[00:05] hey all [00:31] heyy [00:34] wizardslovak: What's up? [00:36] Has anyone here taken an LPI certification test recently? [00:44] ssd7: not much , sup with you? [00:46] nm [01:13] is here anyone who can help me settup email server? [01:16] wizardslovak, gl, mail is EVIL!!! [01:16] andol, u still around? [01:17] i already heard it somewhere [01:17] heard what? [01:18] mail is a pita? [01:28] wizardslovak: what kind of help do you need [01:37] how do i change this?, mailed-by chr1831.xen.....com i wanna make it just my domain name [01:40] zoopster: well someone who can guide me true installation [01:41] FFForever, it depends on which smtp mailer you are using, but its called a mail server banner, you can google the smtp dameon you are using along with smtp banner to figure out how to change it [01:41] thanks, i am using sendmail-bin [01:41] like what mx priority should i set in my domain name ? [01:42] FFForever, sorry, laptop battery going to die...gl with it [01:42] thanks [01:42] foxbuntu: do you just mean the foo in HELO foo? [01:43] zoopster: when i set up in my domain name register site , i have record type "mx"mx priority "10"host name "@" and my ip [01:43] zoopster: but i am getting error" * Forward Address is not RFC compliant." [01:45] if i have a mta can anyone from the world wide web use my send mail to send messages? [01:45] (i should only have the sendmail-bin and sendmail-cf installed) [01:47] FFForever: that depends on how it is configured. [01:47] twb, default configuration? [01:48] I do not use sendmail, and I recommend you don't, either. [01:48] twb, when i telnet ip 25 i get connection refused so i should be fine right? [01:48] FFForever: that depends on how the connection is refused. [01:48] twb, php uses send mail for sending emails [01:48] No, PHP uses sendmail(8). This is a CLI provided by all MTAs. [01:48] sendmail(8)? [01:49] The sendmail binary documented in section 8 of the manual. [01:49] ahhh well as long as i only have the binary's i should be fine no? [01:49] I cannot answer that question. [01:49] ahhh well sendmail is a service/mta =\ [02:03] Can anyone recommend a good Nobish Ubuntu spin off that will fly on an old AMD64? [02:03] Noobish* === ssd7_ is now known as ssd7 [02:09] BlackMinnow: this channel is about Ubuntu Server, not about Ubuntu derivatives. [02:11] For some roles, a Pentium III would be more than adequate to run Ubuntu Server, let alone anything supporting the AMD64 architecture. [02:12] so i am folowinf this https://help.ubuntu.com/community/PostfixBasicSetupHowto [02:13] i settup everything like this [02:13] but still when i do test it doesnt work [02:24] hmmm [02:42] anyone familiar with this "SCGIMount /RPC2 127.0.0.1:5000 " in regards to an apache config [03:12] wizardslovak: What are you trying to get it to do? [03:18] well i want to make it work for my website [03:18] so my email will be xxx@web.us [03:20] okay [03:20] symptom & config? [04:07] Where did the 'netselect' tool go? [04:08] It seems to only exist in dapper and hardy [04:11] tonyyarusso: what does netselect do? [04:11] twb: Chooses the fastest server from a list, eg for selecting apt mirrors or irc servers. [04:12] tonyyarusso: oh, like apt-spy [04:12] I've never seen one that works on Ubuntu, only Debian. [04:12] never heard of apt-spy [04:12] It's not in Ubuntu [04:13] http://packages.debian.org/apt-spy [04:14] yeah, similar to that [04:14] tonyyarusso: See http://www.mail-archive.com/ubuntu-devel-discuss@lists.ubuntu.com/msg05990.html -- looks like it never really worked on Ubuntu. [04:15] Huh, I wonder what sort of bugginess it had. [04:23] https://answers.launchpad.net/ubuntu/+source/netselect/+question/65575 suggests others are interested too. You cabn always grab the source package, build it in Jaunty and see what happens when you use it :) [04:25] twb: It's wierd that you mentioned that, I've actually been looking at the same thing [04:26] I think it would be pretty easy to write an even better tool than netselect-apt by reusing a lot of code that is in software-sources-gtk [04:27] I've actually spent the last few minutes writing something up and I have something built around netselect that isn't near finished yet [04:27] but once I figure out a good way to determine whether a sources.list entry is one of the official mirrors, it should be pretty easy to finish [04:28] There's a list of mirrors on Launchpad that you could parse. [04:28] ssd7: what apt-spy does is use a list of official mirrors [04:28] That list includes the mirror's URL, and what architectures and categories are supported by that mirror. [04:28] yeah, I have a similar list that I stole from /usr/share/update-manager/ [04:28] apt-spy then finds the fastest one (or few?) and writes them to /etc/apt/sources.list.d/apt-spy [04:29] You could probably just make apt-spy an Ubuntu package by changing where it looks for the list [04:49] well, right now I have a python script that will replace any entry in your sources.list that matches one of the mirrors in the file I have with either (a) a mirror you provide or (b) the fastest mirror from the list I have [04:50] is there way i can check what software is installed on server? [04:52] dpkg -L, dpkg --get-selections [04:55] aptitude search '~i!~M' [04:56] The !~M part suppresses automatically installed package, which you don't normally care about when building additional boxes. [04:56] find / piped to apt-file search! [05:06] huh. That seems to be an aptitude-only functionality. [05:22] any idea how much ram a hl engine uses? [05:31] hl server* [05:40] What's HL? [05:43] tonyyarusso: I'm guessing half life ? [05:43] ah, that could be. [05:43] * tonyyarusso is such a non-gamer he doesn't even think of such things [05:43] with the exceptions of m-b, sol, and pp-r [05:43] :P [05:44] I'm a non-gamer too, but I recognize some of the abbreviations, mostly from seeing reviews of video cards that test with the various games :) [06:01] to install or not to install... that is the question === cgkades_ is now known as cgkades [06:03] cgkades: To install what when on which machine for whom, and why ... that is the *real* question :) [06:08] lol [06:09] jmarsden: i just need a decient secure, robust server. [06:09] installed on a p4 [06:09] i'm trying freeBSD right now, but i've been a fan and user of ubuntu desktop since 6 [06:09] solaris completly failed [06:10] What are you already comfortable with and knowledgeable about securing... if you are used to hardening Ubuntu, run Ubuntu server. If you already know *BSD security, try OpenBSD which claims more security than most... [06:10] i'm used to ubuntu [06:11] not used to hardening anything yet [06:11] i have alot of books though :) [06:11] i have worked with mac os x, wich is bsd... but not as much [06:12] i kind of wanted a challenge, but i also want somethign that is secure [06:12] Then it comes down to who you know... if you have security expert friends who know one but not the other, let that guide your choice, maybe. Failing that... run Ubuntu Server and ask your questions about it here :) [06:12] yeah i dont know anyone, so i'm here in the channels learning what i can [06:13] the guys over in ##freebsd are too 31337 for me [06:13] i like the ubuntu comunity more, so perhaps i'll try that if i cant figure out bsd [06:13] It's a gross oversimplification, but IMO Ubuntu people are often friendlier than *BSD people :) [06:14] * cgkades agrees [06:14] just wondering how secure ubuntu-server is [06:14] i know there was a problem with openssh port for debian [06:17] Which was fixed. [06:18] (Ubuntu (all versions) has zero unpatched vulnerabilities, as reported by Secunia.) [06:20] *known* vulnerabilities [06:20] cgkades: assuming you're referring to the entropy of ssh-keygen, that issue was fixed prior to the last Debian stable release. It also affected Ubuntu, and was fixed there, too. [06:22] tonyyarusso: What about CVE-2009-1337? Haven't seen any USN mentioning it yet? [06:22] andol: The exit_notify function in kernel/exit.c in the Linux kernel before 2.6.30-rc1 does not restrict exit signals when the CAP_KILL capability is held, which allows local users to send an arbitrary signal to a process by running a program that modifies the exit_signal field and then uses an exec system call to launch a setuid application. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2009-1337) [06:23] andol: link? [06:23] tonyyarusso: I belive uvirtbot just supplied it :) [06:23] fine, I'll read... [06:27] 2.6.30 isn't in Ubuntu yet, is it? [06:27] Oh, "before" [06:28] I wonder if it depends on configuration options or anything [06:28] Otherwise why is the Secunia link just for Red Hat's RT kernel? [06:29] tonyyarusso: Yeah, could be. But I don't think it is Red hat specific, concidering there is also a new Debian kernel which fixes it, among other CVEs. [06:29] The development release doesn't have security support. [06:29] Jaunty has 2.6.29 [06:30] err 8 [06:30] 2.6.28.11 no less, which isn't listed in the affected versions on the NVD site. [06:30] (that list stops at .9) [06:31] So it's no suprise there's no USN. [06:34] ScottK: What about supported Ubuntu versions not being 9.04? [06:35] ScottK: By the way, I'm not really saying this is a critical problems, I simply jumped at the number zero mentioned earlier. [06:35] Supported is 6.06, 8.04, 8.10. and 9.04 [06:36] So an issue that only affects 2.6.29 or 30 isn't one that gets a security fix. [06:37] ScottK: What makes you say that this CVE only effect .29 and .30? [06:38] I thought that's what was said earlier in this discussion. [06:38] The CVE itself only lists 2.6.30 [06:39] Generally CVEs list all affected versions. [06:39] ScottK: Are we taling about the same CVE-2009-1337? From what I can see it goes up until .30-rc1. [06:39] andol: The exit_notify function in kernel/exit.c in the Linux kernel before 2.6.30-rc1 does not restrict exit signals when the CAP_KILL capability is held, which allows local users to send an arbitrary signal to a process by running a program that modifies the exit_signal field and then uses an exec system call to launch a setuid application. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2009-1337) [06:39] andol: I misread. Sorry. [06:40] I missed the word before when I read it. [06:42] Ohh well, I'm off for work... [07:07] i need any way to drop packets based on tcp window size, is this possible? [07:08] People sometimes do it by accident by adujusting MTU to be too large. [07:10] well, my server is being attacked by a large SYN flood [07:11] and from a sniff the packets have a tcp window size of 65535 (the max), and a tcp sequence number of 0 [07:11] so i need a way to drop them based on that [07:11] because that is obviously a forged packet [07:13] moparisthebest: Does turning on syncookies (with sysctl net.ipv4.tcp_syncookies=1 ) not deal with the problem? [07:13] no [07:14] i also set the syn timeouts as low as possible [07:18] It might be easier to check for sequence number of zero rather than for window size... tcpdump filter syntax should be able to do that... [07:18] * jmarsden reaches for "TCP/IP Illustrated Volume 1" ... [07:22] How do you make server search for the fastest mirror and use it? [07:24] fbc-mx: The tools for doing that are Debian-specific and have not been ported successfully to Ubuntu yet. Look at apt-spy and netselect if you are interested... but they do not work in Ubuntu at the moment, as far as I know. [07:26] jmarsden, hehe... thanks. that's the best one I've heard yet. [07:27] It was being discussed here (or was it on #ubuntu-motu) earlier today... [07:27] It was here, about 3 hours ago... [07:30] moparisthebest: I think a rcpdump filter of the form tcp and tcp[4:4] = 0 might match all TCP packets with sequence number zero. [07:31] s/rcpdump/tcpdump/ [07:33] what about the window size though? it's the combo of the two that makes it a fake packet [07:35] I would like to turn my server into a torrent client that I can control remotely either with a web-gui or gtk-gui. Is there such an animal? Some have mentioned deluge, but I'm still looking for documention for such a function. [07:36] fbc-mx, torrentflux :) [07:37] moparisthebest, thanks.. by the way, I have fond memories of a 440 with a six pack carb. [07:38] torrentflux is awesome, but not as awesome as that setup :) [07:39] moparisthebest: If the window size option is always at the same offset into the TCP header then you can check for it in a similar way. [07:39] hmm, alright [07:39] wait, actually i don't think that helps [07:40] since i can't actually drop them with tcpdump right? [07:40] Well, you can use something that uses tcpdump filter syntax to drop them... snort maybe? [07:43] Would the iptables tcpmss help?? [07:44] Hmmm, there is an iptables u32 thing that should get you the same kind of testing... but I've never used that... [07:55] i'll look into it [10:41] In a screen session, how do you scroll-back? [10:42] <_ruben> i usually do ctrl+a esc .. then use arrow keys/pgup/pgdn/etc [10:42] <_ruben> there are probably better/easier ways [10:43] And ctrl-a esc to get out again too? [10:44] <_ruben> just esc i think [10:45] Thanks! [10:49] <_ruben> its abusing the 'copy mode' .. you can copy/paste data in that mode, which i never used really ;) .. so there's likely to be a 'cleaner' way as well ;) [10:57] Cool, didn't know that. At least I got what I needed, a view of the scroll-back buffer. Next when I have some spare time I'll see if there is a "real" way :) Thanks again. [10:58] ngrep is a wonderful tool :) [10:58] Hmm, just realised, I could just pipe that to more :) [10:59] <_ruben> hehe [10:59] One of those doh! moments :) [11:00] Later all. [11:48] how do I get gnome for Ubuntu server [11:49] <_ruben> you'd install the desktop edition ;) [11:50] whats the difference in ubuntu server and desktop aside no GUI? [11:50] <_ruben> slightly different set of standard packages that are installed .. slightly different kernel [11:51] because Im trying to set up a webserver and xp and vista are resource hogs so...linux is the best option [11:51] New bug: #374819 in krb5 (main) "Missing dependency on update-inetd and other issues" [Undecided,New] https://launchpad.net/bugs/374819 [11:52] wasnt sure if desktop wouldnt have the stuff I needed or not [11:52] <_ruben> the webserver packages are optional for both server and desktop [11:52] sweet [11:53] thanks for the info ruben [11:53] <_ruben> (they're actually the same packages as well, server and desktop share everything except the base install) [11:54] ah ok [11:54] only difference is server has more stuff compared to the desktop size wise [11:54] rest you have to download on desktop right? [12:01] <_ruben> the server install is actually smaller than the desktop install (the gui takes up a fair bit of packages/space) [12:11] hello [12:13] i'm using virt-manager to deploy and manage kvm virtual machines. I'm using bridged networking and all seems fine, but i need to do a port forward. In kvm's documentation i saw the -redir parameter, but i can't seem to find a way to pass that parameter to virt-manager .. is there any way to do that? [12:22] <_ruben> with bridged networking the vm is just like another machine on your network, so portforwarding would be done on your router, not the kvm host === rgreening_ is now known as rgreening [13:40] is there a tutorial/cookbook etc for some sort of a basic storage area network using ubuntu server (ive been trying so far without luck to find it)? [13:41] Hey guys is anyone around that can help me recover from a failed grow operation on a raid 5 array on ubuntu 8.04 [13:54] Hey guys is anyone around that can help me recover from a failed grow operation on a raid 5 array on ubuntu 8.04 [13:56] Doonz, what's the matter? i MIGHT be able to help. [13:56] Hehe [13:56] well where to start [13:56] basically i was growing a raid 5 array when i lost power to my server [13:56] http://pastebin.com/m523cddcb [13:57] thats an output of my drives [13:57] im trying to figure out if is there a way to downgrade my mdadm whithout having to reboot the server [13:57] @Ubuntu-Server:~$ mdadm --version [13:57] mdadm - v2.6.7 - 6th June 2008 [13:57] and apparently this problem isnt in 2.6.4 [13:58] i just skimmed the output, but it looks alright. what's your issue? any reason for downgrading? [13:59] http://ubuntuforums.org/showthread.php?t=833191&highlight=failure+growing+raid+array&page=3 [13:59] this forum [13:59] thie thread* [13:59] im just struggling really im still very new to linux [14:00] heres when i try assebling with a force [14:00] @Ubuntu-Server:~$ sudo mdadm --assemble --verbose /dev/md2 /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1 [14:00] mdadm: looking for devices for /dev/md2 [14:00] Segmentation fault [14:00] downgrading mdadm while running the system shouldn't be an issue. mdadm is just a program for administrating a sw-raid (afaik). the actual work is done by the sw-raid kernel modules, which are impossible to up-/downgrade while the system is running (all the sw-raids would have to be unmounted, which is impossible if your system partition is on the raid). [14:02] can you walk me through this process? my system partition is on a hardware raid controller this raid array is purely for data storage and can be unmounted and remount at will without any impact [14:02] honestly: i've never played with any in-depth operations of sw-raids/mdadm. [14:03] oh [14:03] ok [14:03] im just running a cheap sw raid1 on my dedicated box. [14:03] hehe [14:03] so i got a little experience. [14:03] well this is a home media server [14:03] are you sure the downgrading will help? [14:03] no [14:04] im just looking at that thread on the forum [14:05] if you look at post 3 in that thread it says he patched his version of mdadm [14:08] any ideas? [14:09] ok, it appears mdadm suxx [14:10] that's what you need to fix [14:11] high risk, if-it-fucks-up-everything-it's-your-fault solution: install the corresponding package from a newer ubuntu version. might work, might not. if you're unlucky, hardy wont fulfill some dependencies. [14:14] hmm === asac_ is now known as asac [14:30] Doonz, got another idea: boot the server from a live medium (preferably with jaunty), fix the raid and reboot from your hardy install. might work, might not. [14:30] sorry, i'm hardly a help ... [14:32] no problem [14:32] i just noticed something [14:32] 09:35 <+Doonz> DEVICE /dev/sdc1 /dev/sdd1 /dev/sde1 [14:32] 09:35 <+Doonz> ARRAY /dev/md2 level=raid5 devices=/dev/sdc1,/dev/sdd1,/dev/sde1,/dev/sdb1 [14:33] notice how under devices it only list 3 hdd's [14:33] and the array has the 4 listed [14:33] could that be a a problem [14:55] might be what causes the segfault, but that's only a guess. [15:12] umm, i restarted a machine with fsck option set on reboot, so it booted up and is responding to pings now but no services are being started [15:13] could it be that it's brought up the network and is doing the fcsk now? and will only start services later on? (after an additional reboot) [15:13] seems weird for it to do that if thats the case [15:16] I was running a raid 5 grow operation when the power went out. It looks like my disk are fine but i cant reassemble the array. i get a segmentation fault. There is very little information on the site to help with this problem [15:18] <_ruben> Doonz: chances are big that you lost you're data .. during the grow operation, the array is in an incosistent state most of the time [15:19] is there a tutorial/cookbook etc for some sort of a basic storage area network using ubuntu server (ive been trying so far without luck to find it)? [15:19] <_ruben> s/you're/your/ [15:19] _ruben: http://pastebin.com/m523cddcb [15:19] thats the output of my examine [15:21] psteyn: fsck on root ? , you need to be able to write root to log in.. ( not 100% about it ) and normally its mounted read only during fsck [15:21] psteyn: not sure which partition needs to be rw, was it var home or / [15:21] root [15:21] <_ruben> Doonz: that does look rather sane [15:22] <_ruben> Doonz: segfault's strange though, could be a bug [15:22] yeah [15:22] the forums dont give a noobie like me an easy way of understanding howto fix [15:23] wbi havent seen one. but keywords: samba, nis , nfs, openafs, kerberos may help you to create big picture about options what you have [15:23] W8TAH: read above [15:23] <_ruben> i only resized a software raid5 once .. which went without a hitch, luckily [15:24] hehe [15:36] simplexio: thank you -- im already using samba -- and have about an zillion lil servers scattered around -- id like to consolidate them into some sort of a centralized group so that i can access them all as one volume [15:53] Ok i have this script. its a bash script it checks some condictions and then execuses a bin in the same folder so the path ./binfile is vaild, but in this case it wont work? whats wrong with the system? [15:53] I run this script on all sorts of systems [15:54] Is this a path issue? that ./ doesn't means same folder? or is it making the other script run some where else so that it doesn't have access to that bin? [15:54] y [15:57] seems i can't even launch it without the path. Could there be something missing from the system that would cause it not to be able to execute bin? [16:06] New bug: #374447 in samba4 (universe) "Jaunty: Samba permission problems" [Undecided,New] https://launchpad.net/bugs/374447 [16:12] can someone point me to a how-to to downgrade my version of mdadm [16:15] W8TAH: if you want to keep samba around, you should read stuff about openldpa, kerberos, AD [16:16] New bug: #374925 in php5 (main) "PHP5-MSSQL package needed (dup-of: 87063)" [Undecided,New] https://launchpad.net/bugs/374925 [16:17] Soo nobody [16:34] Doonz, downgrading won't help you. it's rather upgrading you need to do. [16:34] as i said before: give the live cd a shot. [16:35] they said it still is in jaunty [16:36] well. then you still need to upgrade to a fixed version. did the guy who wrote the third post publish a patch? [16:39] well ... im off. [16:43] RoAkSoAx: hi there [16:45] * mathiaz waves at ivoks [16:46] mathiaz: hi === ecanto is now known as edson [16:49] simplexio: thanks [17:06] hi anyone have idea on this error installing zaptel 1.4 http://pastebin.com/m126b3216 my distro si ubuntu server... [17:13] ivoks, heya master, how's it going? [17:13] you are missing headers [17:13] install linux-headers-server package [17:14] RoAkSoAx: fine, and you? [17:14] ivoks, been better :) I'm still hang over since yesterday [17:14] haha [17:15] ivoks:how..? [17:15] ruben23, sudo apt-get install linux-headers-server [17:15] ruben23: how to install oackagte? [17:15] package [17:16] What kind of fileformat should i format my harddrive into if i like to transfer huge files between Ubuntu and Mac OS X? [17:16] files > 4 gb [17:17] ext2 [17:17] RoAkSoAx: THANKS [17:18] ivoks " mac deos`nt support it [17:18] ub_: it does [17:18] ivoks: how? [17:18] mount_ext2 -o rdonly -x /dev/diskXsY /Volumes/Linux [17:19] RoAkSoAx:i already installed the packages get the same error.. [17:19] ruben23: then Makefile doesn't look for headers in right place [17:23] ivoks, sensors-applet. In the last ubuntu version, hardy, theyprovided a patch that is now included upstream, that's why i put that changelog entry [17:24] RoAkSoAx: then put 'name/of/the/patch dropped; included upstream' [17:24] RoAkSoAx: or don't add anything [17:25] RoAkSoAx: since that's a change that isn't preserved in this version [17:25] ivoks, ok I'll erase it then [17:26] does it matter what kind of torrentsoftware you use in? I thinking about the speed. [17:26] ivoks, and about gnome-doc-utils (>= 0.3.2), in the previous ubuntu version they seem to have added that dependency but there was no changelog entry for it. that's why i didn't put it... [17:27] RoAkSoAx: add it [17:28] ivoks, btw.. qemu has been successfully built so i'm removing it from the FTBFS list, i think it was a problem with libc-* package [17:29] could be [17:30] https://edge.launchpad.net/ubuntu/karmic/+source/qemu/0.10.3-1ubuntu1 [17:30] yep, looks ok [17:31] ivoks, yep there was a bug about linux-libc-dev that didn't contain a file and that was why it wouldn't built [17:33] bbl [17:33] take care [17:36] hi, can someone recommend a ticket support system ? [17:36] or tech support system [17:38] How can I make a service startup automatically when the server boots? I have to keep logging in to start deluged everytime we lose power or somethign of the sort. [17:41] Can I put it into rc.local? [17:44] oruwork: I like request-tracker [17:44] fbc-mx, is it a service provided there or have you made a custom script? [17:44] fbc-mx: do sudo update-rc.d servicename default [17:44] fbc-mx: that is if you have an init script in /etc/init.d [17:45] sommer, RoAkSoAx , what I tried was putting 'exec /usr/bin/deluged' before the 'exit 0' in the rc.local file [17:46] sommer, RoAkSoAx , you guys think that would work? [17:46] fbc-mx: yep that should work as well [17:48] sommer, great! no more grief of getting home expecting to watch something that isn't there. :-) [17:49] sommer, My computers bios allows me to power it back on in case of an outage, so I think I'm pretty much covered. [17:50] sommer, as long as I do not have to login for rc.local to execute, I'm ok. [17:52] sommer, but, I'm assuming I don't as a couple of the examples on line show it being used to detect the installed gdm and load it, and that usually happens before a user logs in, at least as far as I'm aware of. [18:10] fbc-mx: ya basically rc.local is the last init script to be executed, so it should work as you expect [18:25] sommer-> url please ? [18:27] oruwork: http://bestpractical.com/rt/ [18:27] oruwork: there's packages in universe as well [18:27] ok was just looking at that [18:33] sommer-> how do you spell the package ? [18:38] oruwork: there's multiple ones depending on your backed db etc: apt-cache search request-tracker [18:39] sommer-> i get this in return http://pastebin.com/m1406b0cc [18:41] oruwork: right so you'll need request-tracker3.6, probably rt3.6-apache2, and whichever database you'd like to use just install the appropriate package [18:54] jaunty's ocfs2 is older than sid :( [18:57] khaije1, yes, we'll merge ocfs2-tools for karmic [18:58] im going to try to import sid's deb, any reports on this working so far? [19:11] hi [19:12] how do I enable ping... when I ing I get a message saying: ping: sendmsg: Operation not permitted [19:18] Are you trying to do some special kind of ping? [19:18] fevel, are you doing this as a regular user or have you tried "sudo ping"? [19:18] im doing it as root [19:18] maybe ufw is blocking [19:18] ufw allow icmp isnt working [19:19] does anyone know the ping port... cant remember [19:19] maxb: nope... normall ping [19:19] unless you changed /etc/ufw/before.rules, ufw allow outgoing ping requests and incoming ping responses [19:19] There's no such thing as a ping port! [19:20] if unsure, 'ufw disable' and try again [19:20] jdstrand: yes I made it restrictive by default [19:20] jdstrand: cant disable it because of the masquerade dependant users [19:21] fevel: you can look in /usr/share/ufw/before.rules for what you deleted from /etc/ufw/before.rules [19:21] where should I insert established related rules [19:21] ok [19:21] specifically: [19:21] -A ufw-before-input -p icmp --icmp-type echo-request -j ACCEPT [19:21] ok thanks [19:22] was wondering where i could insert hand crafted rules [19:22] that is what /etc/ufw/*rules is for [19:22] ok [19:22] see 'man ufw' and /usr/share/doc/ufw/README.gz for details [19:23] (and I meant to say before, incoming and outgoing requests) [19:24] thanks for the help [19:24] got it working [19:25] np [19:51] hey, how do i mesure response time in my apache2 server ? not externaly, but internal without logfile things like %t %d , but in realtime ?# [20:43] Anyone feeling particularly bored? I'd really like apache's mod_status to maintain stats per vhost. [20:44] ...and it doesn't at the moment. I've been looking at the code a bit, and I can point out where to add it, "all you need to do" is implement a hash table mapping vhost to a stats structure of some sort. [20:45] ..and present it in the server-status page. [20:50] soren: why not blog about it? [20:50] soren: you'd reach a broader audience [20:50] mathiaz: Point. [20:50] I'll probably write a spec. [20:50] ...and then blog. [21:02] New bug: #312261 in samba (main) "unable to remove/delete .folder (hidden) from share" [Low,Invalid] https://launchpad.net/bugs/312261 [21:05] New bug: #375092 in dovecot (main) "package dovecot-common 1:1.1.4-0ubuntu1.2 failed to install/upgrade: " [Undecided,New] https://launchpad.net/bugs/375092 [21:08] if I password protect the recovery entry in grub's menu.1st and I forget the password, is there a way to recover? [21:09] err... I guess you could boot from cd and change the file :) [21:14] Hi, I'm running Ubuntu Hardy Server (64bit), and every time I try to do a "shutdown now", it starts to shut down, but then gives me a "recovery console", and whatever option i choose, it just boots back into linux. How do I fix this? [21:15] i've had this problem on multiple computers running ubuntu, also, not just servers [22:05] How long is a raid5 resize supposed to take [22:05] for 4 tb drives [22:06] I've seen a resize take quite a while, hour+ on smaller drives [22:08] i'm worried because the md5_reshape process is in uninterruptible sleep and is not running at all [22:08] er [22:08] md1 [22:09] and the md1_raid5 process is only taking up a few cpu, when it usually takes up around 40% when rebuilding the array === ssd7_ is now known as ssd7 [22:10] i ran: "sudo mdadm --grow --raid-devices=4 --backup-file=/root/grow_md1.bak /dev/md1" [22:10] it said it created the backup and passed the critical section, but the backup file does not exist [22:11] [>....................] reshape = 0.1% (1786880/966992448) finish=13342.0min speed=1203K/sec [22:11] ok [22:11] its resizing, but its very slow [22:12] Hum, sorry I should have been more specific in that I have seen a resize take a while on HW cards...I've never actually done it with mdadm. However the ninterruptible sleep would kind of make sense, I wouldnt worry. [22:12] this says it would take NINE DAYS [22:12] Im pretty sure thats not normal [22:12] Yeah, that's pretty bad [22:13] How long has it been going? [22:13] only 10 minutes [22:14] Oh, shoot. I wouldn't trust the time estimate for at least another 20min [22:15] stickystyle: but the md1_raid5 process is hardly taking up any cpu time at all [22:20] Well on a reasonably new box, recalculating the parity doesn't eat up that much cpu. I believe its largely io bound [22:20] do you have iostat installed? [22:20] no [22:22] That tool would give you a much better indication as to if your array is really working, or just moving along very slowly. [22:22] Has the finish time gone down? [22:22] no, its gone up [22:25] Hummm....then you may need to wait for someone else in this # to speak up, as that is as I understand not the normal behavior. I wouldn't want to steer you down the wrong path as my knowledge with mdadm is fairly limited. [22:25] ok [22:31] is it possible that its a bad drive? [22:31] I could see that as a possibility. [22:31] :/ [22:32] if you're using the drive, it gets _very_ slow [22:32] if you leave the machine alone, then it goes much faster [22:33] and sometimes you can give it hints that make it go really really fast (as in "ok, done then") - though I haven't bothered to figure out what those conditions are [22:33] lamont: i've used that partition while it was rebuilding the array before, and it only took 3 hours [22:34] 2.67TB, on a fresh (raid6) build took about 24 hours on my machine [22:34] SATA, 6 drives [22:34] while using it? [22:35] no. I walked away [22:36] it was threatening 6 days when I was using it [22:36] this is raid 5, 4 drives, 3 tb, by the way [22:36] so does reshaping take that much longer than rebuilding? [22:36] it's basically a rebuild after a little bit, sorta [22:37] but yeah, dunno [22:37] I know that if you kill the box during a rebuild, it can be torturous [22:37] and off with my daughter for her birthday [22:37] ok [22:50] hey folks, I'm trying to connect to my ubuntu server remotely (both on same network though) via telnet on Win7 but I'm getting that the connection is refused, is there a way to see the port it's listening on? [22:51] MkJackson: unless you installed telnet, ubuntu is not running a telnet daemon. [22:51] SSH. and even that you need to install openssh-server [22:53] i would also recommend ssh over telnet [22:56] ssh runs on port 22 by default, and a decent ssh client for windows is putty [22:59] hmm... when I try to "sudo apt-get install openssh-server" on my box I get an error saying there's no candidate or something to that effect [22:59] telnet is installed though [22:59] (and I can ping the server) [23:00] what version are you running? [23:00] try an 'apt-get update' prior to the install [23:01] and you mean a telnet client is installed on your ubuntu box, not a server right? [23:01] will update now (running 9.04 server 32-bit) [23:01] errr... I just did an apt-get install telnet and it responded that it was already installed [23:02] that's just the client [23:02] the telnet package is a virtual package that installs the client. [23:02] telnet-server would be the server package [23:05] update did the trick [23:05] when I did the apt-get on telnet-server though I got a could of packages other than because there was no exact match for telnet-server (or something to that effect) [23:06] Yeah, though don't install a telnet server unless you *really* have to [23:06] as it is 100% unencrypted [23:07] well, maybe I'm going about this totally wrong, basically I'm trying to get my feet wet with making a Linux server and since Ubuntu's been good to me I went with their solution... so I got this laptop running it and it's connected to my network and tucked away, I figured I could just remote connection via telnet, is there a better option? I heard SSH via PuTTY so I'm downloading that now... [23:08] Yes, SSH should be the only thing you consider. [23:08] PuTTY works quite well [23:09] a victory... :-) thanks so much folks... [23:09] No problem, come back anytime :) [23:16] victory! [23:21] New bug: #261831 in gcc-4.2 (main) "gcc-4.2 cannot compile bind 9.5.0.P2 on i386" [Undecided,Incomplete] https://launchpad.net/bugs/261831 [23:33] cjwatson: hai. i have been following putty gtk2 in ubuntu for a while. do you have a sec for a question? Does copy and paste work for you in your testing on karmic with 0.60+2009-04-05-1? [23:34] i compiled putty's dev snapshot in 9.04 and everything seems to be ok, except copy and paste. i have rolled my own on 8.10 for a while and do not remember this issue. trying to determine if it is a regression upstream, an affliction of my shiny new jaunty desktop, or what. [23:37] slestak: I'm using 0.60+2009-04-05-1 on jaunty, and I only really use pterm, but it seems to work for me [23:38] (gah, and I see that it fails to build in karmic; must fix that ...) [23:38] cjwatson: yes, pterm is what I am after. is there a ppa for this? id love to help test. [23:39] must, have, gtk2... [23:40] slestak: just the .deb from Debian installed with dpkg -i [23:40] slestak: you can grab it from http://ftp.debian.org/debian/pool/main/p/putty/ [23:41] cjwatson: thx. i see that now. appreciate your work on bringing this to ubuntu [23:42] slestak: I'm not sure how copy and paste could go wrong. We're talking about standard X copy/paste, aren't we? i.e. highlight to copy, middle-click to paste? [23:47] cjwatson: yes, that is what i dont understand. i could not get any test from my gtk2 putty to transfer to any other x app [23:53] cjwatson: sth is up. still doesnt work. [23:54] I'm stumped, I'm afraid. Feel free to file a bug with as much detail as you can manage [23:55] this is weird, i can paste into gvim, but not gedit. [23:55] canoot paste into my wiki on firefox (ala fckeditor) [23:56] how can gvin see the recent copy [23:56] s/vin/vim [23:56] are you attempting to use edit->paste in gedit? [23:56] I tried edit-paste as well as shift-insert [23:57] do you know that there are two entirely separate copy/paste systems in X? [23:57] highlight/middle-click is one of them [23:57] doh, crap [23:57] edit->copy/edit->paste (or shift-insert or whatever) is the other [23:57] try middle-click to paste instead [23:57] that works [23:58] sorry to waste your time [23:58] np