[00:02] Anybody here who could tell me if ubuntu can use fibreoptic cards? To be specific, lspci gave me this : Fibre Channel: Brocade Communications Systems, Inc. 415 4Gb/815 8Gb PCIe FC HBA (rev 01)... I have multiple of these in a server and must get them to work as one using powerpath.. Anybody who might be able to give a hand? [00:04] sven_oostenbrink: if they are supported by linux, they should work with Ubuntu ;) [00:05] taking the kernel version in account of course [00:05] JanC: But how do I know if they are supported by Linux? How can I see if the card is working? [00:06] sven_oostenbrink: try the live-CD ? [00:06] JanC: well, the server is already installed with ubuntu [00:06] JanC: so I can just do anythign needed to get them to work [00:08] so it doesn't work OOTB... [00:08] ? [00:09] JanC: Nope.. I did just find drivers on the brocade website (http://www.brocade.com/products-solutions/products/fabric-management-software/product-details/efcm/specifications.page) but these seem Redhad or Suse only.. [00:10] Any chance I could utilize those for ubuntu server? [00:10] JanC: AFAIK, the cards are not yet recognized by the kernel, but Im not sure if this is the case.. [00:13] sven_oostenbrink: did you try with a recent karmic version? [00:14] judging from that site, their linux solutions are from the last decade anyway :-( [00:25] JanC: I know.. Im right now downloading the driver for 9.04.. crossing fingers and toes here.. [00:48] Does cron recursively search cron.daily? If I put a symlink into a directory with a bunch of scripts I want to run, would it run them? [01:04] Hey guys I downloaded the latest build of ubuntu server (9.10 from today) and I am getting an error on install [01:04] it wants drivers for the CD drive it is loading off of, or if I try to install off of a flash drive I get the same error. [01:04] I don't even want to have the CD drive in the system after install. [01:05] any suggestions? [01:26] New bug: #455931 in libgoogle-collections-java (main) "google-collections update breaks eucalyptus build" [Undecided,New] https://launchpad.net/bugs/455931 [01:26] hey gang, im trying to complie imagemagick from source and not sure how to get it to see the requisite libs for some components [01:27] like png for example [01:27] i just did, apt-get install libpng3 [01:27] and added --with-png=yes even, however the configure report is not showing that it will be used [01:28] i think i have to tell it the path to the libraries for the linker, but not sure where they are installed when i apt-get libpng3 [01:29] should i build libpng from source? i thought the point of the package was so i dont need too .. === The2morrowMan_ is now known as The2morrowMan [01:33] kirkland: hey! [01:33] kirkland: doing some testing with raid1 === freeflyi1g is now known as freeflying [01:33] kirkland: I've booted one of my guest with only one drive (after a successfull install) [01:33] kirkland: then restarted with the other drive only [01:34] kirkland: logged in and created a file in my home directory [01:34] kirkland: then restarted with both drives [01:34] kirkland: the system booted with a functional raid array [01:34] kirkland: and the file in my home directory went away [01:35] kirkland: there were bunch of errors one the last boot though [01:35] kirkland: what would be the expected behavior in that use case? [01:44] Is Ubuntu 9.10 supposed to bring anything really nifty to the server arena, besides cloud computing integration? [01:53] libpng12-dev : theres the package ;) === SyL is now known as Guest68051 [02:41] hello i have a question hoping that you would help me answer, i dont have any idea about server administration, and im planning to buy a vps , which os should i choose ubuntu or centos, im quite familiar with ubuntu as it is what im using for my desktop, but its different when you talk about server side, which is easy to configure cent os or ubuntu server? [02:42] ramseize, since you use Ubuntu at home, you'll be more used to Ubuntu as a server too. [02:42] ramseize, anything you do with your Ubuntu Server you'll be able to do with your Ubuntu desktop. [02:43] Hi! I just moved from Fedora 7 to Ubuntu Server 9.04. I noticed that the remote desktop locks up the server (I installed the ubuntu-desktop to make it easier for me to administer until I'm more comfortable with the terminal-only commands). Does anybody know of a fix for this? [02:44] but i read some great comparison with centos and ubuntu , but i am more familiar with ubuntu so you think its a great advantage for me to use ubuntu? [02:45] does anyone of you know of a good vps hosting but very cheap, i just need it foe education purposes [02:45] ramseize: This is #ubuntu-server, so most people here are going to recommend Ubuntu :) It is (generally) an advantage to use something you already know -- it means you have less new stuff to learn. So if you already know Ubuntu, especially Ubuntu at the command line, I suggest you use Ubuntu server. [02:46] ramseize: you should check distrowatch and possibly ##linux is a better place to ask [02:46] ok [02:46] thanx [02:46] however i personally recommend debian for servers [02:47] centos is very popular too [02:47] i dont really like the ubuntu practice of installing everything and the kitchen sink [02:47] ok [02:47] As for VPS recommendations, many Ubuntu folks are happy with LiNode http://www.linode.com -- lowest size VPS there is US$19.95/month [02:47] then turning on every service known to man [02:47] its better on the desktop but it puts you at a serious disadvantage security wise [02:47] some other distros install and turn on much less by default [02:48] however there is also a cost for that [02:48] med\weed: Last I kknew Ubuntu server defaults to a minimal install.... [02:48] the cost in this case is additional install / configure time [02:48] so what os should i choose? ubuntu or cent os? [02:48] jmarsden: thats not a REAL minimal [02:48] yeah...very minimal! [02:48] ramseize: do you run ubuntu on your desktop? [02:48] if so just start with it [02:48] its _most_ important to learn! [02:48] med\weed: what unnecessary services are installed? [02:49] med\weed: What services does Ubuntu server install by default that you consider to be a security risk? [02:49] jcastro: i'll leave that as an exersize for you [02:49] +sp [02:49] yes i run ubuntu in my desktop [02:49] well, you're the one making the claim, isn't the onus on you? [02:49] but if you want to really get down i would go so far as to say a centos default install is much more secure by default no matter what is installed. [02:49] as they turn a firewall on by default [02:49] * skuld takes a number and waits in line for the current conversation to conclude... [02:50] oh im done with it [02:50] I just moved from Fedora 7 to Ubuntu Server 9.04. I noticed that the remote desktop locks up the server (I installed the ubuntu-desktop to make it easier for me to administer until I'm more comfortable with the terminal-only commands). Does anybody know of a fix for this? [02:51] med\weed: rapidswitch.com like these kind of vps hosting, [02:53] * skuld hears the crickets chirping outside... [02:54] I also ran the updater and the server claims it's 'everything up to-date'. [02:55] skuld: Ubuntu server doesn't come with a GUI desktop, if you install one you are pretty much on your own regarding support for the resulting server/desktop hybrid setup you have created... you may just want to install Ubuntu desktop, unless this is truly a production server the differences are unlikely to be signifcant to you. [02:56] I was opting for the server because of the built in mail and web server applications. I'm really not that familiar with the terminal only style of the server though [02:56] it is supposed to be a production server, and after fighting for over a week trying to get mail to work I decided to install a fresh OS [02:57] (mail was working fine for over two years before-hand...then I guess I changed something or ran some program that changed something) [02:58] skuld: You can use tasksel on a desktop install to add mail/web/whatever, just like you can on the server installation... no difference there at all. [02:59] If you prefer, set up a fresh Ubuntu server with no GUI and ask for help with the mail setup here or in #postfix and I'll do what I can to help... I've been doing paid network admin work on Linux for over 15 years... [03:00] okay. really quick, is there a remote desktop for the terminal-only server? [03:01] my server is in an awkward location to work on so remote capability is really handy [03:01] Remote desktop? You can run GUI programs on the server over SSH, and remote server admin is usually done at the command line in a shell anyway... [03:01] If you need a web interface to it, there is ebox... [03:02] !ebox [03:02] ebox is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/eBox [03:02] okay. I'll see about reinstalling server with no gui [03:03] skuld: OK. The only "extra" thing you really will want to install at the initial installation time is openssh, so you can SSH into the server. Everything else (mail, web, etc) can be added later. [03:05] can I use 'ssh' with a windows workstation...that's the PC I work remotly from...hence why I was using the gnome desktop with the built in remote desktop [03:06] You can install Putty which is a well known free SSH client for Windows, and use that, yes. [03:06] If you don't know this already you are probably not really ready to be a sysadmin of a production Linux server, though... [03:07] okay cool. I'm just not really all that comfortable with the command line yet...it's been years since I've worked with it [03:07] OK, I can point you at some guides to try and get you up to speed on the shell if you need that. [03:08] http://tldp.org/LDP/intro-linux/html/intro-linux.html is one [03:09] could I just 'uninstall' the ubuntu-desktop and gdm to revert back to a terminal only server, or would you recommend just a full reinstall? [03:10] You can try that, and if it fixes your issue with hangs, that's fine. I'm not sure what causes the hangs, so that worries me a little... but go ahead and try it. [03:10] skuld, you could purge it [03:10] simply uninstalling it would leave a LOT of unused config files [03:10] purge? [03:11] definately [03:11] sudo apt-get purge PACKAGENAME [03:11] apt-get --purge remove [03:11] however there is a caveot with that [03:11] Rather than sudo apt-get remove PACKAGENAME [03:11] it also purges regex'd packages - there can be differences in what it takes out iirc [03:11] jmarsden: your wrong [03:11] my form is correct sorry ^^ [03:11] med\weed: I'd take Ubuntu over Debian since Debian lacks most of the proactive security features in the built packages and kernel. [03:12] so I'd do: sudo apt-get purge ubuntu-desktop gdm [03:12] med\weed, both ways work [03:12] purge is now a valid option [03:12] sudo apt-get remove --purge PACKAGENAME [03:12] what@karmiclyubu:~$ sudo apt-get purge remove gedit [03:12] [sudo] password for what: [03:12] Reading package lists... Done [03:12] Building dependency tree [03:12] Reading state information... Done [03:12] E: Couldn't find package remove [03:12] sudo apt-get purge PACKAGENAME [03:12] i.e. your wrong. [03:12] --purge [03:12] not purge remove, just purge [03:12] e.g. --purge remove or just purge [03:13] qman i just pasted proof [03:13] read what you did [03:13] med\weed: I'm wrong... sure... I do system admin work for a living... do you? You mistyped the command... [03:13] sudo apt-get purge remove gedit does not work [03:13] sudo apt-get --purge remove gedit does work [03:13] that's because that's wrong [03:13] Right, and that is not what I suggested. [03:13] that was the whole problem i had with their examples ffs [03:13] look start reading its irc but i' [03:13] you don't "purge remove", just "purge" [03:13] ll paste it for you one sec [03:14] 02:11 < jmarsden> sudo apt-get purge PACKAGENAME [03:14] 02:11 < jmarsden> Rather than sudo apt-get remove PACKAGENAME [03:14] Could you use a !pastbin this time? [03:14] * skuld sudo apt-get purge ubuntu-dekstop gdm ? [03:14] skuld, yes [03:14] med\weed: Do you see the string "purge remove" in what I typed? [03:15] med\weed: Do you see the string "purge remove" in what you used as "proof"? [03:15] med\weed: Do you understand the difference now? [03:17] ummm...I *thought* it was 'ubuntu-desktop' but when I try running the command, it states that it can not find the package by that name [03:17] skuld, that is the correct name of the meta package for the ubuntu desktop system [03:17] so unless you installed it manually, that should work [03:18] I used apt-get install [03:18] to install it [03:18] yes, but if you installed, say, gdm, and all the other programs you wanted to use by name [03:18] instead of installing ubuntu-desktop by name [03:18] skuld: Typo? Earlier you wrote: sudo apt-get purge ubuntu-dekstop gdm ? # This misspells desktop [03:18] that won't work, you'll have to go back and uninstall all the ones you installed [03:19] oh and me an my stupid fat fingers...I mispelled desktop [03:19] hah [03:20] okay, it says it purged. rebooting the server now [03:23] okay, that worked. terminal only now.... [03:23] what's the package for ssh so I can work remotely? [03:24] sudo tasksel install openssh-server [03:24] I'm trying to set a couple of things on my site and need some help. I'm running my own server and need to set a couple of things, but don't know how. I need to set extra perameters for default pages to load in folders ie "index.html vs index.htm" [03:25] it's not apt-get? [03:26] Pairadimesitty: Use the Apache DirectoryIndex directive to provide the list of index pages you want your server to look for. [03:26] in code please, I'm a total noob, don't know the language [03:26] skuld: You could do that, sure. Learning tasksel prepares you for installing web and mail tasks, though... [03:26] jmarsden: very well your right [03:27] med\weed: Thankyou. [03:27] however. [03:27] i also do sysadmin work for a living. [03:27] not knowing one single command is not enough to break that out :P [03:27] med\weed: OK, then perhaps you should read man pages a bit more carefully :) [03:27] [we're currently using 70gbits] [03:27] cool! it installed, now I use putty to see my server screen? [03:28] jmarsden: more like i should review them occationally as apparently there's value added in linux too ;) [03:28] skuld: You use putty to log into the server, yes. [03:28] jmarsden: if it makes you feel better you probably predate me in linux :P i'm only on year 6? 7? [03:28] since sarge was unstable [03:28] actually I'm running my server on Xubuntu to help my lack of directory and terminal knowledge, so if you could just point out a file to edit that would help too [03:29] jmarsden: btw if your wondering wtf i'm doing here - its because my co runs ubuntu lol [03:30] oh crap! I need to add port 22 to my firewall [03:30] Pairadimesitty: Look at /etc/apache2/mods-available/dir.conf [03:31] skuld: good practice to chang that port in /etc/ssh/sshd_config [03:31] s/chang/change [03:31] jmarsden: thanks, on it. [03:34] I need to go AFK briefly... skuld, do you have Putty connecting to your server yet? [03:34] I'm still trying to get it connected [03:35] What is the problem? You firewalled off your server from your admin workstation? :) [03:35] OK, well, I'll be back soon I hope... [03:36] okay, I changed the port, but how do I open my new port...say it's 553 (it's not) [03:37] I was running firestarter back when I had the gnome desktop installed. now I don't have that utility, so I need to know how to edit my firewall via the terminal. LOL [03:39] jmarsden: if you're still here thanks that did the trick. [03:42] skuld: what firewall? [03:42] The other thing I'm trying to do is integrate my forum inside my home page (/forum.php inside index.htm), but iFrames universally suck, I can't get it to resize automatically inside my content. I'm told to use ssi, but that doesn't seem to load, [03:43] what forum software are you using? [03:43] google also tells me that I need to set the server to accept htm format for it rather than only shtml [03:43] I've been admining my own for 7 years now [03:43] smf [03:43] perfect...that's what I use [03:43] so far I really like it [03:44] if only I had my server up...I could *show* you. [03:44] thanks skuld, that's exactly who I need is your help [03:44] you don't NEED to use SSI, that's just a bunch of functions that make things convienant [03:44] um [03:44] yeah [03:44] I used some, but mostly I coded my own "ssi" using PHP [03:44] iframe shows the content, but doesn't play nice with sizing [03:44] creating the HTML and mySQL calls that way [03:45] what is it you're wanting to do exactly? [03:45] feel free to take a look epicdatanet.co.cc [03:45] there is a button leading to the forum at the moment, I'd like to nest it inside the home page [03:46] why not just tweak the style.css and use the built in banner options in SMF? [03:47] I think that would be much easier to accomplish, and cleaner. [03:47] I'm also planning on putting an autosizing to browser chatbox (I installed an ajaxchat mod in smf) on the left side set to follow the browser, hopefully collapseable [03:47] :) can I give you some adivce? [03:48] don't install a chat box [03:48] but I doubt that would nest it inside the content area [03:48] with a forum [03:48] your forum will die [03:48] becuase most of the talk will be in the chat and not in the forum [03:48] I was wondering about that [03:48] I'm back but need to deal with a real life issue... car with flat (really FLAT) tire... [03:48] I've done that and regretted it [03:48] ewww yuck [03:49] I'd like to give it a shot anyway, perhaps restrict it to forum members? [03:49] I'm still trying to connect via putty. I think I opened the port I needed by using this command, but I don't think it's working yet. [03:49] restrict it to advanced forum members above a certian post group if you want to do that [03:49] skuld: telnet 1.2.3.4 22 will test whether you can access the SSH service on IP 1.2.3.4 ... what happens when you try? [03:50] * jmarsden needs to go again to get a car tire (very FLAT) fixed... [03:50] * skuld iptables -A INPUT -p tcp -d 0/0 -s 0/0 --dport 553 -j ACCEPT [03:50] good thinking, hopefully the mod allows that, truth be told I haven't gone through the forum options yet, just added some themes and mods [03:51] skuld: 553? OK, if you want to use 553 for SSH and have set up sshd to use it... did you do that? [03:51] I'm very glad I came here by the way [03:51] you'll need to code that by hand. I can probably help with that after i get my mess cleaned up LOL [03:52] at least the first part, integrating the forum to the website, that's pretty easy...most of the time is just spent tweaking your smf/style.css .... and there's a little coding in the index.template.php file to remove the SMF logo on the right. [03:52] I was afraid of that [03:52] brb, have to dress my kid for bed.... [03:52] skuld: Whether that iptables command will work depends on the exact set of commands in the INPUT table... if there was a -j DENY at the end then appending to the table fails because the DENY rule gets reached before your new appended ACCEPT rule [03:52] OK... [03:54] actually, with the black silk theme I'm using for smf, other than some text that shouldn't be black all I really want to change is to remove the banner all together [03:55] and of course frame (without frames) it inside my home page [03:55] back [03:56] and no where else [03:56] skuld: So what does telnet 1.2.3.4 553 do # if 553 is your SSH port [03:56] right [03:57] sorry that I'm dividing your attention skuld [03:59] it's okay....I'm also multitasking here :) [03:59] I can use putty to telnet, right? [03:59] fyi, multitasking there makes things worse for you, not better lol [04:01] I know [04:01] yes you can [04:01] skuld: You could, but there is a Windows command called telnet which i was expected a Windows-using person to know and use :) [04:01] well, when I try that, it sits there for a few seconds...like 30 and then disappers [04:01] at least, there was in Win 2000 and XP, not sure about Vista and Windows 7. [04:01] I guess it's timing out [04:02] skuld: Open a console window and do it, so you can see the error message :) [04:02] Start -> Programs -> Accessories -> Command Prompt [04:02] Then type in telnet 1.2.3.4 553 and press Enter. [04:03] Do not "guess" that it is timing out, we need to *know* what it is doing :) [04:03] should there be a : between the IP and port? [04:04] No. a space. [04:04] telnet is not a recognized command in windows 7 according to my command prompt [04:04] you've gotta be kidding me.... I guess windows vista doesn't include telnet! [04:05] putty can do it according to google [04:05] yeah, it can , but it vanishes after about 30 seconds when I try it [04:05] OK, use putty I guess. BTW, you should consider using Ubuntu desktop for your admin workstation :) [04:06] skuld: There should be a Putty option to not close the window on errors... set it and retry. [04:06] http://www.webmasters-central.com/article-blog/tutorials/how-to-telnet-server-using-putty-ssh-client/ [04:06] I would if I could afford to do that (another machine) [04:06] skuld: Dual boot? Or run Ubuntu Desktop in a VM under Vista (better, run Vista in a VM under Ubuntu)... [04:06] okay, trying it [04:07] any recommendations on using ext3 vs ext4 for servers? Is ext4 stable enough to be used in a production environment? [04:07] "connection closed by remote host" [04:07] I've got xubuntu on my server running the desktop right now. I'm looking at three monitors, two keyboards, two mice, a bunch of cords, a cat on my lap, and a bag of reeses [04:07] nice....reeses.... :) [04:07] yes [04:07] skuld: Incidentally, it would have probably been faster to reinstall Ubuntu server than to do all this :) [04:08] anyway... so, is there really an sshd listening on port 553 on your server? [04:08] yeah, but I'd loose my internet connection in the process [04:08] On the server console try sudo netstat -ntlp |grep 553 [04:08] if it weren't for the reeses none of learning Dreamweaver, designing a site, choosing the best free host available, dumping said host, and installing my own linux server would have happened [04:08] OH wait!! do I need to "start" an ssh service? [04:09] skuld: If you dinked with its config file, you need to restart it, yes. [04:09] If you had left it alobne, it would have started on port 22 all by itself... [04:09] lol, it's always something simple like that [04:09] yup, okay...off hand you know what it is? (/etc/init.d/sshd start ?) [04:09] But someone told you to change the port... so we got into all this mess... [04:09] lol [04:10] lol [04:10] sudo service ssh restart [04:10] I would have had to open the port in the firewall anyway [04:10] * Pairadimesitty eats a reeses [04:11] okay trying again [04:11] Check it is there with sudo netstat -ntlp | grep 553 first. [04:11] ok [04:12] looks like it [04:12] Good. OK, now try ssh to port 553 from Putty... [04:13] bet it failed [04:13] connection timed out [04:13] just a hunch [04:13] too easy to work [04:14] LOL [04:14] skuld: OK, can you (from the server console) do ssh -p 553 username@1.2.3.4 # and log in as your username [04:14] how do I review that firewall iptables listing to see if there is a dENY in there [04:14] first let's check the sshd is working. [04:16] I wish I could show you a picture of the mess of hardware in front of me [04:16] if I didn't have this giant thick desk it probably would have caved in [04:17] yeah, it let me log in [04:17] skuld: OK, good, you can exit to get back out of that ssh session. [04:17] threw an error about authenticity of my ip I tried [04:17] logged out [04:17] yes [04:17] sudo /sbin/iptables -L -n -v [04:18] will display all the details of your firewall setup. [04:18] ok [04:18] One issue will be finding where firestarter put its config files and scripts... :) [04:18] cable modem, linksys dd-wrt router, mic, speakers, controllers (tv and game) [04:19] printer [04:19] god I need to clean up [04:19] wow that's confusing! [04:20] It's what firestarter did... now you know why using GUi tools that hide all the details is not always a good idea :) [04:20] I didn't see any denys in there, and it was a pretty short list...lots of * [04:20] yeah...I know... :( [04:20] but I'm learning all kinds of cool stuff now [04:20] On the server if you do sudo apt-get install pastebinit [04:21] we can then have you pastebin that iptables list somewhere I can read it... [04:21] okay [04:21] Once it is installed, do sudo /sbin/iptables -L -n -v | pastebinit # and tell me the URL it returns [04:24] http://pastebin/f7f31f192 [04:24] that's pretty handy [04:24] I think I mis copied somehting... [04:25] It would be if it worked... I'm seeing a page which says no such paste id... [04:25] hmm [04:25] Miscopying the ID would do that ... :) [04:25] no I copied it right [04:26] Did you copy the full URL OK... just pastebin between the // and the / [04:26] Looks odd to me... [04:27] let me try again.... [04:27] http://pastebin/f54c917e4 [04:28] Hmmm, mine look more like http://pastebin.com/f5b874c4e [04:28] Yup.... you omitted the .com [04:28] http://pastebin.com/f54c917e4 [04:28] duh! [04:28] LOL [04:28] I was just surprised that a link without a .com actually showed up correctly on the page [04:29] if I put those in my browser they work [04:29] Pairadimesitty: Firefox will turn WHATEVER it into www.WHATEVER.com if that URL exists... [04:29] but www.pastebin.com and pastebin.com are different sites... [04:30] I've never seen that before, though I suppose the http:// had something to do with it, if it were just text my settings would have turned it into a google search [04:31] my '553' port is 26 [04:32] I see no 553 anywhere in that listing... ? [04:32] 553 = 26 [04:33] I guess it's on line 40? [04:33] ah, that is way lower than all the DROP rules :) [04:33] ...is that a good thing? [04:34] cloudhosting [04:34] it means confusion in virtual form skuld [04:34] skuld: No. The DROPs will drop your packets before the ACCEPT in line 40 will accept them. [04:34] You want that line up around line 6 or so to work right. [04:35] okay, so how do I fix it to do that? [04:36] so is it even possible to embed my forum in my site so that all links to the forum will lead to my homepage with the correct link in the forum showing? [04:36] iptables -I INPUT 3 -p tcp -d 0/0 -s 0/0 --dport 26 -j ACCEPT [04:37] or something close to that :) [04:38] ok, let me try that.... @pairadimsitty: I don't understand [04:38] that makes two of us [04:39] I want the forum framed inside my homepage, nowhere else [04:39] when I try to associate a EBS volume with a UEC image using euca-attach-volume -i i-46C508A6 -d /dev/sdb vol-32BA04A7 the cc.log reports bad input params to vnetAttachTunnels() and doesn't attach the SC volume to the instance [04:40] I've tried using just sdb for the device name , that doesn't seem to make a difference [04:41] and if I should have a link to some part of the forum elsewhere on my site, it should show the homepage, with the appropriate forum post (or whatever) showing [04:41] everything else is working for me to my knowledge [04:41] okay, I can finally ssh into my server! :) [04:41] yay [04:41] New bug: #434836 in augeas (main) "[MIR] augeas" [High,Incomplete] https://launchpad.net/bugs/434836 [04:41] skuld: OK :) Now, do you already have some broken/half working email setup on it? or are we starting email (postfix/dovecot?) from scratch? [04:42] all that's done so far is the package has been installed, nothing further [04:42] OK. What is the overall plan -- you need outgoing SMTP for some PHP forum stuff, only, or you are really doing email hosting? [04:43] @pairadimesitty: if you're forum is embedded on your home page, then all links will also be framed there as well. [04:43] I have squirrel mail, just because that's what the guide I followed recommended [04:43] cool, I was worried about that [04:43] I have six domains that need some number of email addresses. a couple only need one or two, a couple need much more [04:44] i need to have the ability to send and recieve mail (relay if needed with smtp authentication) [04:44] I've tried almost everything....I think that's what blew up my last server...qmail finally did me in LOL [04:45] then I read a couple posts on postfix and dovecot...should be able to handle the multiple domains...and use mysql? [04:46] so that's what I installed was the default mail server (postfix/dovecot) [04:47] Yes, Postfix is a little saner... MySQL only if you need it... are you comfortable editing text files to configure all the email accounts? [04:47] now I just need to embed it, preferably without frames [04:47] I'm not convinced MySQL for email configs is necessary unless you are up into thousands of accounts... but that's personal preference... [04:47] actually if I had my preference, I love the idea of it being a database back end rather than text files, but if that's what I have to do, I'm comfortable enough [04:48] Either way is doable. Let's see if I can find a howto with an approach to using MySQL for you... [04:48] @pairadimesitty: I'm saying you don't need to embed anything. why not just make your smf forum your home page with just your banner, and add the menu extra menu links you need at the top? [04:49] ok [04:50] skuld: Are you running Ubuntu 9.04 Server, or 8.04.3 LTS ? [04:50] 9.04, the latest and greatest [04:50] I like my existing page, and the wrap around of the content, I also plan to embed squirrel mail, I will have pages though that have no framing, and I want to be able to handle all future expansion fluidly and in the same content region [04:51] OK... then http://craigballinger.com/blog/2009/07/postfix-dovecot-mailserver-on-ubuntu-904-jaunty-jackalope/ looks a good fit for what you are trying to do. [04:52] Go through it slowly, ask me questions if anything looks odd or "goes wrong"... [04:52] oddly what I did in that situation was create dynamic web pages using php. using a variable like $content = 'stuff here' I was able to create new pages of content with out mucking up the index.php. just created a new php file that returned $content. [04:53] okay. also be thinking of apache2 in the background. I guess things have changed since the httpd days...I thought I could just copy paste my conf file, but that didn't work lOL I'm reviewing the link you sent now [04:54] skuld: Sure, I can help with the apache stuff too later on. Debian/Ubuntu break out the configuration to make virtual hosting easier... which it is, once you understand how it all works :) I need to go pick up my daughter from somewhere or other... back in ~20 or 30 minutes... [04:54] crohakon: install vsftpd [04:55] bingo skuld, looks good, does it auto size for no frame scrollbars? [04:55] Anyone have experience configuring Tomcat to work with Apache2 ? [04:55] So, apparently I missed the FTP server during the install of ubuntu server edition. How can I install it now post install? [04:55] crohakon: apt-get install vsftpd [04:55] crohakon: apt-get install vsftpd [04:56] wait... [04:56] you guys make me come here... [04:56] it will auto size with the broswer. [04:56] and then you answer my questions that I asked in ubuntu? heheh [04:56] okay. looks like I've gotta fix apache before I continue with postfix... [04:56] ha ha [04:56] skuld: Really? Why? [04:57] that'd be for the install, more indepth requeusts for server version ... this is the best place to ask [04:58] smf by default autosizes with the browser...that is unless you have a banner image that's really large horizontally. [04:58] https://bugs.launchpad.net/eucalyptus/+bug/452185 [04:58] Launchpad bug 452185 in eucalyptus "can't access SC volumes in a UEC instance" [Low,Incomplete] [04:59] skuld: I'd copy your httpd.conf file somewhere safe, then sudo apt-get purge apache2 && sudo apt-get install apache2 # and you should have a very basic working apache2 ... anyway, got to go, back later... [05:01] I know that it does, and like that, it does autosize horizontally with iframe,but iframe sucks and won't autosize virtically [05:01] scroll bars inside the window and stuff like that [05:02] I'm not sure of anyway to do verticle autosizing without an iframe... [05:04] I saw a script online, php that is advertised to do something like that, but I'm not quite sure how to set it up, and I wasn't sure if php is the best solution [05:04] not because I'm against it or anything [05:04] I just don't know [05:04] like I said, this is my first experience with all this stuff [05:04] my first website, my first forum, my first server [05:05] my second host, but awardspace is about the best freehost and it didn't allow some needed php that the forum used [05:06] how do I exit and save when editing files with vi? [05:06] eventually after wading through the many free hosts out there researching I said F* it and pulled out my old optiplex gx 260 [05:06] crohakon: :wq! [05:06] that's why I stick with nano, it's annoying, but all the commands are listed [05:07] russlar; eh? I try to type that and it just... well.. types it.. [05:07] crohakon: hit escape first [05:07] !vi [05:07] Text Editors: gedit (GNOME), Kate (KDE), mousepad (Xfce4) - Terminal-based: nano, vi/vim, emacs - For HTML/CSS editors, see !html - For programming editors and IDE, see !code [05:07] oh! [05:07] Thanks =) [05:08] Does pico still exist? [05:08] yup [05:08] I really liked pico back in the day... back when I remembered how to do stuff. =( [05:10] perhaps ssi then skuld? [05:10] or do you have an idea for php, probably much better to use php? [05:13] How do I edit my iptables? [05:15] nevermind, I am dumb. [05:15] hehe [05:16] sorry, I was reading a tutural.... [05:17] @pairadimesitty: you are asking something about SSI versus PHP? [05:18] sort of, I really don't care what the method is so long as it's sort of clean and I can reimplement it without too much trouble [05:18] So, another dumb question... where would I find the httpd.conf file? [05:18] /var/? [05:19] it's in /etc/apache2/ [05:19] by default it's a zero-byte file [05:19] crohakon: /etc/apache2/httpd.conf [05:19] I was... close.. [05:19] hehe [05:19] I'm still working on configuring apache2 myself [05:19] Read all the files under /etc/apache2 to see how Debian/Ubuntu sets things up. [05:20] * jmarsden is back, by the way :) [05:20] wait... the file is empty? [05:20] *gasp* [05:20] I just recall that some googling told me something about setting the server to accept .htm and that the default was just .shtm [05:20] * skuld jumps up and down [05:20] I have to fill it... by my self? [05:20] No, you have to read the other files under that directory, as I just said... [05:20] I'd like to be able to effectively use css to control the "frame" [05:21] then that's all just HTML coding [05:21] * jmarsden wonders why jumping up and down is part of the process of configuring email servers, ... ? [05:21] healthy body makes a healthy thinking brain? [05:22] does cussing count? [05:22] while jumping up and down? [05:22] only as entertainment [05:22] * jmarsden thinks... but jumping up and down makes it really hard to read the screen, and even harder to type accurately... [05:22] while configuring email servers [05:22] LOL [05:24] I want a virtically resizing "frame" with no margin or border or whatever that fills my content box with proper resizing [05:24] I want to control it with css if I can, but I can adapt [05:25] and I want another reeses [05:25] I want to want fewer things and still be happy [05:25] * Pairadimesitty eats another reeses [05:25] Pairadimesitty: ask in #css, or #html, about those... not sure what channel would help with providing Reeses ... [05:25] can I copy/paste between my browser and putty? [05:26] skuld: Sure. You can set options in putty for what the different mouse buttons do, I usually set it to match xterm, but Windows people find that confusing... [05:27] ah, I missed the html channel at first when I looked [05:28] * jmarsden thinks maybe you should start a #reeses channel :) [05:28] skuld: How is the email config coming along? [05:30] just got finished setting up the postfixadmin superuser [05:30] OK. [05:34] anyone know of a good guide for configuring apache2? The apache website is giving me trouble. [05:35] I'd start with the Ubuntu server guide [05:36] is that at the ubuntu website? [05:36] jmarsden: damn, I think that might just work [05:36] pico = nano I use pico [05:37] ##php pops me into ##overflow is it actually full or does that mean something else? [05:37] crohakon: https://help.ubuntu.com/9.04/serverguide/C/ [05:39] Pairadimesitty: ##php lets me into it just fine. Are you a registered FreeNode user? [05:39] ah, no [05:39] I only irc if there is a problem lol [05:39] I don't know for sure, but that might be it. Register your nick and try again? [05:41] frack [05:41] doesn't accept hotmail accounts? [05:42] that's my main email! [05:42] Then you must not have much interest in receiving mail. [05:42] at least untill I set up my squirrel mail account [05:44] @jmarsden: This instruction isn't clear. [05:44] what the heck do I have to do here [05:44] 1.sudo nano master.cf [05:44] dovecot unix - n n - - pipe flags=DRhu user=vmail:mail argv=/usr/lib/dovecot/deliver -c /etc/dovecot/dovecot-postfix.conf -f ${sender} -d $(recipient) [05:44] do I just put that at the end of the file? [05:45] skuld: Edit that file using the nano editor as root (hence the sudo), and unless there is already a line starting with dovecot in there, yes, add it to the end of the file. [05:46] ok [05:47] That should all be one single line of text, in case that isn't clear either... [05:48] You might need to do sudo nano -w master.cf so it doesn't word wrap the text [05:49] got it :) [05:50] okay... how can I install gnome? I give up on console. =( [05:51] crohakon: personally when I got fed up installing a gui free ubuntu server I switched and installed xubuntu to start over [05:51] crohakon: Installing a GUI on top of Ubuntu server gets you a hybrid mess that is neither desktop nor server, and is best avoided. [05:51] so far it works [05:51] crohakon: But what is so hard about using the command line? [05:51] crohakon: sudo apt-get install ubuntu-desktop, if you insist. Any questions about it are off topic here, see #ubuntu. [05:52] So you suggest installing the desktop version and then just installing the servers? [05:52] crohakon: Yes, if you are 100% set on having a GUI on your machine. [05:52] crohakon: What I really suggest is sticking with the command line :) [05:52] okay, got a problem with the dovecot-postfix.conf. it failed when I restarted the service... [05:52] I guess I will just push on and learn something *sigh* [05:53] sudo /etc/init.d/dovecot restart [05:53] That's the spirit [05:53] I copied / pasted from the website. [05:54] skuld: OK, what does the error say, and pastebinit I am still suffering from the vista/windows never have to use a command line to do anything syndrome. [05:54] crohakon: Really? what is the Windows GUi equivalent of the *Windows* netstat command? or nbtstat? [05:55] simple, make some other program do it for you [05:55] that's what 3rd party is for [05:55] There are plenty of command line Windows commands that a good Windows server sysadmin knows and uses. [05:55] http://pastebin.com/f61603c18 [05:56] <--- not a windows server sysadmin [05:56] Pairadimesitty: Name a 3rd party tool that does what dcpromo does in Windows Server 2003 [05:56] holdon...it didn't say everything [05:56] <--- a Windows battle field 2 player [05:56] sudo /etc/init.d/dovecot restart [05:56] * Restarting IMAP/POP3 mail server dovecot Error: Error in configuration file /etc/dovecot/dovecot-postfix.conf line 5: Unknown section type [05:56] Fatal: Invalid configuration in /etc/dovecot/dovecot-postfix.conf [05:56] I wasn't planning on being so blunt jmsrsden, but yes, maybe he isn't [05:56] crohakon: Now you are becoming a server admin, right -- you are setting up a server. So, Windows or Linux, you will need to use the command line some of the time. [05:57] and how should I know, I'm not either [05:57] you know full well, this adventure is even my first time writing in html [05:57] lol. The sad thing is, I once had a fairly good understanding of the command line and how to do things. Like 8 years ago... [05:58] so... the claim that "vista/windows never have to use a command line to do anything" ... seems to have been somewhat invalidated :) [05:59] skuld: Can you please do pastebinit No no no... I was claiming that I when using vista/windows never had to use command line. I was not speaking for everyone else. =) [05:59] although I know ALOT about windows in all but the earliest of it's forms, I haven't messed with the server side much, and all my networking is done through hamachi and my router [05:59] I guess I could have worded it better though. [05:59] * skuld since I cant figure out how to pastebin a file... LOL [05:59] mail_location = maildir:/var/vmail/%d/%n [05:59] first_valid_uid = 150 [05:59] last_valid_uid = 150 [05:59] passdb sql { [05:59] args = /etc/dovecot/dovecot-sql.conf [05:59] } [05:59] userdb sql { [05:59] args = /etc/dovecot/dovecot-sql.conf [05:59] } [05:59] master { [05:59] path = /var/run/dovecot/auth-master [05:59] mode = 0660 [05:59] user = vmail [05:59] group = mail [06:00] } [06:00] skuld: pastebinit ok [06:00] You will get kicked off if you flood like that :) [06:00] wait... you can do that? [06:00] http://pastebin.com/f1a3134c3 [06:00] crohakon: Sure. pastebinit takes whatever you give it on standard input and pastebins it and then outputs the URL of the pastebin page. [06:01] that's better. [06:01] sorry about taht [06:01] So you can use input redirection with pastebinit just as you can with other commands. [06:01] that is so nice of pastebin.com [06:01] *nods* [06:01] cool, I wasnt' the only one who learned something this morning LOL [06:04] Morning? It is 10pm here :) [06:04] just past 1 am [06:05] hello people [06:05] so i just shut down my system to upgrade memory [06:05] when i turn off pc i cant see my site [06:05] wizzy.us [06:05] that was your first mistake LOL [06:05] @jmarsden, any ideas? [06:06] i can connect with webmin and putty [06:06] but when i type site it doesn show [06:06] sudo service apache2 start [06:06] ? [06:06] skuld: I don't see anything obviously wrong there... I think we might need pastebinit sounds like your webserver service isn't running Wizardlovak [06:08] wizardslovak: ssh in using Putty, then sudo netstat -ntlp | grep :80 and tell us what you see [06:08] i am restaring [06:08] and i got error [06:08] wizardslovak: You'll need to be much more specific about the error :) [06:08] ok ok [06:08] ERG! I get permission denied [06:08] "could not open config file : /etc/squirrelmail/apache.conf [06:09] try: sudo vi /etc/squirlmail/apache.conf [06:09] skuld: sudo cat /etc/dovecot/dovecot-sql.conf |pastebinit [06:10] Also check what perms the howto says that file should have... [06:10] "wait [06:10] there is error "syntax error on line 4 of /etc.apache2/apache2.conf [06:11] wizardslovak: So either the file /etc/squirrelmail/apache.conf does not exist, or it has incorrect permissions, I would think. And apparently you also have a typo or something in your /etc/apache2/apache2.conf file. [06:11] on line 4 it says "include /etc/squirrelmail.apache.conf [06:11] i will just pu # before it [06:12] wizardslovak: Do we have a . vs / problem. No, fix it, don't hack it. [06:12] i did remove squirrelmail loong time ago with "purge" [06:12] Ah, but you forgot to update other files that refer to it... OK, in that case, go ahead and comment it out. [06:12] Hi [06:13] ok now it works [06:13] thank you people very much [06:13] wizardslovak: No problem. === crohakon is now known as crohakon|afk [06:13] well, goodnight [06:13] crohakon|afk: Goodnight! [06:13] skuld: sudo cat /etc/dovecot/dovecot-sql.conf |pastebinit # did this work? [06:14] btw when i reset apache there was comment "httpd (no pid file) not running" but apahce started ok [06:14] I have a server running under Ubuntu 9.04 and am a beginner to DNS configuration. I've configured my DNS in such a way that it seems to work well on my server if "nameserver 127.0.0.1" is the first line of resolv.conf, but it doesn't work from anywhere else [06:15] additionally, when trying to "dig @myserverIP mydomain.tld" I get a timeout whereas [06:15] "dig @localhost mydomain.tld" works fine [06:15] Any idea of what could be causing this? [06:15] Phlogistique: Sounds like bind is configured to only use localhost [06:16] @jmarsden: finally got that pastebinit.... http://pastebin.com/f6e2207ab [06:17] jmarsden: uh, that's right [06:17] okay, I lied....I keep getting permissin denied [06:18] Phlogistique: So check /etc/bind/named.conf* for any options that may be restricting your named to only using localhost, and also check that /etc/hosts is correctly set up [06:18] skuld: OK, sudo -s and then pastebinit jmarsden: yeah, just did that, thank you [06:19] stupid permissions! try this Jmarsden: http://pastebin.com/f1d985c5a [06:20] Phlogistique: You can also try sudo netstat -ntlp | grep named to see what ports named is listening on. [06:23] jmarsden: it was listening on the loopback [06:23] jmarsden: fixed now [06:23] Phlogistique: Good :) [06:24] jmarsden: thank you so much orz [06:24] skuld: Shouldn't that file have more than just the user query in it? [06:24] and now I don't know how to create a user for squirrel mail [06:24] Phlogistique: You're welcome. [06:24] I figured it might by default use my ubuntu login at least by default [06:25] Pairadimesitty: It uses whatever your IMAP server uses for authentication. [06:25] I have squirrel mail properly installed (I think) I selected the D option to use default IMAP settings [06:26] there's a lot of commented stuff. [06:26] squirrelmail configtest passes [06:26] Pairadimesitty: How are you (in general) adding email users and storing information about them? [06:26] I went back through and uncommented/changed the queries to what the tutorial stated. [06:27] skuld: OK, and does that make it work now? [06:27] that's just it, besides installing and setting the automatic IMAP settings from the menu, I haven't added any users [06:27] no. same result [06:27] and don't know how [06:27] because I stooped [06:27] *becuz === WALoeIII_ is now known as WALoeIII [06:28] Pairadimesitty: It is not wise to configure a webmail system before you have basic email (SMTP and POP3/IMAP) working :) [06:28] I just followed the guide [06:28] I think it should work with just one or two settings changed [06:29] Pairadimesitty: Learn what you are doing as you go and test as you go, blindly following a howto leads you... somewhere, but you don't know where :) [06:29] skuld: OK, pastebinit the newly fixed up dovecot-sql.conf file for me. [06:30] and when I try to log in with my ubuntu user name it comes up with the error:connection dropped by IMAP server [06:30] http://pastebin.com/f65189cea [06:30] Pairadimesitty: OK, so check the dovecot logs for what it said was wrong... and recheck your dovecot configs. [06:31] a different username leads me to a normal incorrect user name error [06:33] it would be nice if I could find a guide that went from A-Z that actually explained what things were along the way in understandable english [06:33] Pairadimesitty: As I said: OK, so check the dovecot logs for what it said was wrong... and recheck your dovecot configs :) [06:34] lol [06:34] the guide doesn't use dovecot [06:34] Pairadimesitty: OK, whichever IMAP server you used, check its logs. [06:34] courier [06:35] If you don't want to do it yourself, pay for webhosting on someone else's server :) If you want to do it yourself, learn how. There is no shortcut. [06:36] There are people around who can help you, if you ask good clear questions and are willing to learn. But it's not trivial to set up a working and secure server. [06:36] As skuld is discovering :) [06:36] Of course I let him use SQL and guess where we now have issues... in the SQL stuff I don't use myself :) [06:36] LOL :) yeah [06:40] found the error log [06:41] says ....stuff..... imapd: Connection, ip=[::1]imapd: chdir Maildir: No such file or directory [06:41] insert a second line there [06:41] sorry [06:42] Pairadimesitty: OK, so your user lacks a ~/Maildir directory and you set things up to use Maildir format malboxes... [06:42] that good bad or ugly? [06:43] seems a simple fix though? [06:43] Try creating the ~/Maildir and see if it helps. Sounds like you just forgot to actually provide a place for that user's email. [06:43] Yes, very simple fix -- create the directory :) [06:43] where just under the webmail folder? [06:44] No..... this is IMAP we are talking about, not webmail. Log in as that user and do mkdir ~/Maildir [06:45] I'm already in terminal as my user set to sudo su, so you mean literally type "mkdir ~/Maildir"? [06:45] sorry for being a dumb@$$ [06:46] I blame my parents [06:46] Pairadimesitty: as yourself, not sudoed to anything, log in as that user. Then type in the command mkdir ~/Maildir [06:47] done [06:47] testing [06:47] skuld: OK... still looking for what might be wrong in that SQL config file... looks reasonably sane now. [06:48] ok [06:48] progress, thanks, now it seems it needs mailboxes to work [06:48] I recall seeing something to that effect in the squirrelmail config [06:51] Pairadimesitty: OK, so now you can just create new users (with the useradd command), set their password (with the passwd command) and create their Maildir's (using mkdir ~/Maildir or similar) [06:52] skuld: Just to simply the file a bit, can you try getting rid of all the comment lines so we just have http://pastebin.com/mc71e400 [06:52] by mailbox, I mean when I log in it error saying it is unable to open the inbox [06:52] *errors [06:53] Pairadimesitty: I'm not a courier expert at all... maybe it needs you to create new and cur subdirectories within the Maildir/ , or something? [06:53] that's my guess [06:54] seems strange that is hasn't created them by default on install [06:54] How would it know which system users you want to also be email users? [06:55] http://pastebin.com/f244b6dac [06:56] still get the same error [06:58] would case matter in the queries? I see a couple of AS and a a couple of as in the queries from the tutorial [06:58] well, good point, I was hoping it would set the admin that installed it as a user, but yeah, I see what you mean [06:59] skuld: generally speaking SQL keywords can be either case, but strings within quotes are likely to be case-sensitive [07:01] skuld: OK... I guess I'm going to create a virtual Ubuntu 9.04 server here, and follow that howto and see what happens... if it works for me we can then check what is different between your setup and mine. [07:04] okay. I'll be on again tomorrow. I need to get some sleep [07:04] thanks for your hard work, and I've learned a HUGE amount today [07:04] well,.... yesterday anyway [07:06] OK... no guarantees when I'll be here, but I often am here :) [07:06] okay. thanks. mail's been broken for a couple of weeks, what's another day or so? LOL [07:06] :) OK, goodnight. === Guest23301 is now known as goodboy [07:18] night jmarsden [07:18] Goodnight Pairadimesitty [07:18] found a guide, looks like what I need [07:18] Good :) [07:18] will report back to whoever's here when I've gone through it [07:20] Hi there, I gave permission some folders in my server to a "group" but when the users connect to the file (using samba) the permision change to "Domain users" groups and the rest of the user form my "group" can not acces anymore. How I can keep the permision in a folder? === NCommander is now known as NC|Mobile === NC|Mobile is now known as NCommander [07:31] roxy09: SAMBA permissions and Unix file permissions are not the same thing. You possibly want to edit your smb.conf file so that users in your group have appropriate access to that SAMBA share? [07:36] maybe...I'd give permission to a "group" but when the user modify or edit files the permission in group change to "User Domain" and the users form teh original Group lost the permisions [07:47] yay, it shows my inbox now, I needed mkdir -p /home/usename/Maildir/{cur,new,tmp} [07:47] that's all [07:47] now to mail myself and pray [07:53] that smells like your MTA isn't working as it should [07:54] hmm, interesting [07:54] very interesting [07:55] I got all the right folders set to my user, but when I sent myself mail, nothing showed up, however I DID find the email in /var/mail/ as a file (not folder) with my username [07:56] the only other filer in that folder is one called nobody [07:56] New bug: #455625 in eucalyptus (main) "Eucalyptus Loses Public IP Address" [High,Incomplete] https://launchpad.net/bugs/455625 [07:56] the file type is "mailbox [07:56] : [07:56] sorry, typo [07:56] "mailbox" [07:58] Pairadimesitty: Sounds like half your mailsystem wants to use mailbox format in /var/mail and the other half wants to use Maildir ... fixing things to all the pieces use the same mail spool format would be good :) [07:59] s/to all/so all/ [08:00] which format would you recommend? [08:00] I bow before the master [08:00] It all depends; for high volume mail servers, Maildir, but for just a few users (say 100 or less) the old fashioned defaults are fine. [08:00] If I had a webcam you would see me bowing right now [08:01] :) [08:01] let me put that another way, how would one add a new user? [08:02] Pairadimesitty: Once you have the mail services all set up consistently, that shouldn't matter; you just create the user with useradd and set their password with passwd and the mailsystem creates their mail files as needed. At least, that's how it works for Postfix/Dovecot [08:03] yes, that makes sense [08:04] I think I get how the Maildir sorts out files now, does the other method just create one file per user or one file per email? [08:04] Pairadimesitty: One per user. [08:04] does it allow for custom folders? [08:05] never mind [08:05] I would never use old style mbox files if I could avoid it. [08:05] seems like squirrelmail has the function, should be able to handle it [08:05] It's much less a matter of how many users there are, than it's about how much e-mail each of them will store on the server, and how much you care about the integrity of said e-mails. [08:06] What you see in /var/mail is old style mbox files. It's basically all of the user's inbox stored in one file. [08:06] I'll follow soren's advice, seems like all emails in one file is kinda querky [08:07] Any changes to any e-mail in that inbox requires rewriting the entire file. [08:07] now to figure out how to set courier and imap to the other format [08:08] and no I obviously don't have any emails that need saving or converting over [08:08] That /is/ possible, though. [08:08] I don't know anything about courier. Never used it. [08:08] I saw a guide while looking on google on switching emails over [08:09] let me regoogle to see if it knows what's up [08:23] I'm having trouble finding it, could postfix need any settings changes? [08:27] mdz: ping [08:28] seems like courier is supposed to use maildir by default [08:29] something else must be interfering, or it's not courier [08:29] hey there, sorry to bother, but, I have a problem with Ubuntu Server 9.04... when I run netstat -lu it gives: netstat: no support for `AF INET (udp)' on this system. How do I add UDP support? [08:30] Yuttadhammo: Which kernel are you using? [08:30] just a sec... [08:31] 2.6.28.11 [08:31] What's the exact output of "uname -a"? [08:31] Linux grid1.buddhaverse.net 2.6.29.5-grsec-hostnoc-4.2.0-i386-libata #1 SMP Thu Jul 9 01:00:08 EDT 2009 i686 GNU/Linux [08:31] There's your problem, probably. [08:31] that's not an ubuntu kernel. [08:32] uhoh... [08:32] ...and it's not 2.6.28.11, clearly. [08:32] sorry, aptitude shows it as 11 [08:32] Well, it's not what you're running. [08:33] hmm... why does aptitude say I am? [08:33] aptitude does not know which kernel you're running. It knows which kernels were installed. [08:33] ..from .deb-packages. [08:33] ah [08:34] 2.6.29.5-grsec-hostnoc-4.2.0-i386-libata probably came from somewhere else. Is this a rented server, by any chance? [08:34] yeah [08:34] Physical or virtual? [08:34] physical [08:34] Ok, then you can probably just boot into another kernel somehow. [08:34] See if it uses lilo or grub. [08:35] brb, have to reboot (kernel update). [08:35] okay thanks a lot. [08:36] imapd file says maildirpath=Maildir, that right or does it need home/$user/Maildir or something like that? [08:37] Pairadimesitty: maildir paths are usually relative to the user [08:37] 's home directory. [08:37] In other words, what you have is probably fine. [08:38] then I'm still sure which program has it wrong and what file to edit [08:38] Yuttadhammo: Who are you renting this server from? [08:38] burst.net [08:38] I see they are on aim... I'm going to contact them... [08:38] Disabling UDP seems /quite/ odd. I wouldn [08:38] t even know how. [08:38] yeah, really weird. [08:39] I don [08:39] t see a kernel config option for it. [08:39] lol... I was almost embarassed to ask. [08:39] presumably, grsecurity is being more restrictive about it. [08:40] sbeattie: Oh, right, it may actually support it, but grsecurity may be blocking netstat from showing it? [08:41] Yuttadhammo: Does it work through sudo? [08:41] that's my suspicion. [08:41] oh, I'm really embarassed now. [08:41] yeah, sudo works. [08:42] thanks for the help. [08:42] No problem. [08:42] Hey, it works everywhere else. Don [08:42] t be embarrassed. [08:43] guess I need to find a way to get back to an Ubuntu kernel... [08:43] Shouldn't be too hard, hopefully. [08:43] if I reboot the server, will I be able to ssh into it again? [08:44] Yuttadhammo: Depends on what you changed since last reboot :) [08:45] I think I'll wait until the guy who knows what he's doing gets back... I'm just trying to make a UDP connection... [08:46] * soren reboots again [08:47] * soren managed to uninstall grub2, so grub.cfg did not get updated with new kernels, but grub still chainloaded into grub2, so kept booting and older kernel (which I thankfully hadn't deleted). [08:48] I'm so frustrated, what program actually gets the mail and puts it in a folder, courier is set correctly, squirrelmail is set correctly, but the mail shows up in var/mail [08:48] postfix does that. [08:48] Well.. [08:48] It depends. [08:48] If you [08:48] I have postfix [08:49] re moving stuff around between folders in squirrelmail, it's your imap server. [08:49] and saslauthd, maildrop, getmail4 [08:49] If it's new mail coming in over smtp, it's your SMTP server, which I gather is postfix. [08:49] and a couple other things I don't know [08:49] Err.. [08:49] Why do you have all of those things? [08:50] The "other things I don't know" bit worries me. Is this not a system you set up yourself? [08:50] it is the best guide I could find for what I wanted, but it doesn't explain things worth a damn [08:50] or finish for that matter [08:50] I doubt you need any of maildrop or getmail4. [08:50] what do they do? [08:50] saslauthd, perhaps. [08:51] I suspect saslauthd is set up correctly [08:51] but again, nothing is explained [08:51] getmail4 - mail retriever with support for POP3, IMAP4 and SDPS [08:51] maildrop - mail delivery agent with filtering abilities [08:52] How do you receive e-mail from the outside world+ [08:52] ? [08:52] the guide was rigged for more of a server host -sell webspace sort of thing [08:52] imap [08:52] does that answer your question even? [08:52] Ah. [08:52] It might. [08:52] So people send you an e-mail, and you're expected to go and grab it from someone else's mailserver? [08:53] Over IMAP. [08:54] hmm? no, everything is run directly off this machine next to me [08:54] That's a bit vague.. [08:54] it's it's own mail server, dns, ssh, ftp, and web server [08:54] Say I wanted to e-mail you... How would an e-mail get from me to you? Are you set up to be your own MX? [08:55] the guide is http://howtoforge.org/perfect-server-ubuntu-9.04-ispconfig-3-p4 [08:56] I'm not asking what you've set up. [08:56] I'm asking how you expect mail from the outside world to reach your server. [08:56] I installed on Xubuntu after getting annoyed with not having a gui to work with, and I skipped ISPConfig and Jailkit [08:57] I'd like it to simply come directly to my server from the sender and store it in the per user Maildir's [08:58] with a check to make sure the email sends so it doesn't get lost in the ether, but I really don't know how the protocols work [08:58] as I said, I'm a dumb noob and this is my first time setting up any of this, and most guides are rather difficult to follow for this sort of thing [09:00] If you intend for mail to be delivered directly to your server via SMTP, you certainly don't need getmail4. [09:01] would it be helpfull to grab from my hotmail account or something like that? [09:01] Dear god, no. [09:02] and I found a forum post mentioning postfix, trying it out [09:02] Wait, what? [09:02] You said you were already using postfix? [09:02] yes [09:02] I said that earlier [09:02] But now you're trying it out? [09:03] As opposed to before where you were... what? [09:03] I'm trying out the post to see if it has accurate instructions on how to make postfix use maildir [09:04] Just put "home_mailbox = Maildir/" in your main.cf [09:12] I found that too! yay [09:13] though I found a post right under one that said that [09:13] Setting home_mailbox won't work if you're using procmail. [09:13] Instead of setting those vars on the procmail command line (they're wrong anyway), it's better to create a /etc/procmailrc and set them there:DROPPRIVS=YES [09:13] ORGMAIL=/home/${USER}/Maildir/ [09:13] DEFAULT=${ORGMAIL} [09:13] does that make sense to you? [09:15] Are you going to use procmail? [09:15] I'm not sure if I am using it, hang on a sec and I'll check [09:15] what does procmail filter anyway? [09:16] If in doubt, it's /extremely/ unlikely that you're using it. [09:16] apparently it's installed [09:17] as is spamassassin [09:17] Pairadimesitty: All these guides and forum posts and whatnot you're following... Are they writting with Ubuntu in mind? [09:17] yes, that's one think I make sure of [09:19] I think you can just ignore what it says about procmail. [09:20] :) rebooted into Ubuntu... man, that feels good. Ubuntu rocks. [09:20] (sorry for the spam) [09:24] I'm trying it now, sending an email, then if that fails I'm going with the other and trying it and forgetting the origional [09:28] progress, I got an email back from myself that said the email failed to be reseaved, I'm happy to know that it sends failed email messages [09:28] and this error about failure to write to directory makes me think I need to chmod [09:29] what number should I chmod it to? [09:29] Yuttadhammo: this is the ubuntu irc, how can that be spam? [09:30] hey all.. so i once asked about keeping people from seeing the folders in website and i was told about -option the folders.. or something.. can somone point me in the right direction pls.. tnx [09:31] (It's not the #fanboy channel...) [09:33] got it 700 [09:42] how about #reese [09:47] what is best channel to ask about htaccess [10:34] ttx: as far as I am aware, eucalyptus is ready to release [10:34] There are two bugs bothering me [10:35] mdz: bug 455293 is cosmetic and quite harmless to fix [10:35] Launchpad bug 455293 in eucalyptus "UEC management interface still has Eucalyptus as title" [Wishlist,Triaged] https://launchpad.net/bugs/455293 [10:35] bug 455816 is slightly more disturbing, should we release-note it ? [10:35] Launchpad bug 455816 in eucalyptus "When installing a UEC cluster, the prompt for the private interface is displayed after the "Installation complete" dialog" [Medium,Confirmed] https://launchpad.net/bugs/455816 [10:37] The last one still nominated for karmic is bug 453456 [10:37] Launchpad bug 453456 in eucalyptus "excessive logs in /var/log/eucalyptus" [Low,Triaged] https://launchpad.net/bugs/453456 [10:37] mdz: which, if any, should be fixed before RC, or before GA ? [10:38] I'm fine with keeping the current release, just making sure we are aligned on this. [10:39] (in which case I'd drop the karmic nominations for those) [10:42] ttx: reading [10:45] ttx: 455293> we should go ahead and commit that to bazaar, since if we do an upload there is no reason not to include it [10:46] mdz: OK. [10:47] ttx: 455816> I have never seen that question; is it new? [10:47] oh, it's only when you have multiple NICs [10:47] mdz: yes, submitted by Etienne on that specific multi-NIC testing [10:48] ttx: I've looked at cjwatson's diff and I'm not sure I can evaluate the risk; maybe ask cjwatson what he thinks we should do wrt RC? [10:48] mdz: it's tricky to fix and test in the remaining timeframe [10:48] mdz: sure [10:49] I think it's an ugly wart, but not a fatal one [10:49] mdz: I can't evaluate the risk either, and can't really test the fix with my current hw [10:49] when I thought about it last night, I couldn't think of anything that would actually go wrong due to it [10:50] ttx: 453456> if (and only if) the logs are getting rotated, we can defer this to lucid [10:50] the fix is verbose but in principle straightforward - but it's a pain to test without uploading [10:50] I gave mathiaz directions on testing it last night [10:50] mdz: 453456> agreed, especially since trying to tone down those logs by configuration doesn't seem to work. [10:50] my inclination, though, is to leave it be [10:54] cjwatson: thanks for your input, I tend to agree with you. [10:54] hello, would anyone know how to list what nodes are registered on a cluster? [10:54] nijaba: apparently there is no euca_conf --list-nodes [10:54] ttx: not in the version I am using then [10:55] nijaba: I said "there is *no* euca_conf --list-nodes" [10:55] ttx: hehe sorry... [10:55] ttx: can we confirm that? [10:55] nijaba: grep NODES /etc/eucalyptus/eucalyptus.conf ? [10:56] mdz: it's confirmed, there is no such option. [10:56] ttx: I meant, can we confirm that the logs are getting rotated [10:57] mdz: I guess we need to let a system run without touching it for a few days... that would help in confirming bug 455625 as well [10:57] Launchpad bug 455625 in eucalyptus "Eucalyptus Loses Public IP Address" [High,Incomplete] https://launchpad.net/bugs/455625 [10:58] mdz: however I'm using my hw to test the RC ISo candidates [10:58] mdz: I'm looking in the code to confirm, and we can ask the Eucalyptoids for more detail / confirmation [11:03] ttx: how about we ask for a pointer to the code which implements this as well? [11:03] so we can confirm how it is supposed to work [11:04] mdz: will do when they are awake -- at first glance it's not obvious where it would be implemented [11:05] ttx: I've asked in the bug [11:06] ttx: meanwhile, we should prepare a branch which implements logrotate [11:06] ttx: I don't think we need to do anything about this for GA, but we should be prepared to do an SRU if it turns out that the upstream rotation doesn't work [11:07] (or is not complete enough) [11:07] mdz: agreed. Can't find where it would be implemented in their code, but they might just inherit it from some other Java library [11:09] mdz: should I wontfix the karmic nomination for the multiple-NIC installer issue ? [11:13] OH MY GOD [11:14] I am soooooo annoid that was it [11:14] annoyed [11:14] PO'd [11:15] soren: it was exactly what we both thought it was BUT it needed a / at the end and the guide didn't say that [11:15] home_mailbox = Maildir/ [11:16] not home_mailbox = Maildir [11:16] holy hell [11:16] that was dumb [11:21] Pairadimesitty: Uh, yeah. That's what I said. [11:22] The slash at the end denotes Maildir format. [11:22] omg, you did [11:22] I saw a post at the same time you posted that, I don't think it did have the slash [11:22] I feel like the idiot I am [11:22] give me a virtual slap [11:23] I deserve it [11:23] * soren pats Pairadimesitty on the head [11:23] I'm not a dog [11:23] I'm a pirate [11:24] if you want to pitty me give me booty [11:25] since I took procmail off should I bother putting it back on? I liked what it did but.... [11:25] Does everything work now? [11:26] I can get mail [11:26] I suppose I should try sending it before I celebrate [11:26] ...and before you start changing anything again (like adding procmail back into the mix) [11:27] YAY [11:28] it works [11:28] nicely [11:28] and quickly [11:28] and screw procmail until such time as I decide to start over from scratch [11:29] and I also believe saslauthd is set up properly, but I still don't know what all this stuff does [11:32] yeah, I don't think spamassassin is set up properly [11:32] Pairadimesitty: saslauthd is usually used to authenticate users so that they can send e-mail through your server. [11:33] I get it [11:38] ug, something else I'll skip for now [12:01] ttx: re: multiple NIC issue, I defer to Colin. if we can fix it with an acceptable amount of risk for 9.10, then I would like to see it fixed. if the risk is too great, then we can accept that. [12:37] New bug: #456238 in postfix (main) "empty message-id, no date in Evolution MUA" [Undecided,New] https://launchpad.net/bugs/456238 [12:41] ach, ##php is very helpfull for my issue, they had me turn my home page into an include, which does autosize my forum, but now css is messed up between both and links all lead to the forum page and I loose my homepage surround [12:48] so it's probably posible with php, but it's way to much work for now [12:48] so how do I make an iframe autoresize virtically === aubre_afk is now known as aubre [13:13] Good morning - Mark Spencer told me he'd find out what the status was on Asterisk and get back to me. I guess since he's a CEO he's not as involved with actual development as he once was when he created it. [13:14] aubre: the "status of Asterisk"? [13:17] What shall I call my backup user since "backup" is already taken? :/ [13:17] I lack imagination [13:17] backup2 [13:17] lol [13:18] Bilge: hint, base the name on the backup tool you'll be using [13:18] I wrote the tool myself [13:18] A series of three scripts for batch, iterative and rolling backups [13:19] I would call it Oxygen. [13:19] ..since that's what my naming scheme dictates. [13:19] YMMV. [13:19] What [13:19] If I make a user who is just used to download backup files from a restricted shell, should I add it as a system user (UID < 1000) or a normal user? [13:28] soren: i sill like backup2 [13:30] zul: Yeah, that's pretty good :) [13:31] soren: but all my machines at home are named after south park characters (including my wife's laptop) [13:31] zul: I have a set of virtual machines names roughly by their function. My web server is webster, my DB server is debbie, my DNS server is dennis, etc. [13:32] I have another set of servers (that are not single-purpose) named after the elements. [13:32] soren: heh originally it was simpsons characters but the new simpons episodes are kind of boring [13:33] I stopped doing [13:33] I stopped naming servers after Simpson characters when I got tired of typing comicbookguy. [13:33] mdz: I'd like to make the multiple NIC issue conditional on mathiaz being able to test my branch and confirm it working, then; I don't think we have enough iterations left to be able to justify an untested upload [13:36] New bug: #427094 in dbconfig-common (universe) "Wrong permission on file makes phpmyadmin show blank page" [Low,Incomplete] https://launchpad.net/bugs/427094 [13:37] Afternoon all. Anybody in Eucalyptus mode here this afternoon. I have an issue with keypairs that is either me being an idiot or a bug. [13:37] Can't get onto any instance I spawn. [13:41] New bug: #453739 in kerberos-configs (universe) "FFe: Sync kerberos-configs 2.2 (main) from Debian unstable (main)" [Low,Incomplete] https://launchpad.net/bugs/453739 [13:44] pmatulis: yeah people were asking when asterisk 1.6.2 was going to be stable [13:46] hmm I am getting this now in MANAGED-NOVLAN mode, priv interface 'eth1' must be a bridge, tunneling disabled in my cc.log - not my nc.log [13:47] this is after updating to 1.6~bzr931-0ubuntu7 [13:49] aubre: I'm getting that a well for the priv interface on the cc controller. [13:50] neil-wilson: whew - well finally I'm not alone [13:57] neil-wilson: I'm trying to file a bug but without much success. [13:58] aubre: are you getting vlan errors as well? [13:59] failed to attach tunnels for vlan 10 during maintainNetworkState() [13:59] neil-wilson: I get that when I tried to attach a SC volume to an instance [14:00] neil-wilson: when I could get instances to launch, that is [14:00] I can launch instances. I just can't get onto them. For some reason soren turned off key injection and I'm trying to understand why. [14:02] Wait, what? [14:02] I did what? [14:02] soren: turned off key injection in managed mode? 7th september. [14:03] neil-wilson: Oh, in eucalyptus? Yeah. [14:03] soren: how's it supposed to work then? [14:03] neil-wilson: Like on EC2. I need to run for 10 minutes. I can explain afterwards. [14:04] neil-wilson: hmm well all of a sudden, euca-describe-availability-zones verbose is working after I tried and succeeded in launching an instance [14:05] neil-wilson: now to try attaching a volume [14:08] neil-wilson: still can't attach an SC volume, now to check logfiles [14:20] neil-wilson: Just like on EC2. EC2 does't inject keys either. It offers keys through the meta-data service. At boot time, ec2-init goes and grabs the keys from the meta-data service and puts them somewhere appropriate. [14:21] well when I try to attach an SC volume, I am getting this in my node's nc.log : http://paste.ubuntu.com/297485/ [14:21] I don't even know if this is a runtime or compile time option. I just took a patch from upstream. [14:21] aubre: Is everything completely up-to-date on that system? [14:22] soren: I don't appear to be picking up the metadata - almost as though apache isn't providing it. Any idea whereabouts in the system it lives? [14:22] aubre: yes , I am using 1.6~bzr931-0ubuntu7 [14:23] soren: yes , I am using 1.6~bzr931-0ubuntu7 [14:23] soren: I updated this morning [14:23] aubre: "everything". [14:23] aubre: Not "eucalyptus". [14:24] soren : I did a apt-get update followed by apt-get dist-upgrade [14:24] neil-wilson: It's reachable from the instances on 169.254.169.254. [14:24] soren: let me do it again [14:24] <_ruben> hmm, is / the only invalid character in a filename or are there more? [14:24] aubre: Ok. And everything is from Ubuntu? libvirt, kvm, kernel, etc.? [14:25] soren: yep, that's the way I roll [14:25] _ruben: ASCII 0 is unlikely to be a good idea. [14:25] _ruben: But it depends on the filesystem. [14:25] Not on my system. A get to that address gets the default Apache 'it works' message. What service provides it? [14:25] Can't find anything recent in the logs. [14:26] <_ruben> soren: but in general (or ext3 in particular), it's pretty safe to assume that exluding \000 and / is safe? [14:26] <_ruben> the at first though obvious ones can be escaped properly if needed (like * and :) [14:28] neil-wilson: Eucalyptus provides it. [14:28] neil-wilson: The cloud controller, I would expect. [14:29] _ruben: Define "safe". [14:30] soren: Off to go digging. Thanks for your help. [14:30] <_ruben> soren: safe is wrong term indeed, allowed is a better world, from the FS point of view .. so "foo..bar" might not be considered safe, but is valid [14:34] _ruben: I /think/ if you steer clear of / and ASCII 0, you'll be alright. * and ? are both certainly aceptable filename characters. [14:36] soren: ping [14:36] mathiaz: about bug 455816, do you have the necessary hardware to reproduce the issue and verify the fix ? [14:36] Launchpad bug 455816 in eucalyptus "When installing a UEC cluster, the prompt for the private interface is displayed after the "Installation complete" dialog" [Medium,Confirmed] https://launchpad.net/bugs/455816 [14:36] soren: is there an up-to-date uec appliance that I can try in my cloud? [14:37] kirkland: Nothing more up-to-date than what I posted last Friday. [14:37] kirkland: tested UEC setups from amd64 and i386 20091020 -- I hit bug 444352 and bug 452556 in both cases, otherwise it's ok [14:37] soren: pointed to on ubuntu-devel@ ? [14:37] Launchpad bug 444352 in eucalyptus "DB deadlock on reboot prevents UEC from working, temporarily" [Low,Won't fix] https://launchpad.net/bugs/444352 [14:37] Launchpad bug 452556 in eucalyptus "euca-authorize default failing" [Medium,Confirmed] https://launchpad.net/bugs/452556 [14:37] ttx: okay good; java problems, i didn't get a built iso before i went to bed [14:38] kirkland: doko fixed that [14:38] Good morning all! [14:38] ttx: on the authorize failing, does it have any mal effect? [14:38] ttx: i found that i could ssh just fine [14:38] kirkland: to the public address ? [14:38] ttx: yes [14:38] ttx: have you tried? [14:39] kirkland: Yes. [14:39] smoser: do you know if I can just use migrate-bundle to put an aki/ari in euwest region? [14:39] soren: thanks [14:39] I'm having a problem with apache2. I configured my websites, tested it works fine on my internal ip (192.x.x.x) but when I try to use either the domain or my public IP (216.x.x.x) I get [14:39] "unable to connect" [14:39] kirkland: interesting... no, I didn't try that. I ran the command again, and since run-instances was run in between, it would succeed [14:39] jjohansen, i know (thanks to our friend ross) that you cannot [14:39] :) [14:39] soren: maybe this is related to https://launchpad.net/bugs/452556 , but I found that I couldn't do a euca-describe-available-zones verbose until after I had launched an instance [14:39] Launchpad bug 452556 in eucalyptus "euca-authorize default failing" [Medium,Confirmed] [14:40] smoser: thanks [14:40] jjohansen, but in the future, i will get you some setup information for publish-image so you can use it. [14:40] aubre: I'm not actually working on Eucalyptus these days, so I'm the wrong person to talk to :) [14:40] then one command gets it in both places [14:40] smoser: that sounds nice [14:40] you're up early [14:40] aubre: that would be bug 444352, grep DEADLOCK /var/log/eucalyptus/cloud-output.log to confirm [14:40] Launchpad bug 444352 in eucalyptus "DB deadlock on reboot prevents UEC from working, temporarily" [Low,Won't fix] https://launchpad.net/bugs/444352 [14:40] ttx: try it next time; i'm not sure the command is entirely necessary (which might be another bug) [14:41] i was just about to ping you to ask you to push those to eu [14:42] can anybody get to the site http://www.localexpresscl.com ? [14:42] soren, what would you think about trying to fix bug 407949 for karmic? [14:42] Launchpad bug 407949 in ec2-init "ec2-init: ec2-set-defaults needs better defaults for non US/EU regions" [Medium,Confirmed] https://launchpad.net/bugs/407949 [14:42] smoser: there is no UEC/EC2 image RC candidate yet on the test tracker -- who/when ? [14:42] ttx, oh, sorry, i can put it there, i didn't know that it should be. [14:43] soren, see my comment 3 in that bug, UEC will always take the fail path [14:43] smoser: at one point the candidate for the image and the EC2 AMIs must be officialized on the ISO test tracker [14:43] ttx, should I put one up there? [14:43] smoser: tat would allow us to report test results, yes [14:43] smoser: if you still expect to change it, it might not be necessary though [14:43] ttx, well let me look real quick, fwiw, the 20091020 is up available on ec2 [14:44] ttx: you are correct sir, it was confirmed [14:44] i have no plans on changing it [14:44] soren: sorry then lol [14:44] smoser: ok, who do I/you need to ping to make that happen ? [14:45] ttx, i'll tak 10 minutes here, and just look, but i think that we just call the AMIs that are in http://uec-images.ubuntu.com/karmic/20091020/published-ec2-nightly.txt as RC. [14:45] and then i guess we tell slangesack that [14:45] smoser: ok, take your time [14:45] smoser: Sounds like a good idea. What do you suggest? [14:46] soren, just an absolutely simple catch-set-default [14:46] as en_US.utf-8 [14:46] ttx: perhaps a simple command could be put in the documentation for users which would clear the deadlock? [14:47] smoser: Feel free. [14:47] (make that UTF-8, though, not utf-8) [14:47] aubre: attempting to run the command multiple times in a row usually autofixes it [14:47] :) [14:47] aubre: but not everyone hits this bug, may I ask what your setup looks like ? [14:47] ttx: sure [14:47] i'll work a suggested patch, but need sponsorship. i'll attach bzr karmic branch there. [14:47] * skuld takes a number and waits in line [14:47] ttx; 1 CC, 2 nodes. [14:48] ttx: 1 CC, 2 nodes [14:48] aubre: on CC: fast/slow disk ? [14:48] ttx: CC is on raid across 4 disks [14:48] smoser: Cool, thanks. [14:48] ttx: 7200 rpm [14:49] ttx CC has 12 GB of RAM [14:49] can anybody help me with apache? [14:49] soren, so i'm going to hit 'target to release karmic' there, that will set of some flags ? [14:49] ttx: CC has 12 GB of RAM [14:49] aubre: hm, sounds quite different from what I run (2Gb RAM / 1x5400rpm disk) [14:50] kirkland: planning to do a RC test install today ? [14:50] ttx: this was bought/spec'd to be a proof-of-concept, if we get it working and we feel confident we may purchase a large system w/support [14:50] aubre: also restarting eucalyptus after the reboot clears the deadlock condition. [14:51] ttx: what's the best way to restart eucalyptus [14:51] aubre: I do "sudo stop eucalyptus" then "sudo start eucalyptus" [14:52] ttx: nice, so I don't have to shut down each process? sweet [14:53] ttx: oh and I'm using RAID 0 [14:53] ttx: Thanks for checking out bug 341256. One quick question, how do I change the version number to avoid the conflict? Do I just edit the version displayed when using dch -i or should I be changing somewhere else? [14:53] Launchpad bug 341256 in unattended-upgrades "RFE: unattended-upgrades should identify host." [Medium,Fix released] https://launchpad.net/bugs/341256 [14:53] Omahn: just edit the one dch -i creates [14:54] ttx: Is EBS in reasonable shape in your UEC installations? [14:54] ttx: Cool, thanks, will correct and upload fresh debdiffs. [14:54] Omahn: https://wiki.ubuntu.com/SecurityTeam/UpdatePreparation#Packaging has good rules on picking up a package version number for SRU [14:54] ttx: I'll read up. [14:55] Omahn: like I said, I'm not sure that's appropriate for SRU -- adding a useful feature that changes the behavior of the release is usually not appropriate [14:55] hi, we upgraded some servers to 9.04 and now we are confronted to a bug in smbfs. We can create files from 9.04 clients to a 7.04 samba server, but modification is impossible (e.g. touch failed with invalid argument and the file is created) [14:56] ttx: It's not really a feature for us, it's a necessity. Due to the volume of unattended-upgrade emails we receives its simply unmanageable without the hostname. [14:56] ttx: Although we are happy to just package the updated version ourselves. Just thought I would try and push it back upstream to benefit others. [14:56] Omahn: I understand why you need it, but I also see why others could get broken by the update (imagine some tool relying on the email format) [14:57] soren: I tested attaching a volume and saw it appear in the UEC instance -- does that count ? [14:57] ttx: Agreed. I'm happy with whatever line you and the SRU team recommend. [14:57] Omahn: I defer to them. [14:57] ttx: No problem. [14:57] soren: for more complex tricks, we relied on the eucalyptoids. [14:58] soren: maybe Dustin did something more clever to tset his attach fixes. [14:58] help? [15:00] kirkland: please add the doc item to the meeting agenda, yes [15:00] ttx: will do [15:00] ttx: test "what" attached fixes? [15:00] kirkland: the scsi-attach fixes [15:01] ttx: ah, in the interest of time, I tested using just kvm, libvirt, and virt-manager; nurmi tested the eucalyptus integration [15:01] kirkland: soren was asking about the status of EBS support [15:01] kirkland: ok [15:01] ttx: i was specifically working on the device add/remove from the kvm and libvirt side [15:02] ttx: once i had working reliably, i pushed nurmi the fixes, and he verified the EBS side [15:02] New bug: #400404 in libvirt (main) "virsh --help {option} no longer works in shell" [Wishlist,Incomplete] https://launchpad.net/bugs/400404 [15:02] New bug: #411417 in libvirt (main) "virsh restore causes kernel panic in guest." [High,Incomplete] https://launchpad.net/bugs/411417 [15:02] ttx: and uploaded when we were both independently happy [15:02] kirkland: ok [15:02] soren: so EBS is believed to be in good shape, though we lack in first-hand testing and multiple use cases [15:06] any hints on a bug which affect the open (O_WRONLY|O_CREAT|O_TRUNC) syscall on smbfs when and return ENOENT when file exists and success when file does not exist? [15:06] ttx: I tested attaching a volume and saw it appear in the UEC instance> Yes, that does count. I think that's all I really need. [15:09] nomoa: Perhaps some weirdness caused by inconsistent case? E.g. you're trying to the create "Foo", but there's already a "foo" on the remote system? [15:09] nomoa: Pure guesswork. [15:10] soren: no, eg : dd if=somefile of=newfile is ok, dd if=somefile of=existingfile is ko [15:10] strace indicate that open has failed, but it should not [15:11] nomoa: I understand. I'm just suggesting to look for files on the remote system that have names that correspond to the ones you're trying to use, but with different casing. I've seen that cause all sorts of madness before. [15:12] ah ok, I'm afraid that this is not my problem :( [15:12] Fair enough. [15:13] Er, hang on. [15:14] No, I'm just being silly. [15:16] I need help with either apache or firewall or network...not sure which LOL [15:16] ttx: no 100% sure I'll have access to the hardware [15:16] ttx: probably ask EtienneG - he's got the right hardware [15:17] New bug: #230839 in lm-sensors (main) "The sensors program crashes on 2.6.26+" [Medium,Fix released] https://launchpad.net/bugs/230839 [15:17] ttx: I don't think we need to fix bug 455816 for release [15:17] Launchpad bug 455816 in eucalyptus "When installing a UEC cluster, the prompt for the private interface is displayed after the "Installation complete" dialog" [Medium,Confirmed] https://launchpad.net/bugs/455816 [15:17] I set up my apache, I can get to it from my internal IP (192.x.x.x) but I can't hit any of my sites, nor the public IP (216.x.x.x). I ran a port scanner from a website, said port 80 is open, and linux said 0.0.0.0:80 is listening. [15:17] ttx, huh? [15:17] ttx, I will be testing EBS at some time this week, although probably not today [15:18] ttx, not sure what you mean by "multiple use" [15:18] EtienneG: I think we don't talk about the same thing -- I was talking to mathiaz about testing the fix for bug 455816, which requires a server with multiple NICs [15:18] Launchpad bug 455816 in eucalyptus "When installing a UEC cluster, the prompt for the private interface is displayed after the "Installation complete" dialog" [Medium,Confirmed] https://launchpad.net/bugs/455816 [15:19] * soren breaks until the meeting [15:19] ttx, ah, ok. that's my biug, actually [15:19] ttx, it is easy enough to test with VMs. Create one with two NIC, install a cluster controller [15:19] mathiaz: we had that discussion this morning with cjwatson and mdz, and will mention it again at the meeting. Looks like it's difficult to test, and risky to fix in the remaining time we have. [15:20] EtienneG: true. [15:20] ttx: IMO it's just a cosmetic bug [15:20] ttx: it's doesn't break the installation [15:20] EtienneG: ^^ correct? [15:20] mathiaz: that's Colin's opinion as well. [15:21] ttx: we could add a section to the release notes [15:22] cjwatson: works for me re: multiple NIC issue [15:23] mathiaz: mdz: I'd like to make the multiple NIC issue conditional on mathiaz being able to test my branch and confirm it working, then; I don't think we have enough iterations left to be able to justify an untested upload [15:23] mdz: right - EtienneG just gave an easy way to create a test setup [15:23] Is launchpad fubared today. I can't file any bugs from server. [15:25] ttx: I'll give a try to cjwatson's branch on bug 455816 [15:25] Launchpad bug 455816 in eucalyptus "When installing a UEC cluster, the prompt for the private interface is displayed after the "Installation complete" dialog" [Medium,Confirmed] https://launchpad.net/bugs/455816 [15:25] neil-wilson: it's been slow for me today, but is certainly working [15:26] aubre: I think you also have access to a 2-NIC configuration [15:26] aubre: would you be able to test a fix for bug 455816? [15:26] Launchpad bug 455816 in eucalyptus "When installing a UEC cluster, the prompt for the private interface is displayed after the "Installation complete" dialog" [Medium,Confirmed] https://launchpad.net/bugs/455816 [15:27] hello * [15:28] how can I make a backup of my System ? including users, passwords, permissions mail accounts .. ? [15:30] mathiaz: let me take a look [15:31] mathiaz: yes I would be able to test a fix for you, I have 6 NICs in total on the CC, 2 each on the NCs [15:33] mathiaz: how soon would you need me to do it? I need to do some personal stuff sometime today [15:34] aubre: I need to build the package first [15:35] aubre: and do some testing myself [15:35] aubre: the test instructions are bit complicated [15:35] aubre: (they're outlined in the bug report though) [15:35] mathiaz: no problem, I'll check back with you later today === aubre is now known as aubre_afk [15:35] aubre: so I could provide a package to test in a couple of hours [15:36] aubre: a few hours after the end of the server meeting [15:36] mathiaz: ok sounds good [15:36] is anyone else having trouble with usb-creator and today's server iso? [15:49] kirkland: no [15:49] kirkland: I used usb-creator with both i386 and amd64 images === jono_ is now known as jono [16:02] Team meeting on #ubuntu-meeting [16:13] I need some help with apache/firewall...I think [16:14] I set up my apache, I can get to it from my internal IP (192.x.x.x) but I can't hit any of my sites, nor the public IP (216.x.x.x). I ran a port scanner from a website, said port 80 is open, and linux said 0.0.0.0:80 is listening. [16:14] any ideas? [16:15] Where can I find a bash shell tutorial [16:15] on the net [16:15] or a linux use tutorial or somethin [16:16] what do you need to know about bash? [16:16] well im doing my thesis on linux server [16:16] and i have to first learn basics on how to use linux [16:16] so thats what i need to nknow [16:17] LOL, I've been using linux for six years, and I'm STILL learning 'basics' [16:17] pretty lame if u ask me [16:17] honestly, I would just google "howto linux termianl" [16:17] *terminal [16:17] ok [16:18] * skuld wonders if everybody in here is taking a coffee break... [16:19] It's 17:19! No even time for a coffee break! [16:19] LOL [16:23] mathiaz: have you seen that mysql 5.1 innodb bug? [16:23] zul: yeah - need to check whether it's a regression from jaunty [16:23] zul: and if it's true on karmic [16:24] ok sorry to nag ;) [16:24] jdstrand: Can you pastebin the output of "virsh capabilities"? [16:28] does anybody know of a way to CLEAR all my iptables so I can start over with a new firewall config? [16:29] iptables -F [16:29] iptables -X === RoyK^ is now known as RoyK [16:31] skuld: be sure to *not* lock yourself out when doing so [16:31] skuld: ie - check the default policies for iptables [16:32] soren: http://paste.ubuntu.com/297593/ [16:32] soren: what's going on? === bdmurray_ is now known as bdmurray [16:34] jdstrand: Ok, that looks correct. In that case, it probably /is/ a virt-manager problem, which I suspected it wasn't. That's good :) [16:35] soren: right-- I updated bug #453495 with that information [16:35] Launchpad bug 453495 in virt-manager "virt-manager does not honor other architectures when using qemu" [High,Confirmed] https://launchpad.net/bugs/453495 [16:36] jdstrand: Good boy :) [16:36] jdstrand: Are you attaching the actual output? [16:37] soren: no, I just said they were available. I certainly can [16:37] jdstrand: Please do. If someone else should decide to look at this and don't know how awesome you are, they might not take your word for it :) [16:37] heheh [16:39] soren: done [16:39] * soren hugs jdstrand [16:39] :) [16:40] jdstrand: Could you attach the full domain xml as well, please? [16:40] There's another thing I'd like to check. [16:41] soren: you mean for one of the guests (eg arm) that ends up with x86_64? [16:41] jdstrand: Exactly. [16:42] jdstrand: I.e. your "virsh dumpxml testme |grep qemu" with out the grep. [16:42] sure, hold on [16:42] Heh... You have this in your apport output: LiveMediaBuild: Ubuntu 9.04 "Jaunty Jackalope" - Release amd64 (20090420.1) [16:43] I wonder what that's all about. [16:43] I noticed that too. I figured that must have been the disk I used to install on this machine [16:43] *shrug* [16:44] i'd like to document somewhere that you can configure your dhcp server to ignore d0:0d addresses if you wish for it to ignore UEC instances. [16:44] where would i put that? [16:44] kirkland, ttx ? [16:45] smoser: wiki :-) [16:45] Commented out in the default dhcp3-server config file? [16:45] smoser: we'll need to find the rigtht place [16:47] hi again [16:48] flushing my iptables, killed my internet connection sharing for my internal network [16:48] soren: attached. I even threw in dominfo for giggle [16:48] s [16:49] so, I have the firewall set again as I should be. Apache doesn't seem to be responding to my public ip on port 80 [16:54] skuld: Are you behind a router? [16:56] Hey guys [16:56] is this the right place for the ubuntu cloud server? [16:57] * nrvate this is your mind in a cloud [16:57] should I be afraid? :o [16:58] DARE to be afraid :> [16:58] I dare to be.. insidiously naked. [16:58] How about them apples? [16:59] no. LInux is acting as my firewall, I have firestarter for internet connection sharing for my eth2 [17:00] sorry for my delayed response, my dad needed some help [17:01] skuld: So the web server is directly attached to the internet? [17:01] lol so, this is the right place, or not? [17:01] yes [17:03] skuld: and your server is aware of your public IP? [17:03] skuld: I.e. does one of its interfaces have the public IP you're trying to connect to? [17:03] I think so? [17:04] Well, /I/ don't know :) [17:04] eth0 is set static for my public ip [17:04] Check it. [17:04] Ok. [17:06] Alright well ill just ask my question I guess heh [17:06] New bug: #456224 in bind9 (main) "Installing bind9 with forwarders causes loss of hostname resolution." [Low,Incomplete] https://launchpad.net/bugs/456224 [17:06] Basicall I have 3 servers, and Im wondering if I set them up in a cloud, how that works interoperably [17:06] like for example [17:06] if one server goes down [17:07] do the other two somehow reconfigure to be reciprocal to requests? [17:07] or do I require one server to delegate requests to child servers [17:08] soren: here is my addr info http://pastebin.com/f2a632bfb [17:15] any idea why my eth2 is in a state of unknown? I don't think that would affect apache though [17:15] * soren calls it a day [17:16] buh [17:16] someone somewhere must have used ubuntu cloud [17:16] Hajuu: people mean a lot of different things when they say "cloud" [17:17] in our case, it's about having a self-service API to manage virtual machines, a la EC2 [17:17] what you're describing is more high-availability rather than what we call "cloud" [17:17] when I say ubuntu cloud server [17:17] I mean.. ubuntu cloud server. [17:17] not entirely suprisingly [17:17] http://www.ubuntu.com/products/whatisubuntu/serveredition/cloud === jono_ is now known as jono === jono is now known as Guest86059 === Guest86059 is now known as jono [17:22] I'm not sure what it is...but it's starting to tick me off! LOL [17:22] Hi guys. [17:22] I have NSS LDAP set up on a server. [17:22] But id gives me no such user. [17:22] I've straced id, and found it's not contacting the LDAP server. [17:38] When trying apt-get purge package, I get this error: Package is in a very bad inconsistent state - you should reinstall it before attempting a removal.... How do I do this? I checked man apt-get but it gives -reinstall as an option, but that doesn't work either.. how can I get rid of this package? [17:44] * jdstrand is a bit boggled that running libvirt-bin in a jaunty/schroot actually seems to be working [17:47] even more so that kvm seems to be working... [17:47] kvm seems pretty stable these days [17:47] nice and fast too :D [17:48] yeah, but I am running karmic, with karmic kernel, then schroot into jaunty, install libvirt-bin and kvm, and create machines in there and have them work [17:48] Could anybody take a look at this one? How can I fix this apt-get problem I cant remove, purge, reinstall or just install the package anymore :( http://pastebin.com/m55803a4e [17:48] geez. [17:48] I really wasn't expecting that [17:48] that is a bit surprusing [17:54] sven_oostenbrink apt-get clean [17:55] apt-get autoclean [17:55] then apt-get install sun-java6-jre [18:00] can I ask a favor? can somebody please try to get to http://www.localexpresscl.com and let me know if you can connect? Thank you! [18:03] mathiaz: mdz: nijaba: https://help.ubuntu.com/community/UEC/Glossary [18:04] I started that page, as the documentation glosses over a lot of these terms that are probably unfamiliar to many server admin's [18:04] skuld: got 200 OK from wget --spider http://www.localexpresscl.com [18:04] We are supporting Ubuntu [18:04] By preventing piracy. [18:04] One star trek at a time! [18:04] why can't I connect to the site then? [18:05] also I can't find the logs for apache2...I thought they were in /etc/apache2 [18:05] skuld: /var/log/apache2 [18:07] kirkland: nice, thanks [18:07] kirkland: hopefully thursday we can add "multi-cluster" and "" to that list ;-) [18:07] mdz: it's growing as we speak [18:07] mdz: i'm going through both doc's now [18:08] mdz: while installing today's ISOs [18:08] [01:09] -> -pici- you know, banning someone after they say 'anyway, ill stop too' is really just asking for a botnet attack. Not neccisarilly from me, but I'm just saying [18:08] [01:09] -> -pici- just because you have a lot of users, doesnt mean they will stand for tyranny [18:08] [01:09] -> -pici- I would also direct you to the many people I helped in my time in there. Kthxbye. [18:09] mdz: for the server upgrade test I basically told the user to install jaunty select all the tasks in tasksel and then upgrade and perform the tests in the individual tests (basically the same what the desktop team has) [18:10] zul: is it actually possible to install with all of the tasks (no conflicts)? [18:10] zul: if so, that sounds reasonable [18:10] I can't see anything in the logs that tell me why I can't connect via firefox. [18:11] zul: what's the URL for the test case? [18:11] mdz: it should be able i need to test if the testcase is valid [18:11] mdz: i havent posted it yet i wanted to make sure if it was valid first [18:11] zul: ah, ok [18:12] I'm not behind a router either. I don't get it...I can connect to the internal server ip (192.x.x.x) though [18:13] I should be able to connect using both my public IP (216.x.x.x) and private (192.x.x.x), right? [18:17] anybody? Am I crazy? [18:17] mathiaz: oh btw puppet is finally in main [18:17] skuld: yes unless if you have a firewall blocking port 80 [18:18] my firewall states 0.0.0.0:80 is open [18:18] anything in your error_logs? [18:18] everything looks okay...I see lots of 200's [18:18] unless an upstream router is blocking port 80 [18:19] is there a command I can use to check that? [18:19] nmap [18:20] okay, installing it [18:21] New bug: #456381 in apache2 (main) "/etc/init.d/apache2 missing option to stop gracefully" [Undecided,New] https://launchpad.net/bugs/456381 [18:23] do you know what command I would use with nmap to look for port 80? [18:25] foudn it....it says (nmap --opem 216.x.x.x) port 80 is open [18:28] any other ideas I could try? [18:28] kirkland: nice Golassary, thanks [18:30] skuld: Are you trying to connect to that address from within your network? Some routers don't allow you to make loopback connections like this. [18:30] yes [18:30] I didn't have a problem before using redhat, very simular config [18:31] skuld: If its running on the same address that you're connected to IRC as, then I can see it. (I see index.html, gallery.html, partners.html) [18:31] etc [18:32] hah squeeze will be the last kernel that has dom0 support [18:32] zul|mdz: the jaunty server tasksel options all install with no conflicts. [18:33] sbeattie: cool i just finished an install as well [18:35] I have my Ubuntu Samba File server set up so I can access it from computers my desktop running XP but now and my laptop when its running Ubuntu 9.10 (which is really really easy :) ) but I can't seem to connect to it when I'm running Windows Vista on my laptop and I've sept a few hours trying to figure it out without and luck can anyone help me out? [18:36] @Pici: Can you get to http://www.localexpresscl.com ? [18:39] zul: bug 456381 - I think the bug description is clear [18:39] Launchpad bug 456381 in apache2 "/etc/init.d/apache2 missing option to stop gracefully" [Low,Incomplete] https://launchpad.net/bugs/456381 [18:40] mathiaz: i agreee but he didnt specify a version [18:40] zul: right - just asking which version may be enough [18:41] mathiaz: in hind sight yes [18:41] hello to all ! i have a question about the security of ubuntu server [18:41] could it be that ubuntu server is unsecure and very easy hackable [18:41] becouse i have allways updated my ubuntu server [18:42] but i am seeng some strange logs on my side [18:42] and my network isnt really easy to be access from the outside [18:42] Generally it isn't, but it's not out of the question that you changed something to make it so. [18:42] i am running a standard ubuntu server version [18:43] the strange thing is [18:43] that they do some piping of my server [18:44] will show you just a log now [18:44] 02:12:50�63.131.18.154�static-63-131-18-154.har.onecommunications.net�&nb05.188.251.26:443� [18:44] for some reason somebody call my webpage and do call a another website [18:45] the destination adress is port 443 ip adress 05.188.251.26 [18:46] the call of the website comes from a static-63-131-18-154.har.onecommunications.net [18:46] whats that for a fuck? [18:46] never seen something like this till yet [18:46] Running your own php scripts? [18:47] yes [18:47] self written [18:47] know every line of the script [18:47] skuld: Yes, it points to the same thing I was looking at before. [18:47] Pici: do you have also this problem ? [18:49] you see a directory listing instead of the website? [18:49] skuld: Yes. [18:49] hmmm, where is the DirectoryIndex option supposed to go in the conf file? [18:49] I probably have it in the wrong spot [18:51] does it go in the section or right under DocumentRoot [18:58] New bug: #343898 in dhcp3 (main) "dhclient-script apparmor issues" [Undecided,Fix released] https://launchpad.net/bugs/343898 [19:14] skuld did anyone help you fix your thing [19:19] do I need to do anything with httpd.conf when setting up a new apache server? seems like all of my virtual domains are pointing to the very first directory in my /var/www folder instead of pointing to where they are supposed to be. [19:19] still working on it...I sort of fixed the direcotry listing problem... now i have a new issue [19:24] could I have a problem with a "loopback" [19:27] and...I have a problem with all of my virtual domains aren't working, only the first directory in my /var/www folder is showing up when I browse to any of my domians [19:30] skuld, the default apache configuration doesn't actually use httpd.conf [19:30] you have to set the document root for each site, in the sites-available/site-name file [19:31] right. [19:32] do I have to end each site....intelatech.net with .conf (intelatech.net.conf) for the virual hosting to work? [19:32] you also have to set the domains for each site [19:32] no, not at all [19:32] example, > [19:32] if you have one that listens by IP, or on *:80, it will break all the other sites [19:33] and if you use domain names, DNS has to work [19:33] AH! that's what I'm missing...let me try that [19:33] the other option is using a different IP for each site, and configuring virtual interfaces [19:33] only really practical with a couple sites [19:34] so if I do do I have to add another one under the entire section for www. ( ? [19:35] no, there's another way to work around the www. thing [19:36] the best way I know of is rewrite rule [19:37] I'm all ears! :) [19:38] RewriteEngine on [19:38] RewriteCond %{HTTP_HOST} ^www\.domain\.com [19:38] RewriteRule ^(.*)$ http://domain.com/$1 [R=permanent,L] [19:38] I think that's correct [19:38] the other thing you need to do is enable the rewrite module [19:38] sudo a2enmod rewrite [19:40] the reason I like this way, as opposed to working around it [19:40] is it forces the user to use domain.com, instead of having two different sites [19:40] fixes issues with cookies [19:42] okay. now I just need to fix this possible 'loopback' problem where I can't see any of my websites [19:42] well, you need to check each site [19:42] make sure none of them are listening on *, *:80, or your IP [19:42] and then make sure that the local system is resolving DNS names correctly [19:43] apache flips if it can't find all your FQDNs [19:43] Hi all - if anyone cares, I got a freeze exception for monit - https://bugs.edge.launchpad.net/ubuntu/karmic/+source/monit/+bug/426402 [19:43] Launchpad bug 426402 in monit "Sync monit (1:5.0.3-3) from Debian Sid" [Undecided,Fix released] [19:43] It's here: https://edge.launchpad.net/ubuntu/+source/monit/1:5.0.3-3 [19:44] MagicFab: cool [19:44] oh. now I changed all my config files so that , but I got a bunch of warnings and mentionings of things taking precedence [19:45] well, the change is correct, so you'll have to hunt down those warnings [19:46] don't I need like a servername directive? [19:46] yes [19:47] "If you are using name-based virtual hosts, the ServerName inside a section specifies what hostname must appear in the request's Host: header to match this virtual host." [19:47] so, each site should have a servername matching the one in the virtualhost directive [19:50] sommer: Thanks for the server guide... I'm reading it now. [19:56] brb [20:04] mdz: i've revamped, refactored, improved, and cleaned up https://help.ubuntu.com/community/UEC [20:05] mdz: i'll work on the multi-cluster one after i take a break and get some food [20:05] hi [20:05] what is the config file for x server in ubuntu-server...? i mean the equivalent of xorg.conf [20:06] hey all, i'm running ubuntu hardy server and i ran updates, now when i try and use certain programs, /dev/null is chmodded to 600, so a non-root user trying to access anything that needs it dies, i saw a launchpad bug about it but no real solution except to add an init script to change it to 666, is there a real solution yet? === crohakon|afk is now known as crohakon [20:09] Kirsch: I seen that happen before when someone specifies /dev/null as a log file in a configuration file somewhere [20:09] Kirsch: did you set anything to /dev/null in /etc? === adam is now known as Guest72722 [20:12] hi all, i dunno if my question about /dev/null went through if anyone saw it, but i'm back [20:13] Guest72722: Kirsch: I seen that happen before when someone specifies /dev/null as a log file in a configuration file somewhere [20:13] Kirsch: did you set anything to /dev/null in /etc? [20:15] okay, so I need both and ServerName localexrepsscl.com ? [20:23] hey all, i am having a /dev/null permission issue as 600, i've googled around and i see a lot of temp solutions, like setting an init script to change it to 600, is the real cause found? [20:27] I think they know it was bad for hardy, but they cant change the original chematic of the dist because its an LTS. Id like to see more people to the coders, to change things like this. Or to add in new versions of programs into older dists that are still active on the net. [20:27] and more power to the coders. [20:27] hey guys [20:28] Kirsch: I seen that happen before when someone specifies /dev/null as a log file in a configuration file somewhere [20:28] Mornin Doug [20:28] Kirsch: did you set anything to /dev/null in /etc? [20:28] I am short a php exten [20:28] the php5-rewrite module [20:30] remove and purge php then reinstall it, unless its a source install ? [20:31] I got issue with trying install the RTL8101E driver. I am getting r8101.ko no such file [20:32] Its a very new RTL card ? [20:33] this computer is one year old [20:34] modprobe --list | grep -i rtl [20:34] Its a Realtech card [20:35] ok I got a list of stuff === nrvate5i2 is now known as nrvate [20:35] How many lines ? [20:35] 4 [20:36] rtl8180.ko could work perhaps. Try to insmod it [20:49] Panzer: Working ? [20:50] no. the r8180 is a wireless mod. [20:52] 8139too.ko or 8139cp.ko ? [20:53] where? [20:53] Thought you had a wireless, sorry about that. [20:53] modprobe 8139too [20:53] Hello, general server spec question: ~100 (frequent) accessors querying a 1.5m+ record database... the actual DB and the files referenced by the DB will be on a RAID array with 15k rpm drives, but the system partition may only be on the internal RAID 5 7.5k RPM drives... would this be a serious speed hindrance? All equipment mentioned is theoretical right now, I'm just trying to determine if I should spring for faster drives for the system partitions as w [20:55] Tohuw: the system disc access is only required if the DB wants access to system files.' [20:56] Panzer: remove the modules that didnt work, with rmmod ModuleName [20:56] none of them including the ones you have listed have worked [20:57] google for the solution [21:00] yea I have. none of the solutions say anything about r8168.ko or r8101.ko file not found issues. [21:01] see http://ubuntuforums.org/showthread.php?t=843398 and http://ubuntuforums.org/showthread.php?p=4210510#post4210510 [21:02] I don't think there is a 9.04 patch [21:14] this is driving me crazy. [21:15] skuld: yea my issue too [21:16] I can ping intelatech.net, I get the public server eth0 IP address. yet from within my network, I use firefox, http://intelatech.net and I get "connection not established" [21:16] I have no router [21:17] I'm sure I'm probably missing something really small in my configuration....somewhere...It worked before I installed ubuntu with fedora [21:20] and I can't test any of my virtualhosts until I get this resolved! [21:21] skuld: my first gues: a firewall rule? [21:26] Question, I need information on a 10Gbit network card that also supports iSCSI.. Could I use such a card for bot iSCSI and normal traffic, or would that require an extra, other, network card? [21:27] I'm thinking it has to be somewhere in the firewall, but I can't find it. I don't even remember which commands I ran to see that 0.0.0.0:80 is open [21:34] skuld: try netstat -an | more [21:37] that shows it in the list [21:37] what a long listing [21:40] somebody care to examine my iptables...it looks all greek to me [21:41] skuld, sure, pastebin it [21:41] http://pastebin.com/f1192ec32 [21:41] thanks! [21:42] that's pretty complex [21:42] okay, so I'm not crazy [21:43] I wouldn't doubt if that's the problem [21:43] any reason it's set up that way? [21:44] I had to use firestarter to get my laptop to connect to the internet using it's Internet Connection Sharing. If I knew how to do that and only have the handful of ports open that I need...plus being able to see my own hosted web pages from my laptop, I'd gladly get rid of firestarter and do it that way [21:45] well [21:45] it definitely doesn't need to be that complex to do so [21:45] but it's not exactly simple either [21:45] you need masquerading, but also port forwarding [21:46] I'm guessing, anyway? [21:46] or are all the ports you need opened on that machine [21:48] went back for an example, if I want a window to move on buzzed then I should intercept the signal, get the window handle and move it that's it ?? [21:48] sorry, I had a phone call [21:49] I only need 21, 25, 26, 80, 100, 4000-4050 (forwarded to spec IP), 9100-9115 (forwarded to spec IP) open on the server [21:49] I guess everything else in there was for the ICS (Internet connection sharing) [21:50] and since I use IRC, I guess 6667 too [21:50] but I didn't explicitly open that port [21:50] and it works [22:05] kirkland: why are you moving everything in the UEC tuto to sub pages? I am not sure this is making it easier to follow [22:05] nijaba: it will [22:06] nijaba: so much of this is common operations, whether doing a simple install, or a complex install [22:06] nijaba: getting credentials [22:06] nijaba: bundling an image [22:06] nijaba: running an image [22:06] nijaba: registering a node [22:06] nijaba: otherwise, we will have so much duplicated operations [22:06] nijaba: we *already* do [22:07] nijaba: i'm trying to clean this up [22:07] nijaba: then we build pages that point to the incremental steps [22:07] kirkland: yes, I understand the factorisation, but it makes it quite difficult to print, and since in each step you now offer many options, I am afraid first time iser may get more confused [22:08] nijaba: it may be slightly harder to print [22:08] which is the opposite of the goals that were set when ttx and I were asked to do a simple step by step set of instructions [22:08] nijaba: but i disagree that it's more confusing [22:08] nijaba: i think it's less confusing [22:08] skuld, so you do need port forwarding as well [22:08] kirkland: well, we'll see Neil's and Mark's reaction to that, as they were the ones requesting it in the first place [22:09] skuld, I don't have any time right now, I have to go to class, but you might try fixing it with firestarter [22:09] skuld, add exceptions to allow connections from the LAN to port 80 [22:12] okay, I'll give that a ttry [22:15] mdz: hmm, I'm extending the UEC documentation now in the wiki, starting at https://help.ubuntu.com/community/UEC [22:15] kirkland, have you considered using includes rather than linking to the sub-pages? [22:16] mdz: didn't know that was possible; will use that [22:16] kirkland, that would let you avoid duplication while still providing all of the necessary information on one page [22:16] mdz: that's perfect, thanks [22:16] kirkland, <> [22:19] kirkland: and to probably use Include, I usually add delimiters via comments in the wiki pages [22:19] well, that didn't work [22:19] LOL [22:20] kirkland: things like: ## StartGeneratePubKey [22:20] kirkland: and then you can use the Include macro with start='## StartGeneratePubKey' options [22:20] kirkland: similarly there is an end option [22:21] mdz: perfect [22:21] mathiaz: cool, thanks [22:21] mathiaz, neat, I didn't know about that [22:22] mdz: see https://wiki.ubuntu.com/TeamReports/May2009 [22:22] nijaba: https://help.ubuntu.com/community/UEC [22:22] nijaba: there you go ;-) [22:23] mdz: that's how I usually include the server team report in the team report [22:23] mdz, kirkland: wonderfull -> best of both worlds [22:29] nijaba: do you mind if I move UEC -> UEC/CDInstall, and make a new UEC page that links to 1. UEC/CDInstall (simple, basic setup) and 2. UEC/PackageInstall (complex, custom setup) [22:29] kirkland: not at all, I think that would be fine [22:29] nijaba: thx [22:32] kirkland: you have a UEC/CDINSTALL page? [22:35] SyL: https://help.ubuntu.com/community/UEC/CDInstall [22:35] SyL: work-in-progress [22:35] kirkland: would suggest a page with all-in-one as well [22:36] SyL: what do you mean? [22:36] cluster and node on the same machine [22:37] kirkland: and have you gotten it working after updating the packages? [22:40] SyL: i have not gotten nc on the same system as the cc [22:41] SyL: i understand from nurmi that this is not recommended, or even possible, due to some networking configuration reasons [22:43] oh perfect, i just wanted to ask how many machines i need to build my own uec for testing purposes [22:43] heh [22:43] did i get that right that i need at leas two physical machines? [22:43] ahe: 2, yes [22:45] so i think i need to go shopping because i don't think that any of my spare machines fits the minimum hardware requirements [22:45] kirkland: I thought it could work with only one machine [22:46] SyL: where did you see this described? [22:46] SyL: i'm willing to entertain the possibilty, but I have not succeeded in doing such things [22:48] kirkland: I remember someone getting it working a few months ago. I don't see why it wouldn't even if it was for a quick proof of concept [22:48] SyL: you're welcome to try [22:56] New bug: #328550 in udev (main) "qla2xxx takes ~one minute to initialize per device" [Medium,Won't fix] https://launchpad.net/bugs/328550 [23:09] kirkland: EtienneG got it to work on a single machine, but it is a ugly that he is not willing to document. [23:09] nijaba: gotcha, that's what I suspect [23:10] SyL: ^ [23:10] kirkland: the minimum requirements are maybe a bit too high for a simple "discovery" install. Will work fine one lower machine, just a bit slower... [23:10] s/one/on [23:11] nijaba: okay, set the Minimum ones to what you think [23:11] kirkland: ok, I'll do that and then -> bed [23:11] nijaba: i know that my installs were *failing* because I was installing from a 32GB SSD, and it didn't have enough space to run a single VM (this was back when we had 10GB images, though) [23:13] kirkland: disk wise, I think your reco are fine. It is just that you can start a cc + cloud machine with just 512Mo ram, and 1G is what I would say is minimum for the nc [23:13] nijaba: fair enough; you're welcome to update those Minimums [23:15] kirkland: done. and now, goodnight :) [23:16] zul: have you seen some issue with karmic and nmbd? [23:16] zul: it seems that sometimes nmbd doesn't stat [23:16] zul: *start* [23:16] kirkland: but it works for you on 2 machiens out of the box before updating packages? [23:16] zul: at boot [23:17] SyL: it will if you take the 20091020 iso from cdimage.ubuntu.com [23:17] zul: I suspect that nmbd starts when the network is not ready yet - the last line in the log.nmbd is 2009/10/20 18:15:01, 0] nmbd/nmbd.c:693(open_sockets) [23:17] there is no data, only zul. [23:19] nijaba: night [23:27] nijaba: ok, I will try this again [23:28] are there any common reasons for why it takes roughly 35 seconds to get a prompt after entering my password at login? [23:28] nijaba: server too? [23:49] hi, is it possible to setup vsftpd server using this settings..------->http://pastebin.com/m589bf6a1