=== medberry is now known as med_out [00:22] soren: there's the Hungarian Phrase Book [00:26] Is there some magic I need to do so that Ubuntu doesn't take up gobs and gobs of CPU time when copying, moving, or deleting files? Just deleting a bunch of files has bumped my load average to close to "3". [01:00] any good ubuntu based email appliance distrubution [01:03] I'm using this tutorial to configure a dhcp server [01:03] http://www.howtoforge.com/dhcp_server_linux_debian_sarge [01:03] However, I'm not sure how to start it, the command in the tutorial didn't work. [01:03] 'no such file or dir' [01:09] Calif, cause its called dhcp3-server? [01:10] well, I'm sure its sometihng stupid like that [01:10] I've got dhcpd.conf [01:11] I don't know, maybe this isn't even the right tutorial, I thuoght that was what they had promoted [01:12] Calif, dhcpd.conf is used here on lucid-server, but in /etc/dhcp3 [01:13] ok, let me try something [01:13] one sec [01:14] ok, I've messed this up good [01:14] I should probably start over [01:15] I just want a simple dhcp server, whats something that I can get with apt-get, that you would recommend? [01:15] is this dhcp3-server good? Would you have used something else? [01:19] fuck it, I made a backup ill just start over [01:21] Calif: dhcp3-server is good, but if you only need a simple DHCP/DNS server for a small LAN, you can also look at dnsmasq [01:23] Here's another question [01:23] # No service will be given on this subnet, but declaring it helps the [01:23] # DHCP server to understand the network topology. [01:23] #subnet 10.152.187.0 netmask 255.255.255.0 { [01:23] #} [01:23] What is this subnet - is this the external, internal, whats the point of this part? [01:26] Calif, that only applies if you have a 10.x.x.0 private lan [01:27] what if I have a different address, but it's still a private address scheme on the lan I wish to use dhcp with? [01:27] eg: 192, or 172 [01:27] Calif, keep reading the samples, they show an example [01:28] /usr/share/doc/dhcp3-server/examples/dhcpd.conf [01:28] I'm looking at a few under that [01:29] So the purpose of these lines is just to show different config examples, the above one being the simplest? [01:29] further more, the above one can remain commented without affecting anything? [01:30] I'm running this on a linux box thats doubling as a router/fw for my network [01:30] so I want to offer dhcp services to the internal [01:31] that said, the external interface is dhcp as well, but it gets its address from an upstream router [01:32] I'm sorry for all the questions, I'm just trying to wrap my head around this. [01:33] Calif: for a small LAN, dnsmasq is much easier to set up & includes a DNS relay server (but lacks several more advanced features that dhcp3-server & bind support) [01:33] Calif, your internals are limited to certain nics: in dhcp.conf I have - INTERFACES="eth1 eth2" [01:33] eth0 is for my WAN [01:34] dhcpd.conf doesnt allow me to specify the interfaces [01:35] That said, if I just uncomment one of these, and give it the addresses of my internal interface, can I expect that to work? [01:35] Or is there some other place I specify the interface I want to use? [01:36] * ChmEarl slaps Calif with a trout: I just gave you the line [01:36] you said dhcp.conf [01:36] So i have to edit that file as well? [01:36] or did you typo it [01:36] and meant dhcpd [01:37] I swear, I wish I knew this stuff like you guys do ;p [01:38] I'm like 10 years behind or something until a month ago, the most advanced thing I did was setup an eggdrpo [01:38] now im running a server with a firewall, and dhcp, I'm setting up nagios [01:38] but I am fairly challenged by these tasks, it's going to be a long time before I understand nix like some of you guys [01:39] Calif: reading up on basic networking and what DHCP does might be useful ;) [01:39] I've got basic networking, and I understand dhcp, but I'm coming from windows so there's a disconnect [01:40] it works exactly the same in Windows ;) [01:40] it works the same, but the setup is radically different ;p [01:40] it's similiar in the physical settings [01:40] but how you put it all together, and knowing what to edit where, I get confused easily [01:41] I can set this up on windows server with my eyes closed - setting up reservations, static assignments, ranges, all sorts of stuff [01:41] And I've read stuff on linux too [01:42] You all probably know the 'linux newbie administrator guide; [01:42] I've read that twice over the years, (not to say I'd not benefit from another round mind you.) [01:42] with me, I think, the trouble is I read stuff and it makes sense, but I need to _do_ it for it to sink in [01:43] and half the tutorials I find on stuff leave things out, or expect and assume you know some small detail... You can't ask a tutorial a question... so I wind up here [01:44] anyway, so back to this - I think if I can figure this out, I'll be able to do the rest and run this thing in a few minutes [01:44] I'm going to go look at your dhcp.conf, and see if that answers my question [01:45] well, my dhcp.conf ;p [01:46] ok, so using the locate command, I found a copy of that in the /etc/samba dir [01:46] is it supposed to be an empty file ? [01:46] probably.. [01:47] ChmEarl, I just add the INTERFACES="ethx" - x being the interface going out to my internal network that I want to offer dhcp services on? [01:48] I suggest you don't go looking for similarly named files in random other directories ;) [01:48] lol [01:49] So was he meaning to tell me to edit dhcpd.conf with that line, or did I miss something entirely here, because locate only came up with one file [01:50] if it is dhcpd.conf where in the file do I stick that line at? [01:51] I'm sure once I get the interface and that line I pasted earlier figured out, that would be all my questions for tonight ;p [01:51] There is an end to the madness I promise ;p [01:53] I know im close, it'd suck to abandon this 70% through to pick another dhcp server, that I'll probably likewise have dumb questions about... [01:55] chm, I didn't pm you without asking, I used notice - I'm making an effort :P [01:57] Calif: I configured my first gateway with dnsmasq in less than 30 minutes ;) [01:58] less than 15 maybe [01:58] If I can figure out what I need to do next I'd be done in 2 [01:58] and I suppose dhcpd.conf is the file for the DHCP server config [01:58] I believe so [01:59] I can paste you the contents in a pastebin if you think you could help [01:59] I think it's simple, but I know there's something to what chm said too about defining my interface [01:59] I know the info I want to put in for my network, but there's some stuff it says mainly for the examples that confuses me as to why they worded it like that [02:00] I'm tempted to ignore that part and just uncomment one of teh examples, change some stuff to my info, but that leaves the interface thing chm spoke of, that I can't figure out where to put it "INTERFACES="eth1 eth2"" [02:02] Sorry, lost my connection [02:04] My room mates coming home soon, I may have to give up soon [02:05] well, 3am here, so I give up now :P [02:05] dnsmasq ROCKS [02:06] Cool, thansk for your help JanC [02:06] Calif: interface lines go in dnsmasq.conf [02:06] I'm not using dnsmasq I don't think [02:06] Oh sorry [02:06] dhcpd.conf [02:06] Then I can't help :P [02:06] crap :/ [02:06] lol [02:07] ok, I promise if I don't figure this out in a day or so [02:07] that will be the first thing I try [02:07] I'm already mostly through this one [02:07] so I'm hoping to finish it before I jump to another one [02:07] it'd be nice to figure out both in the end [02:28] Is there a way to tell nmap to operate over an ssh hop (e.g. SOCKS with ssh -D) ? [02:29] nmap is on my laptop, I can ssh into bastion.example.net, and from there I can connect to foo.example.net -- but I can't reach foo.example.net directly from my laptop, and I can't install nmap on bastio [02:30] ssh -I === jjohansen is now known as jj-afk === CrazyGir is now known as Guest74692 [03:31] how can i setup a default pipe for email that doesn't have a real account? [03:31] like a catch all [03:56] goddard: with postfix? [03:56] New bug: #777541 in php5 (main) "package libapache2-mod-php5 5.3.5-1ubuntu7.2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/777541 [03:56] yes [03:56] ok, I read a bit about dhcpd [03:56] goddard: dunno, try #postfix [03:56] I'm a bit confused with declaring a subnet/interface [03:57] I'm going to run dhcp on a box which acts as a gateway/firewall, and the external interface is a public address but it dynamically changes. [03:57] So how would I declare such a subnet in dhcpd.conf (isc)? [04:08] Someone please explain this: /bin/sh: hardcopy.7: not found [04:08] Someone please explain this: http://paste.debian.net/116054/ [04:08] Ah, -h [04:09] never mind === hikeonpast___ is now known as hikeonpast === hikeonpast___ is now known as hikeonpast [05:39] Any grub experts out there? I've got a client getting grub error 24 "Attempt to access block outside partition" on a server that was previously running perfectly. Bug #353071 seems to be related, but it is for much older versions, and i've tried all the suggested remedies. [05:39] Launchpad bug 353071 in grub "grub error 24 with Jaunty beta" [Undecided,Invalid] https://launchpad.net/bugs/353071 === hikeonpast__ is now known as hikeonpast [06:14] And i forgot to mention that there's no ext4 involved - all filesystems are ext3 [06:15] blahdeblah: What release were you using when the systems were initially installed? [06:15] I don't know - i am only new on the site. I think it was hardy. [06:15] Certainly it was upgraded from hardy [06:16] Then you at least have grub and not grub2. [06:16] yes, definitely [06:17] The upgrade from hardy was fine for a few weeks, then something happened (don't know what, since i am only on site when required) and on the next reboot it exhibited this problem. [06:17] New bug: #777585 in bacula (main) "package bacula-director-mysql 5.0.3-0ubuntu2 failed to install/upgrade: サブプロセス インストール済みの post-installation スクリプト はエラー終了ステータス 1 を返しました" [Undecided,New] https://launchpad.net/bugs/777585 [06:18] ScottK: I wondered whether i should upgrade to grub2, but i don't like to do upgrades in the middle of an unexpected downtime. [06:18] I wouldn't. [06:19] You're sure some enterprising young admin didn't decide these file systems should be converted to ext4? [06:19] I'd guess not. [06:19] yes [06:19] Very sure [06:19] blahdeblah: grub2 is a downgrade [06:20] I also ran a full fsck which came back with no errors [06:20] I've reinstalled grub a few times, using both the /dev/sdX devices, and the /dev/md0 device, and all result in the same behaviour. [06:20] extlinux FTW [06:20] Sorry, no idea. [06:20] You've used it on CDs, you've used it from PXE, now use it for normal booting! [06:36] <_ruben> hooray for co-workers not showing up for planned maintainance.. grr === Guest33391 is now known as phate408 [08:35] What is your favorite file system for hosting/scaling a large MySQL database and why? [08:37] talntid: That may be a better question for the mysql room. [08:42] New bug: #777626 in php5 (main) "package libapache2-mod-php5 5.3.5-1ubuntu7.2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 10" [Undecided,New] https://launchpad.net/bugs/777626 [08:51] I removed the stupid NAT interface from libvirt, because I only want bridging [08:52] I just upgraded libvirt, and the postinst has recreated libvirt/qemu/networks/autostart/default.xml -- why? [08:52] That is not how conffiles are supposed to behave [08:55] hello, I'm trying to upgrade ubuntu server to 11.04 and I get following errors: http://pastebin.com/9whvVhs1 . Can i throw out one of those offending packages? [08:58] Try dpkg -i /var/cache/apt/archives/console-setup_....deb /var/cache/apt/archives/keyboard-configuration_....deb [08:58] ok, I'll try that [08:59] It's trying to upgrade keyboard-configuration first due to bad dependency declarations, I think [08:59] Note that you should be upgrading only one version at a time [08:59] e.g. 10.04 directly to 11.04 is not supported [09:00] oh, that could be the problem [09:33] can some1 tell me where i can find the changelog for libapache2-mod-php5 last night update please ? [09:33] DISTRIB_DESCRIPTION="Ubuntu 8.04.4 LTS" [09:34] i had some trouble with various CMS since last w.e. which has been corrected last night. [09:35] th0mz: aptitude changelog libapache2-mod-php5 === taneli is now known as xampart [09:37] xampart: great, thanks. [09:45] Did you know that in d-i, if you drop to a shell and do "mkfs.ext2 /dev/sda", you can convince d-i that you don't need a partition table? [09:46] This is especially good for /dev/vda VM virtual disks === mrmist is now known as to === to is now known as mrmist [10:37] Oh, NICE. Warning: Permanently added 'example-natty' (ECDSA) to the list of known hosts. [10:38] natty has ECDSA host keys OOTB [11:18] can anyone help with a virtual host configuration? I have 2 internal ip ś with 2 webservers. I want mail.domain.tld to refer to webmail and domain.tld to my website. [11:19] and both are port 80 [11:27] dany_: http://httpd.apache.org/docs/2.0/mod/core.html#servername [11:28] thank you [11:29] dns should be properly configured of course [11:31] any idea what might cause this? http://paste.ubuntu.com/603635/ [11:32] dany_: you don't need two IPs either - it's easier to just use VirtualHost with host header (ServerName / ServerAlias) [11:32] Hi! I've got a server running my application. [11:32] I've created an Upstart script to start it. [11:32] well i cannopt get zarafa an a joomla website to play nice on an OpenVZ container [11:32] so thats why i am trying this way [11:33] ok - never tried openvz [11:33] Previously, we were using start-stop-daemon, but I've removed it (because I made some changes and it was causing problems), so Upstart is starting the service directly now. [11:33] I notice that the Debian guidelines recommend start-stop-daemon, but is this the recommendation for Ubuntu+Upstart? [11:33] jkakar: both work [11:34] RoyK: I guess they're kind of equivalent. [11:34] just different ways to do things [11:34] In this case my application is Twisted based and started by twistd. I guess the reason to use start-stop-daemon would because (apparently) twistd's ability to switch to the right user can be dodgy. [11:35] but - anyone that knows what might cause this? http://paste.ubuntu.com/603635/ [11:38] jkakar: start-stop-daemon is used to daemonise things that aren't very good at daemonising on their own. upstart is a proper process supervisor. You don't need start-stop-daemon with upstart. [11:39] soren: Cool, that's my impression. [11:39] I just need to test my application to make sure it switches users properly. [11:40] jkakar: If not, I use su to do that in upstart. [11:40] e.g.: [11:40] exec su -c "nova-api --flagfile=/etc/nova/nova.conf" nova [11:41] soren: Ah, interesting. [11:41] soren: Thanks for the advice. It helps. :) [11:41] jkakar: It's not pretty, but until upstart learns how to do this natively, that seems to be the common pattern. [11:41] * soren wanders off for lunch [11:41] jkakar: np [11:55] I'm using ubuntu on ec2, one of my instances has stopped being able to connect to the apt repo.. anyone had this before? [11:55] actually it can't even connect to google [11:56] tim_s: Have you done something to the security group the instance is in that could cause such breakage? [11:57] I can however see pages from it's apache server from my laptop. [11:57] jkakar, not that I know of. [11:58] jkakar: we haven't [11:59] tim_s: Hrm, I'm not sure what to suggest then. :/ [12:00] tim_s: I've seen my EC2 instances (in eu-west-1) occasionally lose connectivity to the archive, but never to the internet in general. [12:00] tim_s: Maybe something b0rked DNS on that machine? [12:00] sounds like dns [12:00] it can ping things [12:00] oh I can ping google.com [12:16] can anyone help me with updating apache and mysql on a 10.04.2 box? I ran apt-get update and apt-get upgrade but it doesnt pick up the newer version [12:17] apache 2.2.14 is the latest it has, and that apparently has some security issues that are causing it to fail a PCI scan [12:32] doh, now hw is gone... [12:36] apache failing to do a PCI scan... [12:36] oh well [12:39] RoyK: "Payment Card Industry" scan; you know "security experts" certified by credit card companies who are too stupid/lazy to test for actual vulnerabilities but just compare version numbers instead... ;) === rsaidan is now known as nerens [12:45] <_johnny> hey, i use pubkey for my login to a box i have. now i'm adding another user (but still me who will use it), and i've added the same pubkey (from my computer obviously) to that users .ssh/authorized_keys. however, i'm thrown to password being the only allowed method http://pastebin.com/x3L6s8Uq [12:46] <_johnny> can anyone help me debug this? [12:48] <_johnny> there are no different rules from sshd to this new user. and the old user works fine with pubkey. [12:48] <_ruben> _johnny: check the logs of the ssh daemon, and verify the permissions/ownership of the keys [12:49] <_johnny> ownership might be wrong, good point [12:54] Hi! I did a fresh installation of Ubuntu Server Natty, choosed btrfs for '/' and wanted grub to reside in /dev/sda [12:55] <_johnny> _ruben: should i set it to verbose first? auth.log just tells me the pw is wrong (no mention of the pubkey attempts) [12:55] installation finished, but booting does not work. i can chroot into the natty-server via live-usb [12:56] <_johnny> _ruben: turns out my screen isn't found (although it exists). thanks for the help :) [12:56] and now "# grub-install /dev/sda" gives me: [12:56] /usr/sbin/grub-probe: error: cannot find a device for /boot/grub (is /dev mounted?) [12:57] dev is mounted and i can see /boot/grub .. Any idea someone? ;) [13:03] <_johnny> _ruben: oh no. now it's even harder to debug :( http://pastebin.com/6WYHpEVp i just get failed, and client mentions no reason for failure either [13:10] <_johnny> ok, the bad perms might be due to encrypted homedir. my bad [13:17] <_ruben> encrypted homedirs and key auth aren't best friends, you'll need to store you key in both your encrypted and unencrypted homedir [13:24] webwurst: is the live usb also natty? [13:25] JanC: yes [13:26] well, I guess it doesn't really matter when you chroot into the server [13:26] I think btrfs.mod is rather new though [13:27] JanC: boot is a directory. would it maybe help to make it subvolume and put it indo /etc/fstab? is grub-installer searching for something like that? [13:28] grub installer should work either way [13:28] ok [13:29] webwurst: /dev is mounted in the chroot too? [13:30] yes: i mounted proc, dev, sys with --bind [13:33] JanC: grub-installer fails at this line: [13:33] /usr/sbin/grub-probe --device-map=/boot/grub/device.map --target=device /boot/grub [13:35] JanC: ..and there ist no device.map in /boot/grub/ [13:42] JanC: ist --target=device right in this case? [13:50] webwurst: eh, I guess not [13:51] JanC: changed it to "fs" but same error ("cannot find a device for /boot/grub") [14:07] where can I find mac-fdisk for ubuntu? [14:08] Aison: is it packaged? [14:09] no idea, maybe I can also get the sources somewhere [14:09] I need to fix a mac partition table on an external drive === ScottK2 is now known as ScottK [14:09] Aison: i cannot find such a command anyway [14:12] New bug: #777820 in mcollective (universe) "mcollective does not start." [Undecided,New] https://launchpad.net/bugs/777820 [14:12] pmatulis, maybe there are other tools to manipulate hfs partitions with linux, no idea === nijaba_afk is now known as nijaba [14:31] Recently, I had an attack on my wordpress site, which relayed over 160,000 e-mails through www-data using a PHP form. Is there a way to specify that www-data@mydomain.tld can only send to mydomain.tld? Any suggestions otherwise? Any help is greatly appreciated. [14:45] ok, so in fact you CAN dist-upgrade directly from 09.04 to 11.04 ... altough with a few issues [14:48] rafalk42: Just because you can doesn't mean you should. [14:52] yeah, but I didn't knew about that before i started [14:55] hey [14:55] hi [14:56] anyone have any ideas on this one? I have a U 10.04.2 box that needs apache and mysql updated... when I run apt-get update, apt-get upgrade it doesnt pick up the newer versions [14:56] it has apache 2.2.14, and the latest is 2.2.16 [14:56] any ideas? this one is kicking my butt :-( [14:57] I see I could go to 10.10, but this has 10.04.2 LTS, not sure if I can upgrade... looking at this for a friend [14:58] 2.2.14-5ubuntu8.4 is the latest version in 10.04. [14:58] so, when will this get updated? 2.2.14 has some security holes that make it fail a PCI scan [14:59] or is really my only option to install 10.10? [14:59] 10.10 I saw has apache 2.2.16 [15:03] is there anything that would limit the number of concurrent SSH connections to a Lucid server? (besides h/w resources) === koolhead11 is now known as koolhead11|afk [15:05] the maximum number of PTY's (assuming the SSH connections have PTY's) is one thing. [15:05] think the default is 4096 [15:06] i'm seeing disconnections after about 80 [15:06] The default number per connection I think is like 10 [15:07] Hi everyone, can anyone tell me why aptitude would tell me the following, or how i can get it to install mplayer on ubuntu server? http://pastie.org/1868062 [15:07] genii-around: default number per connection? isn't each session a separate connection (unless using multiplexing) [15:11] <_johnny> how come update-motd doesn't have a /usr/sbin executable anymore? [15:12] <_johnny> i mean, sure, in the long run anyone can wait 10 mins, but if you were to debug/edit, then waiting 10 mins for each change could get boring ;) [15:16] New bug: #777868 in postfix (main) "Postfix smtp can no longer resolve localhost - regression from maverick." [Undecided,New] https://launchpad.net/bugs/777868 === JanC_ is now known as JanC [15:17] hw: No, actually it doesn't have security holes. What's happening is some brain dead security tool is looking at version numbers and not checking for actual vulnerabilities. Ubuntu backports security patches. [15:17] pmatulis: From manpage: " MaxSessions - Specifies the maximum number of open sessions permitted per network connection. The default is 10." [15:17] ( sshd_config ) [15:17] ScottK Yea, basically the issue is securitymetrics did a quarterly PCI scan, and that failed. [15:18] Is there documentation of that somewhere? I can throw that back at securitymetrics [15:18] hw: the changelog and the cve tracker? [15:19] hw: https://wiki.ubuntu.com/SecurityTeam/FAQ#Versions [15:19] hw: http://changelogs.ubuntu.com/changelogs/pool/main/a/apache2/apache2_2.2.14-5ubuntu8.4/changelog [15:20] on this note, anyone see that postfix security question, allowing the use of EHLO is a security issue, according to some scanners :) [15:20] I did. [15:20] kees: ScottK: thank you!! let me review what the tool reported [15:20] Clearly I'm not leveraging other people's ignorance nearly hard enough. [15:20] use of EHLO is not a security issue. :P [15:20] these whole security scan things seem so silly, the more I see people use them [15:20] kees: It was on postfix-users. [15:21] yea, I agree. I've had to call the security scan providers a few times to get them to remove bogus stuff [15:22] usually they do it without any issue, as long as you are technical enough [15:22] now that, make me even more scared :) [15:22] your paying them to be technical enough to do the scan [15:22] not for you to tell them how to scan [15:22] indeed. [15:23] if your a 3rd party in this, then ya, I can understand :( (customer hired x to scan their hosting stuff on your equipment) [15:23] patdk-wk: you pay them to get some stupid worthless piece of paper that some financial company requires, I guess [15:24] janc, pretty much :( [15:24] nice way to create new jobs for the economy :P [15:24] heh [15:24] it's what burocrates are good at [15:25] ok, the PCI scan says CVE-2010-1452 and CVE-2010-0425 are the issue.. the changelog says CVE-2010-1452 was fixed, but the other one is not on there [15:25] hw: The (1) mod_cache and (2) mod_dav modules in the Apache HTTP Server 2.2.x before 2.2.16 allow remote attackers to cause a denial of service (process crash) via a request that lacks a path. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2010-1452) [15:25] hw: modules/arch/win32/mod_isapi.c in mod_isapi in the Apache HTTP Server 2.0.37 through 2.0.63, 2.2.0 through 2.2.14, and 2.3.x before 2.3.7, when running on Windows, does not ensure that request processing is complete before calling isapi_unload for an ISAPI .dll module, which allows remote attackers to execute arbitrary code via unspecified vectors related to a crafted request, a reset packet, and "orphaned callback pointers." (http://cve.mitr [15:25] hw: The (1) mod_cache and (2) mod_dav modules in the Apache HTTP Server 2.2.x before 2.2.16 allow remote attackers to cause a denial of service (process crash) via a request that lacks a path. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2010-1452) [15:27] hw, CVE-2010-0425 is an IIS issue, not linux :) [15:27] patdk-wk: modules/arch/win32/mod_isapi.c in mod_isapi in the Apache HTTP Server 2.0.37 through 2.0.63, 2.2.0 through 2.2.14, and 2.3.x before 2.3.7, when running on Windows, does not ensure that request processing is complete before calling isapi_unload for an ISAPI .dll module, which allows remote attackers to execute arbitrary code via unspecified vectors related to a crafted request, a reset packet, and "orphaned callback pointers." (http://cv [15:27] so no, it won't be *fixed* [15:28] not an IIS issue, but in Apache's compatibility for IIS plugins or something like that [15:28] in any case, it's Windows-only [15:28] and shows how stupid that scan is :P [15:28] genii-around: interesting, thx [15:29] pmatulis: np [15:29] * patdk-wk never realized you could run apache as an iis isapi [15:29] * patdk-wk doesn't think he would ever want too [15:29] patdk-wk: I think it's the other way around, you can run ISAPI plugins on Apache ;) [15:30] oh? [15:30] guess for .asp and .net stuff then [15:30] maybe [15:30] shows you how much I run apache on windows :) [15:41] patdk-wk: LOL I just read that a bit closer. Too Funny. Gotta love the security scan protecting this server from the dangers of Windows [15:43] hw: tell them you didn't put "Wine" on your server, see if they understand ;) [15:43] lol [15:50] damn server is drunk again :( [15:52] <_ruben> patdk-wk: using alcohol instead of watercooling? :) [15:53] for *freeze* protection :) [15:53] <_ruben> heh, hadnt thought of that one ;) [15:54] * patdk-wk would also assume alcohol fumes would smell nicer :) [15:56] <_ruben> assuming you'd be concious still ;) [16:04] Hey, thanks everyone for your help! I really appreciate it. I checked all of the "security issues" on the PCI scan, and they were patched a long time ago. [16:05] For some reason, securitymetrics.com started failing these versions in the last few months [16:06] going to call them today to get this fixed, I'm armed with the changelogs :-D [16:07] * patdk-wk hands hw a larger bat [16:07] seriously [16:07] what was that comment about helping the economy? [16:07] lol [16:12] helping the economy is when you pay somebody to go over there with a bat ;) [16:12] hi: if i add a mount to /etc/fstab and it's options are .... nfs ro 0 1 -- the mount doesn't get added at boot. Any suggestions? [16:14] robo: the mount options should be separtated by comma nfs,ro -not sure if that's your entire issue though. You don't have to reboot to test. Just run 'sudo mount -t nfs -a' [16:14] *separated [16:15] eh, nfs is not an option but the fs type ;) [16:16] oh right :) [16:16] heh [16:19] oCean, mount -t nfs -a mounts it [16:19] just not after a reboot. [16:32] so the console is showing errors about how mountall failed. Before I get into why I'm trying to figure out how I get that same error when I'm ssh'd in to the machine [16:32] I would think dmesg would have that, but it doesn't [16:32] robo: dmesg only has kernel messages (for now) [16:32] robo: you want /var/log/boot.log [16:33] ah, ty [16:33] oh, that doesn't exist [16:33] (ubuntu 9.10) [16:34] there is a /var/log/boot that only has "(Nothing has been logged yet.)" [16:34] I have configured my virtual host like this but my subdomain mail.domain.TLD will only refer to my website and not the webmail as you can see in the config > [16:34] ServerName mail.domain.TLD [16:34] DocumentRoot /usr/share/zarafa-webaccess [16:34] [16:34] [16:35] Am i just being a complete noob or is something else a problem [16:35] can somebody assist me with debugging an email issue with postfix? [16:37] i have a user running Apple Mail on a Mac. He quite often receives emails that don't show the attachments but some email source code instead (looks like attachment boundaries). [16:37] I am trying to figure out if his Mac is butchering the email (or doesn't display it right) [16:37] or if the email is somewhat malformed [16:39] darnit. Just got pulled away by my boss. Need to bring this up later. [16:39] Bye [16:42] is there a way to show what programs are going to start at which runlevels? A centos chkconfig --list equivalent is what I'm searching for [16:44] there are no runlevels :) [16:44] ? [16:44] patdk-lap, that's something that confuses me too that I need to look into. It shows I'm in runlevel 2 -- that's strange to me [16:45] Hi, I'm attempting to run dhcpd - I'm running it and getting the following error: http://pastebin.com/TRp4TiQC [16:45] ubuntu only has runlevels to keep compatability with old stuff [16:45] but it doesn't *use them* [16:45] the same with almost every distro now [16:45] RIP run levels :/ [16:46] hmm [16:46] so how do I know if my server is booted into a gui? [16:47] normally if I see runlevel 3 I know I'm all text. Runlevel 5 I know I have X booted [16:47] heh, a server with a gui is just pure evil :) [16:48] i know. That's why I'd like to figure out if a gui is running [16:48] i just inherited these servers and I know next to nothing about ubuntu :-/ [16:48] service gdm status? [16:49] says un-recognized. Good thing! [16:49] heh? [16:49] what is un-recognized? [16:49] Hi everyone, can anyone tell me why aptitude would tell me the following, or how i can get it to install mplayer on ubuntu server? http://pastie.org/1868062 [16:49] gdm: unrecognized service [16:50] ah, not installed [16:50] so it's not running, which is good :-) [16:50] So, if ubuntu doesn't use runlevels how do I know which scripts are executed at boot? [16:51] Defusal: What release of Ubuntu is that? [16:51] 10.04 [16:51] i guess if runlevel shows i'm in runlevel 2 then I look in /etc/rc2.d ? [16:51] heh, I'm lazy to figure it out :) 'grep ^start /etc/init/*' [16:54] Defusal: Can you pastebin the output of: apt-cache policy mplayer ? [16:54] man patdk-wk, ubuntu is so confusing [16:54] heh, I don't think so :) [16:55] i need to learn it [16:55] I ran slackware for decades [16:55] hated centos/redhat [16:55] is ubuntu-server and debian pretty much the same? [16:55] Pici: http://pastebin.com/4vZVcRQ0 [16:56] Defusal: You aren't trying to install the 10.04 version of mplayer. [16:56] is it because it didnt remove the debian-multimedia source yet? [16:57] You need to remove that. === med_out is now known as medberry [16:57] From /etc/apt/sources/list or list.d [16:57] ah, thanks, my bad :) [16:57] needed some packages and forgot to remove it earlier [16:59] Does anyone know if Can i change the default url froom /webaccess to mail.domain.TLD ? [16:59] robo: In many ways. Ubuntu Server will generally have newer packages than the most recently Debian release. Ubuntu uses apparmor by default instead of selinux. Ubuntu has also implemented a number of security features that aren't in Debian yet: https://wiki.ubuntu.com/Security/Features [16:59] in zarafa [16:59] Anyone able to help me make a murmur upstart job? I run mine and it says murmur starts, but infact does not. This is the line I am using "exec /home/user/murmur/murmur.x86" [16:59] SpamapS: ping upstart question for you [17:13] zul: sure in a bit [17:36] New bug: #777972 in dhcp3 (main) "package dhcp3-server 3.1.3-2ubuntu3.2 failed to install/upgrade: Unterprozess installiertes post-installation-Skript gab den Fehlerwert 1 zurück" [Undecided,New] https://launchpad.net/bugs/777972 [18:00] kirkland: around? [18:03] Hello. When I log in to my server it says there are 50+ updateds but when I try to upgrade or dist-upgrade it says there are no updates? === foxbuntu` is now known as foxbuntu [18:10] hi: I'm getting this error on boot: mount.nfs: rpc.statd is not running but is required for remote locking [18:10] any suggestions on why this is happening? [18:10] start portmap [18:11] or even - enable it [18:11] I see rpc.statd running [18:11] service portmap status shows portmap start/running, process 898 [18:11] perhaps portmap is started after the nfs mounts? [18:11] yeah, seems that way RoyK [18:11] any suggestions on how to figure this out? [18:12] which version are you on? [18:12] 9.10 [18:12] that's not supported anymore [18:12] i know :-/ [18:12] I'd upgrade to 10.04LTS [18:12] guess as of last week [18:12] that's where I was trying to decide if 11.04 or 10.04LTS [18:12] there are far more users on LTS releases, so bugs are found and fixed quicker [18:13] oh? Nice to know [18:13] if it's just a server, I'd suggest 10.04LTS [18:13] cool; ty for the suggestion [18:13] * RoyK has 20ish 10.04 servers running [18:13] nice! [18:13] i have closer to 50 servers that will soon be ubuntu [18:13] :) [18:14] a bunch of solaris, ubuntu and centos right now [18:14] I use the latest Ubuntu release on all my personal computers (including my VPS), but for work stuff I stick with 10.04 [18:14] I prob have around 15-20 ubuntu but they are all running 9.10 [18:14] robo: we still stick with solaris^WOpenIndiana for storage [18:14] ZFS FTW! [18:14] We use Netapp here [18:14] heh - that costs a bit more ;) [18:14] yes, yes it does [18:15] I found the answer. motd not updateing so I deleted /etc/motd.tail - problem solved [18:15] cost us some EUR 20k per 100TB box with supermicro hardware and openindiana on the top [18:15] I just inherited all these ubuntu servers. I don't have a clue what i'm doing [18:15] 100TB _net_ storage after a load of redundancy [18:15] load of redundancy? Is that data de-duplication? [18:16] no dedup - that's not really stable yet [18:16] gotcha [18:16] but redundancy as of small RAIDz2 VDEVs [18:16] oh, right [18:17] and with SSDs for caching, the boxes can sustain 10Gbps throughput, which is rather neat for a machine at that price [18:18] s/a machine/machines/ [18:19] so RoyK, any ideas how i can get portmapper to start before nfs? [18:20] i'm not really sure how to adjust this type of thing in ubuntu. I don't think it's as easy as editing files in /etc/rcx.d [18:20] robo: upgrade to 10.04 first - it might be fixed there, and it'll give you updates [18:20] so you think it's just a bug? [18:20] robo: ubuntu uses upstart for most stuff [18:21] robo: yes [18:21] upstart, let me look into that [18:21] robo: it won't take too long to upgrade, though [18:21] is there a recommended way? [18:21] do-release-upgrade [18:21] i was thinking I'd have to build a fresh O/S and move everything over [18:21] oh, nice [18:21] upgrading is trivial [18:22] just check the release notes etc. [18:22] nice! :-) [18:22] and maybe try first with the least important one ;) [18:23] janc that is defently not any fun [18:23] patdk-wk: why? [18:24] there is nothing fun. We use vmware [18:24] to test it? before attempting it on production? :) [18:24] So snapshot -> upgrade :-D [18:24] yeah :) [18:24] patdk-wk: coward! [18:24] if you have vmware you can probably just test on a copy too... [18:25] ya, I'll clone, then test [18:25] but only if I feel something will break [18:25] and my guts don't work most of the time :) [18:25] patdk-wk: for most servers, an upgrade won't even need a clone, so long as you have a snapshot [18:26] well, it depends on how you test [18:26] I'll play with it for a week or so [18:26] depends more of what sort sort of server [18:27] all our servers are N+1. So I snapshot it and pull it out of the F5 pool for testing [18:27] so formal :) [18:27] robo: where do you work? [18:28] in the US for a medium sized corporation [18:28] how specific ;) [18:28] we have cool toys here [18:28] :-D [18:29] F5, Netapp, getting some EMC storage [18:29] vmware [18:30] what's F5? [18:30] firewall/loadbalaner company [18:31] * RoyK has some rather nice toys as well - just setup this new 2U box with four servers, each with a dual 12-core opteron and 64GB RAM [18:31] mine are the other way around [18:31] s/a dual/dual/ [18:31] dual quadcore 5630's with 192gb ram [18:31] and still keep running out of ram :( [18:32] patdk-wk: virtualisation? [18:32] yep [18:32] never really overload the cpu's ever [18:32] what hypervisor? [18:32] esxi [18:32] ok [18:32] mine are just compute nodes [18:33] this is 50+ win2003/8 machines :( [18:33] I'd quit :-) [18:33] have two other clusters that are almost all ubuntu [18:33] setting up mpich2 on them these days... [18:33] dosen't ESXi support memory overcommit? [18:33] royk, yep [18:34] and memory dedup [18:34] and does it nicely too [18:34] for some stupid reason (price, that was, we're an institute, and get rather good prices from MS), my boss chose hyper-v over vmware [18:34] but when each vm is running 4-8gigs of ram [18:34] hyper-v is not for the faint-hearted [18:35] oh, I setup hyperv once, cause well, we had win2008 lic [18:35] the application failed to work properly in a win2003 hyperv guest [18:35] really strange issue [18:35] it was a webserver [18:35] excessive network traffic seems to bring down the VMs on hyper-v [18:35] ya :) [18:36] there's a hotfix available now, though - the MS guys installed that on one of the nodes today [18:36] the hyperv guest would get the http headers, and send a reply, but the reply would never make it to the network wire [18:36] I dumped hyperv after that [18:36] hehe [18:36] between that and lack of memory overcommit [18:37] my san just arrived at my house last week, haven't had any time to set it up though :( === medberry is now known as med_out [18:37] I don't get why MS hasn't figured out memory overcommit [18:37] * patdk-wk starts playing with 8gb fc at home :) [18:37] patdk-wk: probably useful in the winter, if you live in a place where it gets cold :P [18:38] hehe [18:38] I already hack a shelf of servers that are all poe [18:38] thinking about converting them all to san boot, diskless poe test machines [18:38] hmm, not poe, but wol [18:40] \ [18:42] wol? [18:42] Wake On Lan [18:42] k [18:43] but poe can't really drive a PC, can it? [18:43] maybe if it's an ARM blade, but AFAIK those are not available yet ;) [18:44] 25.5w for a PC seems a little low [18:44] RoyK: depends on what "PC" [18:44] for an ARM or Atom, perhaps [18:44] there are x86 SoCs that run at 1 W (GPU included) [18:45] SoCs? [18:45] but I doubt those are used a lot in servers, outside of small home servers [18:45] SoC = System-On-a-Chip [18:46] basically most of your motherboard integrated in 1 chip, like they also use in smartphones etc. [18:46] ah - ok - but I don't think those will be much fun if you try to start a 3D game on them [18:46] well, old 3D games should be possible ;) [18:47] and the x86 SoCs I know have much worse 3D GPU than high-end ARM SoCs [18:48] * RoyK likes ARM [18:48] this guruplug is rather neat [18:48] RoyK: you can play Full HD video on some ARM SoCs nowadays [18:48] I know [18:49] HTC Desire HD with HDMI output.... [18:49] * RoyK goes to find his old C128 [18:49] or a PandaBoard ;) [18:49] but there is a company working on 4-core & more ARM SoCs for blade servers AFAIK [18:50] aren't ARMs rather slow on FP? [18:50] I guess that depends [18:51] also, even if you can pack them more densely than, say, Opterons, there's still a cost of splitting jobs into smaller and smaller pieces === NG_ is now known as ng_ [18:53] ARM just defines the main CPU architecture, and in recent designs they have some SSE-like instructions, but you could probably use specialised co-processors if you need heavy FP (similar to the DSP they use to decode video) [18:53] RoyK: that depends on the type of job of course [18:53] met stuff [18:53] windfield simulations etc [18:54] your users probably know best how well they can parallellize that ;) [18:54] royk, I didn't mean poe, I meant to say wol instead, keep the rack of computers off, unless I need them :) [18:54] :) [18:55] but the ones I'm using are 2.6ghz p4's, and use about 46watts [18:55] except on boot, I guess? :P [18:55] why? [18:55] no harddrives :) [18:55] shouldn't use *that* much more on boot, I think the psu on them is 90watts [18:57] then you are wasting a lot probably [18:57] wasting what? [18:58] 90 W PSU's (and certainly older ones) probably aren't very efficient at 45 W === ng_ is now known as NG_ [18:59] well, lets say it this way [18:59] IIRC most PSUs are most efficient at about 50% load [18:59] I dunno how much power the computer is using [18:59] the PSU draws 45 watts :) [18:59] or was that 30%? [19:00] my watt meter on the 120v line going to the psu says 45watts :) [19:01] how strange of me also, my workstation runs on 230v, but I haven't ran a 230v line to the *rack* [19:02] most PSUs today take anything from 100 to 250V [19:02] yep [19:02] I have always found 120v power annoying, and neutral wires even more annoying [19:03] * RoyK just read about the .jp power grids - there are two of them! 50Hz and 60Hz [19:03] oh? are they converting to 50hz? [19:04] japan has always been a strange mix of usa and eu standards [19:04] the north runs on 50Hz, where the south runs on 60Hz [19:04] both on 110V, though [19:05] I don't see the point of running 110V systems when 380/400V systems are so much flexible [19:07] I don't see the point of 110v since a neutral is a waste of a power wire [19:25] hey guys, anybody had any luck getting any of the newer(SCST,LIO) iscsi targets working?? [19:25] on natty that is... [19:26] I can get SCST running but as soon as I attach vmware to a lun it wigs out... lio, well I cannot even come close to making that one work [19:28] Hi, I'm trying to get dhcpd running. I am getting a permissions error when attempting to create the pid file. See: http://pastebin.com/TRp4TiQC [19:28] aBs0lut30, how about a Linux client? does it react the same way? [19:28] I have done a chmmod 777 /var/run [19:28] removed the old pid file [19:28] and i still get the permission error when I run dhcpd [19:29] ppetraki: havent tried a linux client, but the win7 client seems to work just fine... [19:29] well. lio is new, I don't believe it's suppost to be that *stable* till kernel 2.6.42 [19:29] ppetraki: and I should clarify, I can connect vmware to the target, but as soon as I try and format a lun to VMFS it just hangs the server [19:30] aBs0lut30, so LUN discovery, with any transport, is usually free of block level access [19:30] am a bit curious as to why IETD is still the "default" iscsi package at this point... its a complete POS... IMHO [19:30] aBs0lut30, e.g. connecting and writing to it are two separate things [19:30] aBs0lut30, the win7 client, can you perform IO? [19:31] not much, but I partitioned and formated a lun with no problems [19:32] on the vm side, I do see a TON of command aborts once it tries to start writing to the lun [19:32] aBs0lut30, well, that's sufficient, a handful of reads and writes [19:32] aBs0lut30, so that tells me the Linux iSCSI server side of things isn't disfunctional [19:33] yeah its really odd... [19:33] ohh, there is one more thing... let me pastebin this real quick... [19:33] aBs0lut30, I was just about to ask for logs :) [19:33] http://pastebin.com/4Xszaxs5 [19:33] that is a fault I saw pop up once... [19:34] let me grab some of the regular logs and throw out there as well... [19:34] aBs0lut30, oh, vmware running on Linux, I automatically thought ESX :-p [19:34] well it is ESX [19:34] that fault is from the SCST side [19:35] yeah, that looks real [19:35] http://pastebin.com/cLrBkbB2 there is a snap of the dmesg output from SCST after I connect to the lun [19:36] aBs0lut30, task wakes up to process a command and looks like it dies doing some refcount housekeeping [19:36] aBs0lut30, bug it [19:36] aBs0lut30, sorry :( [19:37] me too :( stuck with the crappy IETD I guess... which sucks cause ESX is killing it like once a day... [19:37] aBs0lut30, IETD? what's that? [19:37] the old iscsi enterprise target... [19:37] aka the iscsitarget package [19:38] I have only started playing with scst last week [19:39] plan on doing a lot of it soon, but didn't plan on much iscsi with it, mainly going use it for fc lun's [19:39] if I could get it to work, SCST looks pretty easy to use... [19:40] and from what I read/hear it works pretty well too... [19:45] aBs0lut30, I know lots of about SCSI, haven't really had the opportunity to play with iSCSI yet though [19:46] aBs0lut30, I can tell you though, that emulating a SCSI target is one of the most difficult things you can attempt in storage [19:46] heh, and I am pretty much the other side of that coin... [19:46] yeah... I can tell, hah [19:47] aBs0lut30, this doesn't sound like a terrible bug, I'm surprised we didn't catch this in regression testing [19:47] oh? scst is included in natty? [19:47] well, wouldnt be supprised to find out its something strange in my setup that is causing it... [19:47] patdk: not really... [19:48] aBs0lut30, if your clients obey the SCSI spec well enough, which I expect as much from VMware, then it's likely the target emulator screwed up somewhere in the translation [19:50] based on what I am seeing that would sound about right... would love to know what the win7 client is doing that lets it work... [19:50] aBs0lut30, that lets the win7 client run up until this point :). put some stress on that with iometer and I think you'll come to the same conclusion [19:50] plus I had to jump through a number of hoops to get SCST to build/load... and one of the patches for the 2.6.38 kernel wont apply correctly... so there is no telling [19:51] aBs0lut30, ooooh, now I get it :) [19:51] yeah, forgot about that patch not going on until just now... === NG_ is now known as ng_ [20:15] adam_g: ARE YOU READY? [20:41] New bug: #778076 in setserial (main) "package setserial (not installed) failed to install/upgrade:" [Undecided,New] https://launchpad.net/bugs/778076 [20:48] SpamapS: regarding bug 280421, did you push my proposed package already? [20:48] Launchpad bug 280421 in ethtool "ethtool not available in post-ifup script" [Undecided,Confirmed] https://launchpad.net/bugs/280421 [20:50] hallyn: no I didn't see it in the queue. I suppose I can sponsor and then approve.. ;) [20:51] SpamapS: no, no [20:51] SpamapS: it's no good, pls reject [20:52] hallyn: there's nothing to reject [20:52] [20:58] SpamapS: new one comin' atcha [20:58] SpamapS: oh, right, that's right. you must have not seen my other irc msg. I couldn't push, no perms. [20:58] hallyn: hurry I want to go eat lunch. :) [20:59] well you can do this after lunch if you like [21:00] http://people.canonical.com/~serge/ethtool-src.tar.gz [21:00] has the src pkg [21:00] a debdiff is fine [21:00] http://people.canonical.com/~serge/debdiff [21:01] alright, will review later.. might be a while. :-/ [21:02] hi: so I installed ubuntu but didn't have a network interface setup in vmware. It's now there but after I reboot it doesn't look like ubuntu detects it (dmesg |grep eth shows nothing and no eth0 in /etc/network/interfaces.) Is there a way to get ubuntu to detect the nic? lspci does show the vmware vmxnet3 ethernet controller [21:03] SpamapS: np, i consider myself done, if you dont' like how i did the changelog pls just hack it right in the debdiff since the two of us seem to have horrible communications latency :) [21:18] to use webmin I need to unlock root? if not what is the default user/passwd after instalation? [21:27] !webmin | Andre_Gondim [21:27] Andre_Gondim: webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. [21:27] thanks === ng_ is now known as NG_ [21:39] lol i put debian server on my old ubuntu server box to try it out and the load is so much less [21:40] load? [21:40] it's 0.01 now but with ubuntu server it was always around 0.80 [21:40] confused :p [21:40] load means nothing :) [21:40] what was running? [21:41] using all the cpu time? [21:42] I just had a clean install and it never seemed to work very well, but debian is much better [21:42] does ubuntu put extra stuff on it that you don't really need? [21:42] The default install is similar for servers [21:42] :) [21:42] if you have a load of 0.8 on an idle system, you should investigate what's wrong... [21:43] it's weird though it only did that with ubuntu [21:43] idle with debian is from 0.00 to 0.01 [21:43] Some of the security features (see https://wiki.ubuntu.com/Security/Features ) do cause some marginal CPU usage increase, but nothing like that. [22:31] does ubuntu use /etc/resolv.conf? I don't see it [22:32] oh, looks like i have to manually create it. All because i didn't have a ethernet card added === cmagina-afk is now known as cmagina