[00:00] ntoombs_: is the web server running? [00:00] i.e. the apache part [00:01] TheEvilPhoenix: Yes I can access it just fine localy [00:02] and is it binding to *:80? [00:02] just to make sure that its reading outside connections [00:03] http://pastebin.com/Rzyukbit [00:04] ntoombs_: can I /query ya for a sec [00:04] yes === sidnei_ is now known as sidnei === utlemming is now known as utlemming_307 === slug is now known as Guest86677 === Datz_ is now known as Datz [02:00] Hi. I'm having problems opeing port 80 for my LAMP server. I have opened port 22 sucessfully for an ssh connection the same way I opened port 80 however, port 80 still remains closed. I have called my ISP multiple times to see if they are not blocking it and they confermed that they are not. The LAMP server is working perfectally on a local connection but no one can see it externally. Can anyone think of a reason why port 80 [02:14] ntoombs: that happened here in Brazil to me, and the isp guys didn't know what they're talking about... turns out port 80 was blocked by them [02:14] ntoombs: have you tried setting the port to 8080, just in case? [02:15] Yea and 8080 gets through. I've gotten throug port 80 before with this isp. [02:15] if you set the port locally and redirected in your router or whatever links you to the external world, it should work [02:15] hm [02:17] to be sure, is there a program i could install that uses a port that I can change to 80? === guampa_ is now known as guampa [03:01] Hi. I'm having problems opeing port 80 for my LAMP server. I have opened port 22 sucessfully for an ssh connection the same way I opened port 80 however, port 80 still remains closed. I have called my ISP multiple times to see if they are not blocking it and they confermed that they are not. The LAMP server is working perfectally on a local connection but no one can see it externally. Can anyone think of a reason why port 80 === Ursinha is now known as Ursinha-afk [05:37] part [05:37] opps === smb` is now known as smb [07:40] set up postfix but it refuses to send mailsto foreign hosts: relay access denied === koolhead11|afk is now known as koolhead17 [07:42] <_johnny> hi, i'm having an odd networkin problem. i've followed this guide http://ubuntuforums.org/showthread.php?t=318539 to get wlan working. and i can do a scan and see all APs, but as soon as a pull the ethernet cable, but ip's are lost and no internet connection [07:42] <_johnny> plug the cable back in, and both ip's are regained, and internet connectivity on both ip's [07:42] <_johnny> would anyone mind having a look at my networking config ? [07:59] Hi] [07:59] Is there anyone who can un-install libvirt? [08:08] Hello [08:10] Is there anyone who can un-install libvirt? Since I have two version conflict [08:21] Anybody in England please ? out of subject question. I'm looking for a dedicated/vps server prvider based in england please. [08:26] th0mz: bitfolk.com and bytemark.co.uk are the two names I usually see for vps. couldn't answer for dedicated === ikonia_ is now known as ikonia [08:34] thanks shauno [08:43] th0mz: I have a VPS with Bitfolk, and I am very happy with them. [08:45] th0mz: Yet, if you want a more "cloudy" solution, where you can spawn new virtual machines yourself etc I guess you might want to go with Linode. In the Bitfolk case there is still the need to talk to support@ when creating a new virtual machines, modifying amount of resources to it, etc. [08:49] to be exact, i'm just looking to a small vps to get an ip based in england to access BBC video ;) [08:50] i just need to run openvpn or kind of proxy... So i'm looking to the cheapest solution. [08:50] i have nothing to host (i'm working in a hosting compagny, all my personal need are hosted here). [09:36] New bug: #809218 in cobbler (universe) "Cobbler should generate a usable /etc/rsyncd.conf file but doesn't" [Undecided,New] https://launchpad.net/bugs/809218 [10:06] New bug: #809239 in dhcp3 (main) "Default interface config is not deleted after upgrade to Natty (dhcp3-server)" [Undecided,New] https://launchpad.net/bugs/809239 === zz_ng_ is now known as ng_ === ng_ is now known as zz_ng_ [11:03] RoAkSoAx: pong, was travelling yesterday [11:12] how does one remove libraries added from make install? [11:12] is it only doable manually? [11:21] rurufufuss: unless you changed the --prefix, they should be in /usr/local... [11:22] rurufufuss: some things have a 'make uninstall', try that first [11:22] rurufufuss: if not, you can just rm the files... just make sure you get rid of *all* the files. [11:22] cool, thanks [12:14] setfacl: Option -m: Invalid argument near character 3 [12:14] setfacl -m u:lighttpd:rx /var/www [12:14] does my command wrong? [12:15] anyone? [12:18] ?????????/ [12:19] Calm down. [12:35] SpamapS: still around? [12:43] hi all, i just found out, that i can start ubuntu server in rescue mode and become root without being asked for a password [12:44] bug or feature? [12:44] does it really matter? [12:45] well, i guess [12:45] anyone with physical access doesn't need a password to copy your drive [12:45] unless it's encrypted [12:45] amelin: that's fine. if you ever set a p/w for root user then it would be different [12:46] pmatulis: ah okay, but there is a root pw set [12:47] patdk-wk: of course it does, i dont want a technican in any center to have it that easy [12:47] :) [12:47] amelin: why did you set a root p/w? [12:48] my bios locks the keyboard, no password, the bios won't accept entry from the keyboard [12:48] * patdk-wk always sets a root password [12:48] you can get into rescue mode as root even if you set a root password [12:49] patdk-wk: if my system fails, i dont want that everybody can access my system, okay, maybe a technican call me sometimes an tell that he already fix it [12:49] Remove the entry from GRUB and add a GRUB password if you're really paranoid about it. [12:49] if you want to protect data against attackers with physical access, you need to do disk encryption [12:50] maswan: but doesn't it ask you to enter that p/w? [12:50] pmatulis, ideally, yes [12:50] otherwise local users would still have access [12:50] pmatulis: depends on which rescue mode you go to. [12:50] booting init=/bin/bash etc won't care about root password being set [12:51] maswan: sigh, i'm not talking about that stuff [12:51] pmatulis: then what are you talking about? I thought we were talking about booting rescue modes with root password set or not [12:51] maswan: i'm talking about the official rescue mode [12:51] maswan: boot cd and choose from menu [12:52] pmatulis: oh, external boot, then it is up to whatever you boot if they want to care about that. [12:52] pmatulis: so no added security from a root password [12:52] maswan: well you're confusing my conversation with amelin [12:53] amelin: what did you do to get into your rescue mode? [12:53] pmatulis: reboot, open grub menu, choose rescue mode, choose the shell menu [12:54] okay, i choose the shell option to become root [12:54] so after reboot i choose rescue in the grub menu [12:54] amelin: and you can get root access w/o p/w even with a p/w previously set? [12:54] pmatulis: yes [12:55] amelin: k, guess i'll need to go over that scenario. i didn't think you could do that [12:55] sounds resonable, one of the uses of rescue mode would be to recover from forgotten passwords === Ursinha-afk is now known as Ursinha [13:02] good morning [13:04] it's difficult to work around root from a local boot. if they edit at the grub menu (or provide their own boot media), they can just use init=/bin/sh to bypass *everything* [13:05] so you've bot bios-level & physical defences, and encryption. anything past that is a false sense of security [13:06] this is what locked racks or cages are for [13:08] jeah, okay, maybe thats right, maybe i was just a bit to confused :-) [13:09] SpamapS: howdy! yeah I heard... we were wondering how did you get ensemble + cobbler to work as we were unable to [13:29] smoser: ping [13:29] Ursinha: bom dia [13:30] lynxman: bon dia :) [13:31] Ursinha: ooh very good ;) [13:31] :D [13:39] === manjo` is now known as manjo [13:54] hello [13:54] could I get help with bind here? [13:55] https://help.ubuntu.com/community/BIND9ServerHowto [13:55] according to it, I can configure bind9 as hybrid cache and master, then could I achieve it? [13:55] Did you look in the server guide? [13:56] It discusses bind9 configuration. [13:56] that's it? [13:57] RoAkSoAx, in now [14:02] Oh, good grief. [14:04] van7hu: If you aren't willing to look at the existing documentation, I think it's unlikely you'll find the level of help you want for free. [14:05] ScottK, okay [14:05] zul: Do you have a stack of changes to the glance packaging that you forgot to push? [14:06] soren: no [14:06] soren: i dont think so [14:08] zul: There are *5* uploads to the archive that aren't in bzr. [14:08] soren: frig [14:09] zul: One of them even has changes from two people. How did you pull that off? [14:09] soren: ill clean it up today [14:09] Where are you keeping this stuff? If yo uhave changes from two people, you must be sharing packaging somewhere. Where? [14:09] im not [14:10] it was done locally and uploaded [14:10] as i said ill clean it up today [14:10] Ok. [14:15] zul: Before you actually do fix it, I'd like to hear what you intend to do. [14:15] zul: Whenever. [14:36] RoAkSoAx: there was a lot of stuff necessariy [14:38] SpamapS: do you think you can write a small howto for me/us to be able to do it? [14:38] RoAkSoAx: are you trying to do it w/ my branch or something else? [14:39] SpamapS: we (smoser and I) were tryiung to do it with your branch without success [14:39] SpamapS, woohoo [14:39] RoAkSoAx: did you setup a webdav? [14:39] we did not. [14:39] that part is a pita [14:43] it would probably help if you guys just said where you're stuck [14:43] jamespage: ping [14:47] RoAkSoAx: how is it going? [14:48] Daviey: its going we are currentlyu trying to get somethjing to work with smoser [14:49] something? :) [14:49] SpamapS, we're trying to be secretive, and not tell you things so you can't laugh at how incompetant we are :) [14:49] lol [14:50] but, spamaps, if you could write down everything you need to do to get it functional , then that would be great. [14:50] smoser: good idea, I am a heartless bastard that way [14:50] s/that way// [14:50] lynxman: 10 mins [14:50] cobbler-devenv at lp:~smoser/+junk/cobbler-devenv/ [14:50] jamespage: k [14:50] is pretty close to creating the clean environment [14:50] you coudl start there. [14:51] smoser: uh.. install cobbler, setup a writable webdav at /formulas .. [14:51] smoser: oh I bet a full environments.yaml would help [14:51] yea. that would help [14:51] and "setup a writable webdav" is not dummy proof. [14:52] and i think that you need to have an ensemble agent running on the cobbler === medberry is now known as med_out [14:59] smoser: http://paste.ubuntu.com/642690/ [15:00] I'm having difficulties that apache suddenly cannot access the /testing subdirectory of my web server... can someone help me out with this? [15:00] smoser: http://paste.ubuntu.com/642691/ === utlemming_307 is now known as utlemming [15:12] stgraber: is the lxc srcarch +=arm patch from you? [15:12] just wondering whether to put your signed-off-by on it [15:12] recon i could check the bzr tree [15:12] lynxman: back now - wassup? [15:14] jamespage: hey === zz_ng_ is now known as ng_ [15:28] hmm, anyone here use zfs on ubuntu server? I've currently installed OpenIndiana, but I'm quickly starting to dislike the lack of packages for just about everything I want to run on the server. [15:28] How's btrfs? Can it now reliably fsck? [15:28] s/reliably/do/ [15:29] I do see this, at least: http://www.mail-archive.com/linux-btrfs@vger.kernel.org/msg10864.html [15:30] DanaG: from what I understand, its still experimental. :( === The-Comp1ler is now known as The-Compiler [15:34] "it" being btrfs? [15:37] yes [15:37] Though a few distros are pushing harder to make it the default [15:40] stgraber: nm, found the original author :) [15:47] hmm, installing ubuntu-server to an 8GB usb stick, in a 1.5GB-RAM VM. [15:47] The system it'll be running on has 5GB RAM. [15:47] Do I really need swap, or is swap a bad idea? [15:48] altijd swap [15:48] DanaG: swap is not great in a vm [15:48] I've made a rawdisk vmdk for the flash drive. [15:48] And the VM is just for the install, not the final place it'll run. [15:48] DanaG: if you have enough ram then don't use swap [15:49] DanaG: it won't run in a vm? [15:49] No, I'm just doing the install in a VM. [15:49] So then I can take the drive home and plug it into the microserver. [15:50] DanaG: swap can be nice when something goes bad - it can also be a nice way to swap out things that aren't in use, especially on desktops and when hosting large applications, VMs etc [15:50] There's no way the installer need more than 1.5GB, right? [15:50] Well, the install, and post-install upgrades. [15:51] DanaG: setting /proc/sys/vm/swappiness to 100 will make linux swap out earlier, thus keeping more memory available for useful stuff - a lot of allocations are for memory that's not in use, and that's important indeed for VMs [15:51] guys having a major issue with the server here [15:51] !ask [15:51] 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. :-) [15:51] mysql, apache stopped because apparently / is full [15:51] however, du -chs shows this: [15:51] Filesystem Size Used Avail Use% Mounted on [15:51] /dev/sda3 108G 103G 0 100% / [15:51] ??? [15:52] BuenGenio: usually 5% is reserved for the root account [15:52] BuenGenio: you should never fill up a file system that much anyway [15:52] i don't know what did [15:52] probably a runaway log [15:52] but I'll check [15:52] need to start mysql first thing [15:53] you can change the amount of reserved space with tune2fs -m [15:53] man tune2fs [15:53] setting that to zero is usually a very bad idea for the root === med_out is now known as med [15:54] BuenGenio: try to reclaim some space quickly with 'sudo apt-get clean' [15:54] Then you can install ncdu -- ncurses disk usage tree thingy. [15:55] DanaG: installing unnecessary s/w on a system with a space problem? === med is now known as medberry [15:55] it's some log somewhere [15:56] pmatulis: it's a small tool, and it'll tell you where all the space went. [15:56] But yeah, check logs first. [15:56] BuenGenio: first get space with my command and then 'sudo du -sh /var/log/*' [15:58] That ncdu package is just 92 kilobytes, by the way. [15:59] no, wait, 22. [16:01] RoAkSoAx: so , did you guys still want a howto first? I feel it will be faster if you just show me whats broken === serge_af1 is now known as serue [16:04] hi all [16:07] apt-cache showpkg moodle shows mysql-client as deps but does not install it. Using dbconfig-common for installation throws error saying mysql-client needs to be installed before this process to finish. [16:07] do i need to file a bug for this? [16:09] koolhead17: postgresql-client | mysql-client [16:09] koolhead17: if you already have postgresql-client installed, that dependency is satisfied [16:11] SpamapS, http://pastebin.com/sxscTTyU this is what getting installed once i do apt-get install moodle [16:12] i can see postgresql-client getting installed with it [16:12] koolhead17: yeah, that just means that the moodle maintainer prefers postgres over mysql [16:12] koolhead17: if you want to be explicit.. just do apt-get install moodle mysql-client [16:12] koolhead17: you can just install both also.. since you're writing a formula.. eventually you should support both. [16:13] SpamapS, so its not a bug, just a confusing dbconfig-common configuration of moodle where once clicking on mysql as db option ceases the deployment [16:14] koolhead17: dbconfig-common is basically the predecessor of ensemble. ;) [16:14] koolhead17: it is to ensemble as neanderthal man is to ensemble .. a fork along the same chain that will become less relevant as ensemble becomes more popular. :) [16:15] dbconfig is evil [16:15] SpamapS, :D [16:15] zul, +1 [16:15] am still stuck with phpmyadmin preseed [16:15] SpamapS: yes it would be helpful for us here at the sprint [16:15] SpamapS, its just am not interested in paying amazon to test the preseed stuff :D [16:16] before putting it on ensemble :P [16:16] * koolhead17 bows to RoAkSoAx [16:17] hmm, anyone use the kq Linux zfs? [16:18] Random: http://www.dailymail.co.uk/news/article-2011051/Black-macaque-takes-self-portrait-Monkey-borrows-photographers-camera.html [16:20] /20 [16:21] * Ursinha fails with irssi [16:23] DanaG: that's off-topic, kindly refrain... [16:24] SpamapS, can i atleast request for showing some love to mysql :P [16:25] anyway, I think I'll try ubuntu-server with zfs. [16:26] zul: there? [16:28] -ETOOMANYNICKSWITHS [16:28] SpamapS: so [16:28] SpamapS: I'm fixing the old SRU reports script, so you can have it for now [16:28] Ursinha: thats great! [16:29] SpamapS: I'm sorry I've been quiet these days, I'm bootstraping yet [16:29] ::) [16:29] Ursinha: I was hidden away in the mountains all last week.. AFAIK, you have been shouting with a bull horn all week. ;) [16:30] haha [16:30] hmm, in Linux, how significantly much better is an Intel NIC than a Broadcom 5723? [16:32] you can't compare them like that [16:32] and that is a very old broadcom [16:33] That's what's in the HP microserver I have. [16:33] I have that, and a 5721 PCIe 1x, and one of these: http://www.newegg.com/Product/Product.aspx?Item=N82E16833106033 [16:36] odd, the netextreme ii's have lower numbers [16:37] heh, that is a budget intel card [16:37] that broadcom should be much nicer than that [16:37] Really? Interesting. [16:38] why? you just picked the lowest cheapest thing intel makes [16:38] and think it will perform? [16:39] I have an 82574L somewhere else (not sure where), and it worked pretty well, I believe. [16:40] http://www.newegg.com/Product/Product.aspx?Item=N82E16833106011 [16:40] that is more rated along that broadcom chip [16:41] Some peope in #openindiana suggested that Intel would be better than Broadcom. [16:41] I'll have to try the Broadcom again. Maybe it's just their broadcom driver that's the problem. [16:42] dunno [16:42] never had an issue with broadcom nic's myself [16:42] zul: could you push http://people.canonical.com/~serge/lxc_0.7.4.2-0.3ubuntu3-pkg/lxc_0.7.4.2-0.3ubuntu3.dsc ? [16:44] jamespage: howdy, around? [16:45] kirkland: yep === koolhead17 is now known as koolhead17|afk [16:55] well, that's weird... I boot a server (a spare one, not the microserver that's at home), from a USB stick... [16:55] The screen is blank for like 70 seconds... and then pops up, fully booted. [16:55] This is without "quiet". [16:57] danag, normal issue, grub put it into video mode card doesn't like or something [16:57] easy enough to fix [16:57] It was fine at grub, actually. [16:58] Or do you mean the interaction between grub and kernel? [17:01] I'll try grub console instead of gfxterm. [17:03] Nope, still blank, but this time with a blinking cursor. === jMCg is now known as igalic [17:04] Before doing much troubleshooting, I'll have to try it in the system I'll actually want it in. [17:11] SpamapS: ping [17:12] Any of you use zfsonlinux? [17:12] kirkland: pong, on a call but wassup? [17:13] SpamapS: forgive a dumb question ... [17:13] SpamapS: but i'm spinning my wheels here :-) [17:13] SpamapS: what's the best way to see the current list of formulas that have landed in principia? [17:14] kirkland: *good* question.. I was wondering the same thing. :-P [17:14] /7/7 [17:14] argh [17:15] yuck, lots of oopses at __raw_callee_save_xen_make_pte+0x11/0x1e [17:15] SpamapS: :-P [17:16] SpamapS: we need the equivalent of "apt-cache search" [17:17] kirkland: Yeah, a lot of that is delayed as we wait for ensemble to have a proper repo. [17:17] kirkland: we could certainly hack it in. [17:25] kirkland: http://code.launchpad.net/principia has a lot actually [17:25] kirkland: but has stuff in dev too [17:26] SpamapS: hmm, hard to tell what's in, and not [17:26] SpamapS: so, for instance, I'm trying to tell if my "musica" formula has been accepted ... how do i do that? [17:27] kirkland: if there's an lp:xxxx branch its "accepted" [17:27] kirkland: but I think it also needs to go into the mrconfig [17:28] kirkland: so 'principia getall' grabs it === Ursinha is now known as Ursinha-lunch [17:40] SpamapS: regarding lp:~clint-fewbar/ensemble/lxc-container/, are you still ok keeping that in the ensemble source, rather than the lxc source? [17:40] serue: I am not sure where exactly it belongs actually. [17:41] serue: I think there's value in distributing that with lxc [17:41] SpamapS: oh what am i thinking. let's discuss it at the sprint [17:41] SpamapS: since zul seems to not be around, do you mind sponsoring my little lxc cleanup? [17:41] serue: especially if we can genercize it enough where one can just say 'lxc-create -t uec -d oneiric -a amd64 ...' [17:42] SpamapS: i don't want to make the templates more complicated... [17:42] oh [17:42] serue: I 'spose I understand [17:42] yeah, but you can just have ensemble do that, [17:42] serue: do what? [17:42] just install it at /usr/lib/lxc/templates/lxc-uec [17:42] then you can do lxc-create -t uec [17:43] serue: I see value in people being able to boot the same image that is used on the cloud in lxc. [17:43] sure [17:44] i prefer ensemble isntalling the template right now in case the template is going to change a lot for awhile [17:44] serue: that sounds good then [17:44] cool [17:44] SpamapS: so do you mind pushing http://people.canonical.com/~serge/lxc_0.7.4.2-0.3ubuntu3-pkg/lxc_0.7.4.2-0.3ubuntu3.dsc ? [17:46] serue: is that attached to a bug? [17:46] no [17:46] serue: you need to run clean before building that source [17:47] --- lxc-0.7.4.2/debian/lxc.debhelper.log 1969-12-31 16:00:00.000000000 -0800 [17:47] +++ lxc-0.7.4.2/debian/lxc.debhelper.log 2011-07-12 08:00:54.000000000 -0700 [17:47] hmm, is ebox supported, or is it known to break things? [17:47] !ebox [17:47] zentyal is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/Zentyal (Project formally known as eBox - including in Lucid/10.04). [17:47] DanaG: ^ [17:48] i don't see a .log file. but i guess i tarred it on the other laptop. sigh. [17:49] Does ebox run its own ldap server, or can I do it without ldap? [17:49] serue: why aren't you using bzr-buildpackage ? [17:49] oh, I see... usersandgroups is not just for local groups. [17:49] or bzr-builddeb, whatever its called [17:49] It's explicitly for managing ldap. [17:50] spam i use debuild -S -sa [17:55] SpamapS: (jinkeys, network is killing me) i don't use bzr bd bc I don't use bzr right now bc the bzr trees are always out of date. [17:56] I know that's been recently addressed, and I need to try going back to a udd dworkflow [17:56] spam anyway, I thought that debuild -S did a debian/rules clean. but i see the .log file. huh === igalic is now known as jMCg [17:57] * serue wants to smack irssi for not expanding SpamapS [17:59] serue: why are the bzr trees getting out of date? [17:59] serue: we can fix that usually by importing the versions that have failed [18:01] SpamapS: dunno. i don't think they had failed per se [18:04] does anyone know what's the hyphen sign in front of log filenames in rsyslog,conf mean? [18:05] eg mail.* -/var/log/mail.log [18:06] SpamapS: making a f'ing bzr tree [18:11] quilt patches applied [18:11] yay, i get to start over. [18:12] sometimes that's a good thing ;) [18:13] hmm, looks like ebox is more than I need. [18:13] all I'd really want is something like what FreeNAS gives. [18:28] SpamapS: well, i don't understand it. New version, with source at lp:~serge-hallyn/ubuntu/oneiric/lxc/lxc-0.7.4.2-cleanup-patches and pkg src at http://people.canonical.com/~serge/lxc_0.7.4.2-0.3ubuntu3-pkg created with 'bzr bd -S -- sa', the debhelper.log files are still there [18:28] (but not in the bzr tree) [18:29] serue: *weird* [18:30] not like it's a helpful log either :) [18:34] SpamapS: eh well, pls push if you feel comfortable. bbl. [18:39] serue: will take a look ASAP.. need to go find some food [18:43] SpamapS: take out spear? [18:45] pmatulis: the only thing I'll be able to hunt and kill with a spear in west L.A. is Chihuahua .. and then I'd have to deal with angry UCLA juniors and their giant pink purses. [18:46] SpamapS: LOL (at chihuahua), still reading [18:46] SpamapS: wow, another LOL [18:48] has anyone heard of idmapd spontaneously stop running? have natty desktop client randomly getting the nobody-user using NFS/autofs with solaris server [18:49] * SpamapS wonders if Chipotle would be interested in some locally sourced Canine.. [19:20] hello to all. i am trying to get the modrewrite loging to work on my Ubuntu Server maschine but i fail now the whole day to get this to work. [19:20] Anybody can give me a advice how to enable logging for modrewrite ? [19:23] is the php suhosin patch installed by default in lucid? do i need/want the php5-suhosin package in apt? [19:24] using iscsiadm how can I detach from an iscsi server? === ng_ is now known as zz_ng_ [19:24] zeknox: --logout i think [19:25] pmatulis: thanks buddy! [19:26] Daviey: any update or new comments on spice upload? [19:34] Has anybody tinkered with network interface bonding? I'm curious about a few things and was hoping someone could clear the air... [19:36] works great for me [19:36] what does? [19:37] i guess bonding [19:37] wasn't sure if that was directed at me or not [19:37] what's your problem with bonding? [19:38] well I found a guide for it, but I wanted to talk to someone who knows a lot about it to help me understand it a bit more. [19:38] how about we start simple [19:38] question? [19:38] I'm running edubuntu 10.04 on a server with 4 gigabit network ports, I want to combine them into 1. [19:38] would that be the broadcast method? [19:38] none :) [19:38] what is the goal? [19:38] what are you attempting to protect? [19:38] to make all 4 NICs look as one [19:38] I'm just trying to maximize throughput [19:39] you can make all 4 look like 1 several ways [19:39] not all of them will maximize throughput, infact most won't [19:39] cause maximizing throughput over 4 nics is not easy to do [19:39] well, here's my dilemma so you know what I'm doing [19:39] I'm running an LTSP server with thin clients. [19:39] I previously ran a server with a single gig port to a lab of 30 clients without issue. [19:40] this year we're moving to 60-70 clients, so naturally my thought process is to utilize the other nic's [19:40] to help out with the load, ya know? [19:40] I've always set up LTSP on smaller scale setups, so bonding was never a necessity. but this time I think it'd be wise to utilize bonding, based on my client load this year as well as the benefits I read about. [19:40] what do you think would be best? [19:41] roasted: the best option is if you can afford a managed switch, use IEEE 802.3ad Dynamic link aggregation [19:41] well, I'm here in our server room, we have some decent gear [19:41] let me see... [19:41] dell powerconnect 6248 [19:42] wish I knew what that was, non-rebranded [19:42] hmm [19:42] yup [19:42] so my optimal solution would be to work at the switch level and let ubuntu alone? [19:42] no, you configure both ends [19:42] ahh [19:42] I thought you meant one or the other. [19:42]