conf [16:21] they match the fast server, except.... === SirFiChi is now known as ihCiFriS [16:21] kaushal: the Ubuntu Certified Professional certification was suspended, currently there's nothing like RHCE. You can still do the Ubuntu Professional course though [16:21] 10.21.2.21 is listed before 10.21.1.21 on the sluggish machines, and on the speedy one, 1.21 is listed first [16:23] talntid: test them with `dig` on each machine [16:23] actually, 10.21.1.21 isn't pingable...... [16:23] talntid: so? icmp echo replies can be filtered [16:24] right, but it's not... it was -down- [16:24] now the machine is up [16:24] ping is fast now. [16:24] =] [16:24] thanks! [16:34] I'm running Ubuntu Server 10.04 LTS on a DL-380, serving primarily as Apache+php+mysql server and I've suffered recurring failures... about once a month, the webserver would run and serve some pages but I wouldn't be able to access the file system or the database and eventually the whole thing would come to a grinding halt... after a hard reboot it would work fine and I'd find no clue other than that syslog stopped logging a few days before [16:35] I have the same problem at this very moment... I can ssh, list processes, kill processes and some other tasks... but shell will freeze completely on commands such as 'top' and 'ls /' [16:35] does anyone have any clue what it could be or what I could try to test, right now, to find out more? [16:40] teep_: regular harddrives? hw raid? or san? [16:40] teep_: see bug 407862 for the syslog. About the rest: have you tried logging in with another username? I have the same problem when encountering bug 635181. Logging in with another user works, and then I can reboot the system. [16:40] Launchpad bug 407862 in rsyslog "Messages not being sent to system logs" [Undecided,Confirmed] https://launchpad.net/bugs/407862 [16:40] Launchpad bug 635181 in linux "High load average on Lucid for nominal/idle system use" [Undecided,New] https://launchpad.net/bugs/635181 [16:40] zul: pong I'm here now, wassup? [16:40] hw raid [16:41] SpamapS: we are behind in the merges wanna divide up the list and get cracking? [16:41] zul: sure, I'm still working on php [16:41] I can log in at the moment and I think I can do a normal reboot... but I'd rather find out what's up [16:41] SpamapS: really? [16:41] zul: well I had to stop to get my specs done [16:42] SpamapS: heh [16:42] teep_: hard drives full (df)? [16:43] zul: is there any point in sending stuff back to debian while they're still frozen? [16:44] no, checked that, hard drives not full [16:44] SpamapS: not really...just queue it up [16:44] zul: I mean I guess its good to get the stuff into BTS [16:44] I was killing syslog processes to see if they locked things up but that didn't help [16:45] interestingly, shutting down apache with /etc/init.d/apache2 stop worked fine, but restarting it with start threw an error that the socket is already in use [16:47] SpamapS: As long as you give it an appropriate priority (like the DSO linking stuff is still wishlist in Debian), it's good to go ahead. [16:49] commands like w and netstat cause the shell to freeze [16:49] zul: I'm still wrapping up my last spec, if you want to pick a few merges I'll pick whats left and get started in about an hour. [16:49] and I notice now that doing a ps, those commands aren't listed [16:49] teep_, hows memory usage? [16:49] SpamapS: [16:49] SpamapS: sure [16:50] talntid: how can I see this without running top? [16:50] free [16:50] SpamapS: but first i must eat [16:51] free causes the shell to freeze [16:51] i'm thinking your memory is full. memory leak maybe... but likely. it's full [16:52] cat /proc/meminfo [16:52] talntid: I can still open new shells though... and if I don't use certain commands, the system appears quite responsive [16:52] can anyone reccomend a good hosting company in DE? [16:53] talntid: ah, cat /proc/meminfo works and lists 6GB free out of 10GB [16:53] ok.. hmm.. [16:55] cat /proc/loadavg yields 40.98 40.19 38.36 ... that seems a bit high? [16:56] very high [16:56] teep_: Nice. [16:57] now how to determine which process is running amock, without the help of top? [16:57] and without ps [16:57] lsof? [16:58] see what files are open? [16:58] lsof causes a freeze, alas [16:59] indeed [16:59] I am attempting to patch openssh to add the LPK patch.. but the process is failing and I have a question about some things I am finding in the rej files [16:59] http://pastebin.ca/1994725 [17:00] to me it looks like (among other things) am missing some libraries [17:00] shouldn't the ubuntu equivalent of red hat yum whatprovides be apt-cache search to find these missing libaries? [17:01] brb === jdstrand is now known as jdstrand_ [17:11] teep_: That's an enormous load. Can you run iotop? [17:11] alvin, nope, iotop freezes [17:12] I notice now that apache2 never really stopped... trying to kill it [17:13] there is no pid file for apache2 [17:16] killall apache2 [17:17] killall freezes [17:19] funny thing is... doing a kill -9 on an individual apache2 process doesn't seem to respawn it... instead, it just keeps running [17:25] is there any ubuntu specific version of the openssh source? [17:27] right... I'm off for now... alvin , talntid and the rest... thanks for your help so far! [17:28] I am attempting to patch it on 10.10 but dpkg is complaining about missing the debian/changelog [17:29] hey people not sure if you're the right groupt o ask this or not but I need an open source alternative to active directory [17:29] burntoutlamp, LDAP [17:29] -.- anything easier than manually configging LDAP and Kerberos etc? [17:29] apps perhaps similar to AD? [17:30] configuring LDAP and kerberos IS easy ;) [17:30] for you lol I am a n00b [17:30] best way to become a non n00b is to practice my friend .. heh.. persistence pays [17:30] doesn't samba3 have PDC support that is a bit more noob friendly? [17:31] I think it does [17:31] hi [17:31] or was that samba4 [17:31] hu spike [17:31] bluethundr, to be honest I am on a time constraint as well. I would love to take the time to do that but I am somewhat crunched. [17:32] hmm ok.. not sure what to recommend other than LDAP.. sorry [17:32] I've deboostrapped lucid and bundled it for ec2 with --kernel $aki . the instance boots up and on getconsole I can see a login prompt but it seems to have done nothing else [17:32] anyone have any idea on best way to deal with the lack of debian/changelog in patching openssh? [17:32] ie it doesn't seem like it even tried to start the network or any daemon [17:32] bluethundr, thanks dude [17:33] burntoutlamp, no prob [17:33] any idea what it might be? [17:34] I did find 389 dir server but it's fedora :< [17:35] 389 indicates LDAP does it not? is this some slick simplification of LDAP of which I am unaware? [17:35] Hi all [17:37] maybe I dunno? like I said I am new. It is LDAP but I think makes it easy for people like me. I think I just might have stumbled upon an Apache solution also [17:37] apache directory server [17:37] I have problem using this: smbmount //192.168.0.1/ti /home/pc/net -o username=test,password=test => mount error(1): Operation not permitted. Refer to the mount.cifs(8) manual page (e.g. man mount.cifs) [17:38] But if I execute that as superuser works.. [17:38] so if anyone else is interested I managed to just stumble upon that. and the site osalt.com in case anyone is not aware. this is great [17:38] But I need to mount with mormal user.. How I do? [17:38] does anyone know where I can download karmic koala 9.10 server off the top of their head? [17:39] nevermind :) http://releases.ubuntu.com/karmic/ [17:40] Daviey: ping [17:41] zul: o/ [17:41] Daviey: umm...did you review the merge? [17:42] zul: i branched it.... === NG_ is now known as ng_ [17:42] zul: started looking :) [17:42] Daviey: uh huh :) [17:44] anyone? [17:49] mathiaz: ping [17:49] zul: o/ [17:49] mathiaz: so puppet in natty right now is a rc mind if we update it? [17:50] zul: about to upload the final release [17:50] zul: ie 2.6.3 [17:50] mathiaz: excelente thanks! [17:51] zul: branch looks awesome too me! [17:51] Daviey: cool ill upload it then [17:51] zul: sorry for the delay... really, i am [17:51] Daviey: riiiiight ;) [17:54] hello all [17:54] question: i have a Dell PE 6800, is there a way for me to install DRAC on it? [17:54] specifically, the DRAC gui would be nice [17:55] isn't the whole point of drac is remote management? not local? [17:55] yes, but i'm not sure how to access it [17:56] (When you buy your server from an auction for 140 bux, documentation isnt really a big priority for the seller :-/) [17:56] right now i can mess with IPMI a little bit through 'ipmitool' [17:56] it probably doesn't have real drac then, just drac-express [17:56] no, the thing has a full blown DRAC if i remember correctly [17:56] *2600, sorry [17:57] well actually [17:57] how do i check this? [17:57] lspci? [17:57] zul / mathiaz: Should we look to unify the debian and ubuntu puppet packages? [17:58] Daviey: in the long run yes imho [17:58] What is everyone using to set services to start at boot? Is chkconfig taking hold? [17:58] zul: If you are looking to do some work on puppet, would it be worth discussing it with the DM? We are currently on a higher version than they are [17:58] init.d [17:58] oh yuk [17:59] pe2600 + drac is pretty much crap [17:59] :-/ [17:59] it requires os support to do anything at all [17:59] aah shit [17:59] http://support.dell.com/support/edocs/software/smdrac3/RAC/en/is/index.htm [17:59] stupid windows [17:59] Daviey: yeah when debian unfreezes probably :) [17:59] well, at least it makes a hell of a samba server for 140 bux :) [17:59] zul: i thought it had now thawed? [18:00] actually, patdk i have a DRAC card at the ofic [18:00] Daviey: i thought it was still forzen [18:00] *office [18:00] for a 2650 [18:00] its probably plug-and-play compatable, yes? [18:01] no idea [18:01] I have never used drac [18:01] I use ilo2 a lot though [18:01] whassat? [18:01] right now it would be great to be able to manage hardware features through a GUI of some sort on my 2600. [18:02] it seems that DRAC does that, but if theres something else, then i'm all for it [18:02] zul: Unstable has had lotta uploads recently [18:02] right now my only problem with the server whatsoever is that i'm getting correctable ECC memory errors that i have to clear [18:05] Daviey: meh.... [18:11] Daviey: we're already in sync [18:11] Daviey: I'm working from the debian maintainer tree [18:11] mathiaz: rly? [18:11] Daviey: http://git.debian.org/?p=pkg-puppet/puppet.git;a=summary [18:12] mathiaz: ahh, rather than syncing? [18:12] Daviey: natty and maverick branches are already hosted on alioth [18:12] Daviey: because debian is frozen, the latest version of puppet is not in unstable [18:12] Daviey: and the debian maintainers are focusing on squeeze [18:13] Daviey: but we're already in sync [18:13] Daviey: ie all the ubuntu changes are in the master branch [18:13] mathiaz: Ah, ok - that sounds great [18:14] mathiaz: Is there an ETA for an updated Unstable package? [18:14] Daviey: nope [18:14] Daviey: and unstable is currently used for getting packages into squeeze [18:14] Daviey: so as long as squeeze is not released there won't be any new packages uploaded to unstable that are syncable [18:14] mathiaz: experimental ? [18:15] Daviey: yes - that would be the case [18:15] Daviey: however the master is not uploaded to experimental [18:15] mathiaz: are you one of the DM's? [18:15] Daviey: no [18:15] Daviey: I only have commit access to the git repo [18:16] JamesPage: what would we need to run Hudson now -- would, for example, http://cloudbees.com/ work? [18:16] mathiaz: zul seems to be keen to update the package to Final... if he pushes you a diff, can you sponsor it into debian git tree? :) [18:16] Daviey: I'm about to upload the latest release to natty [18:16] Daviey: the work is already done [18:17] mathiaz: hmm... the debian git tree still looks like rc3? [18:17] Daviey: yes - because I haven't pushed it yet [18:17] ahh! [18:18] oh great... i'll shut up then :) === luis__lopez is now known as luis_lopez [18:22] hey.. anyone can help me with mount samba shared? [18:22] I have problem using this: smbmount //192.168.0.1/ti /home/pc/net -o username=test,password=test => mount error(1): Operation not permitted. Refer to the mount.cifs(8) manual page (e.g. man mount.cifs) [18:23] with super user works [18:36] Hiya. [18:38] kirkland: hey, approx is pretty nice.. :) [18:38] kirkland: a lot easier than trying to make sure squid-deb-proxy-client gets installed. :) [18:40] SpamapS: But you still have an SRU to finish. [18:41] ScottK: I have so .. so much to do. ;) Which one? the ipvsadm sru? The mongodb sru? :-P [18:41] SpamapS: Wasn't there one for squid-deb-proxy? [18:42] ScottK: yes, there is... I suppose I should throw a branch up with "the solution" rather than just pointing out the problem. :) [18:42] Yes. [18:43] I actually think for a simple LAN setup, pre-seeding squid-deb-proxy-client is probably easier than trying to override DNS... [18:44] * ScottK doesn't have an opinion on the implementation, just that it ought to get fixed. === oubiwann is now known as oubiwann-away [18:46] Indeed, its severely broken as it stands right now. === oubiwann-away is now known as oubiwann === zul_ is now known as zul [18:49] * SpamapS wonders if somebody has already written a command like 'ec2build' that does your build on a fresh ec2/euca instance... [18:53] SpamapS: yes its called vmbuilder :) [18:54] \/win 11 [18:54] arrrgghhh [18:56] zul: err, huh? [18:56] zul: vmbuilder makes me a VM ... I want it to spawn a VM, copy the source package up, and build it. [18:57] SpamapS: vmbuilder makes your euca/ec2 iimages for you [18:57] zul: no I want to use the stock ec2 image [18:58] like.. ec2build -d maverick-amd64-c1.medium php5_blah.src [18:58] SpamapS: ah ok...thats different [18:58] yeah... its that or buy a cheap 4 core box for doing builds here at home [19:02] SpamapS: you might be iobound rather than cpubound when you are using the ec2 instances to build though [19:03] but cheap virtualized (fake) 4 core box [19:05] Riddell recently built all of KDE pretty quickly on EC2. [19:09] zul: but I will not be killing my machine, which is the main goal. :) [19:09] any reason that my 10.04 servers would be periodically contacting bignay|auckland|atemoya.canonical.com? it's not updates, as I have a local apt-proxy [19:10] wiredfool: did your servers come pre-installed w/ Ubuntu by any chance? [19:10] they're virtual machines [19:11] so, no. [19:11] wiredfool: ah, then probably just the root kit canonical puts in the regular builds phoning home waiting for world domination orders. ;) [19:11] I'm sure that's what an auditor would think [19:12] wiredfool: whether or not you have a local apt-proxy, it may be that apt isn't using it all the time. [19:12] auditors think what they're paid to think [19:12] hmm. maybe that's a dist-upgrade check [19:13] NTP defaults to using ntp.ubuntu.com. Not sure if that relates. === xfaf is now known as zul [19:13] grrr.... [19:13] * wiredfool has a supershort sources.list -- three lines. [19:13] port 80 [19:13] That's a good posibility too [19:13] SpamapS: thats why you should have a second machine [19:14] wiredfool: I'd grab it with tcpdump and look rather than speculate. [19:14] zul: I do.. "in the cloud!" [19:15] scottk: yeah. I was hoping that there was a quick answer [19:15] SpamapS: meh....im not going to argue about this right now ;) [19:15] zul: I do builds all day on ec2 c1.medium's .. very quick and costs about $0.10 [19:15] zul: much cheaper than a second machine, and no power usage incurred at Los Angeles's ridiculous rates. [19:15] SpamapS: yes but i dont want to give amazon 10cents everytime i do something [19:16] zul: I can reasonably build a 4 core box with a few cheap SATA disks for $250 .. thats a lot of $0.10 builds. ;) === 15SAA4B7R is now known as JJman6_ [19:22] hello, when I logging to one of my ubuntu server with ssh, I get this nice welcom screen with the message about the current load and how many packages can be upgraded [19:22] but on my 2nd ubuntu server I don't see anything, where can I turn that on? [19:25] Aison, /etc/update-motd.d/ I believe [19:28] Server is now web page serving! :D yay beyond-sight.com [19:28] the php is probably coded wrong, I need to fix the path links [19:28] anyone here? [19:28] the php is always coded wrong [19:29] poorly handled pages [19:29] or something [19:29] I like the black-on-black though [19:29] well [19:29] in this case [19:30] I've had to redo the php every time I put the site on another server [19:30] it's jst how it handles directories [19:30] It looks really nice when it's working [19:31] the navigation menu is distributed nicely among the leaves [19:31] and the links change colors to match the color scheme [19:33] Xalem: I need to configure the software firewall, mail server, printer server [19:36] ZacharyNewb: you can help yourself a lot by never ever using local files in your code.. thats sooo 2001 [19:37] SpamapS: lol, what am I supposed to do? link to other sites? or link to my own site? [19:37] I can't trust other sites to keep the files up [19:42] ZacharyNewb: if you're talking about includes, there's a path for that [19:46] having trouble killing a process... [19:46] tried kill, kill -9... it's still there.. it's firefox-bin.. any ideas? [19:53] talntid: sudo apt-get autoremove firefox ? [19:53] talntid: uninstalls it [19:53] sudo apt-get remove firefox [19:53] sudo apt-get autoremove firefox [19:53] sudo apt-get purge firefox [19:56] it will still be in memory [19:56] so long as its running [19:59] talntid: It's also not a very server related issue. [20:08] . [20:19] I need some help setting up the ubuntu print server? [20:31] New bug: #676672 in php5 (main) "FTBFS on natty: SASL LDAP check fails (and other DSO problems)" [Undecided,In progress] https://launchpad.net/bugs/676672 [20:39] SpamapS: https://lists.ubuntu.com/archives/ubuntu-devel/2010-November/031991.html [20:43] zul: indeed, working on it how with php5 [20:44] zul: its actually dumber than that, php actually thought sasl_version *should* be in -lldap .. even though it has always been a direct part of sasl2 [20:44] hehe [20:44] reason 1001 to hate php [20:44] zul: maintaining PHP *will* teach you how to work around things done "the wrong way" [20:45] SpamapS: heh....im out of here l [20:45] zul: Yeah I have to run some errands.. will hopefully wrap up php merge and ftbfs later today or tomorrow then get on top of those other merges [20:46] SpamapS: cool beans === ng_ is now known as NG_ === NG_ is now known as ng_ [21:32] Hi. i can't ssh to my server, but I can ping it. Also apache doesn't seem to respond. What's happening? [21:34] PeterNL: no services running? [21:34] There should be loots of things running [21:35] PeterNL: and please don't crosspost [21:35] Sorry, but I thought there was noone here [21:35] I take it you cannot get to the server? [21:35] )in person) [21:36] I'm sitting (almost) right next to it [21:36] But theres no way I can connect a monitor or a keyboard [21:36] thinking... [21:38] PeterNL: test it with nmap [21:38] ok [21:38] Lots of ports are open, as they should be [21:39] (exactly the ones I expexted to be open) [21:41] PeterNL: then you could try to reach other services... (just to see if it is alive at all - I sometimes had such failures with an old suse server, there only a manual reboot would help) [21:42] ...nothing [21:43] But I don't want to reboot it. The PSU fan has some problems with spinning up... Really... [21:43] PeterNL: with only a keyboard you could reboot it without powering down... [21:44] I should try that. But the last time I booted it there was no keyboard present. I don't have any USB keyboards lying around, and PS/2 isn't hot swappable [21:47] PeterNL: well, I'm at the end of ideas then - sorry - but if it is an old machine it might have a reset button... [21:47] It might... [21:47] * PeterNL graps a flashlight [21:47] grabs* [21:48] ...there goes my almost 300-days uptime... [21:48] Wait [21:48] I might have a usb keyboard after all... [21:48] ...maybe... [21:58] Anyone in that can help me, updated to kernel 2.6.37 to try and fix one thing but didnt work. Now I want to reinstall the original 10.10 kernel. 2.6.35.22-generic [21:59] reboot and select the proper kernel in grub? [21:59] I followed this steps and it was realy easy to change. But where do I finde the generic. [21:59] http://ubuntuforums.org/showthread.php?p=10112818 [22:00] eriksson25: server on a notebook? [22:00] I want to reinstall it, since It was giving me alot of errors that I didnt get with a new installed .37 [22:00] server [22:01] Trying to install mantis drivers, installed with dkms but didnt work so removed them and that made my system throwing me alot of weird errors. [22:03] could someone help me with the ubuntu server print server? CUPS? [22:03] eriksson25: sorry, no help from me there, I only wondered because the .37 was supposed to fix a wireless issue with iwg4395... [22:03] ZacharyNewb: as always, ask a questiin, don't ask to ask [22:03] mm, it was realy easy to setup [22:04] But didnt help me =/ [22:04] eriksson25: Thank YOU for your help yesterday. beyond-sight.com is now page serving thanks to you. ;) [22:05] RoyK: Well, I followed some tutorial, which described configuring the CUPS to listen on a certain port so I could configure it on a webpage over the network [22:05] I just did connect a monitor to my server, and it says "[timestamp] BUG: soft lockup - CPU#0 stuck for 11s! [apache2:PID]" [22:06] No problem, my turn to sith with a problem. Just a litle more advance and hard to explain. [22:07] ...and it doesn't respond to any key presses === _TechAway_ is now known as _Techie_ [22:08] RoyK: Current problem, is I don't know why cups doesn't seem to be working on my server. I've rebooted it (managing it through putty on a windows 7 netbook) set which port to use for the http:631 configuration, but the web configuration page doesn't show up when I try to access it over the network [22:09] zul: there? [22:09] smoser: there? [22:11] New bug: #676714 in euca2ools (main) "pull euca2ools 1.3.1 into natty" [Undecided,New] https://launchpad.net/bugs/676714 [22:12] RoyK: eriksson25 I've reinstalled the CUPs with sudo apt-get CUPS [22:12] RoyK: eriksson25 still doesn't seem work. [22:15] SpamapS: re https://code.launchpad.net/~clint-fewbar/ubuntu/natty/cheetah/natty-merge-with-debian/+merge/39881 [22:16] SpamapS: are planning to go back to the merge before the EOW? [22:21] I dont know anything about cups sorry. [22:21] But google is your friend. [22:24] maybe dpkg-reconfigure cups [22:37] hallyn: merging kvm now ... sorry for the delay [22:37] kirkland: np, thanks [22:37] hey does anyone know where the heck to get the openldap HEAD ? [22:38] Would anyone have some tips getting ubuntu server to install from a usb drive? [22:40] Easy, read on the ubuntu.com [22:40] ah there it is [22:40] hallyn: Are they still on cvs? [22:40] eriksson25: Oh, I've tried that. Unfortunately I've run into some problems. [22:41] hallyn: You could just use the bzr mirror on Launchpad. [22:41] Then say them. Hard for us to know what errors you have had. [22:41] soren: no i want to see the very latest [22:41] hallyn: I think that's HEAD. [22:41] http://www.openldap.org/devel/cvsweb.cgi [22:41] so i guess i'ts still cvs :) [22:41] no wonder googling for svn openldap was getting me nowhere [22:42] Aye, sorry. Basically, I'm trying to get it to install on an eeebox, and it tells me it can't find the cd-rom. Of course, the machine doesn't have a cd-rom drive and it's trying to install from a usb stick... [22:42] soren: did you get a chance to look at that live iso? [22:42] kirkland: Not at all, sorry. [22:42] Well, that is a boot problem in the box, have you set it to boot from the usb? [22:42] In my asus nootbook it is shown as a hdd. [22:42] Yeah, I've done that. And it finds the usb just fine, since it can start and run the installer. [22:43] It just fails partway through saying it can't copy the files from the cdrom. [22:43] soren: bummer; any chance you could boot that puppy in a kvm and see if it's something silly/easy I'm missing? [22:43] soren: give it 10 minutes or so attention? [22:43] And you did the usb with the usb creater software? [22:43] (as an aside, I was able to install the desktop version no problem with the same method, but then I realized I need the server edition) [22:43] I tried making it with the usb creator software, as well as unetbootn. [22:43] Why did you need it? [22:44] server eddi. [22:44] Been searching around various websites and tried everything for a few hours now. =) [22:44] Because I want to use the machine as a webserver? [22:44] Well, no problem, install desktop eddi, remove genome, and instal LAMP and you are done. [22:45] You got some more stuf there that is not needed but you wount notice. I run my server as desktop version for 3 years. Before I did a reinstall. [22:46] I'll have a look at that then... I guess I need it to accept my network card properly before I can get LAMP installed, though? [22:46] Well, yes to dl stuf its needed. And to be a web server its surly needed. [22:46] Yeah, I'll have to agree with that =) [22:47] =) [22:47] It seems mostly to have problems with the DHCP rather than the network card itself, at least. [22:47] Do it, and skip serching for a loong time. [22:47] Well, turn that of and put static ip on it then. [22:47] If I manually set the IP, it'll find and connect to my old webserver, but it won't find the gateway, nor will it get out to the 'net. So something seems a little wonky there. [22:48] (I'm basically trying to move over from my old Pentium-2 which is sounding like a lawnmower atm, to the new machine) [22:48] Thats becouse you havent set a static dns and gatway. [22:48] Aye, sounds likely. [22:48] Where would I do that? [22:49] hmm, dont remember on the top of my head. But google static dns ubuntu and you will find it. [22:50] Okay, thanks a lot for the help. =) [22:57] hallyn: done === ng_ is now known as NG_ [23:45] is there an automatic account management tool in ubuntu that is similar to auth-config under red hat that would allow automatic configuration of pam to do ldap lookups for it's information? [23:50] bluethundr: there's pam-auth-update [23:50] mdeslaur, that's great thanks! [23:51] bluethundr: not sure if it'll do what you're looking for, but take a look [23:51] hoping to get all of pam to listen to an ldap server [23:51] I will look into it [23:56] mdeslaur, yes I gave it a try, but no luck [23:58] at any rate, here is how I tried to configure my /etc/pam.d/common-auth file: http://pastebin.ca/1995092 [23:58] as an example of how I tried to set this up