[00:00] update-manager-core yay [02:11] hi === dendro-afk is now known as dendrobates [02:53] hello [02:54] is there anyone that can help me for a minut [02:54] minute* [02:55] !anyone > Guest49270 [02:55] Guest49270, please see my private message [03:00] hi how do i add up a temporarry sudoers..on my ubuntu server.. [03:03] anyone..? [03:04] yes i see , does anyone know how to use cat to write to the /etc/network/interfaces file, [03:04] ruben23: Add them to the admin group [03:05] Guest49270: You don't use cat to write. Use nono or vim or so. Like: sudo nano /etc/network/interfaces or: sudo vim /etc/network/interfaces [03:05] I am attempting to make a script to automate the proccess [03:07] the problem is when i use cat to append to the file even with su i get the "bash: /etc/network/interfaces: Permission denied" error [03:10] ?is there another way i can do it === dendrobates is now known as dendro-afk [03:11] Guest49270: cat somefile | sudo tee /etc/network/interfaces (optionally with -a for append instead of replace/create) [03:11] Guest49270: I suggest using sudoedit [03:12] where would the -a go? [03:12] Guest49270: if you don't know, you probably shouldn't be using tee. [03:12] and sudo edit doesn;t work ether [03:12] i can learn [03:13] Guest49270: sudoedit has no space between sudo and edit. One command, sudoedit. [03:13] Guest49270: To learn about tee, man tee [03:13] rutri@Athena:/media/STORAGE/Scripts$ sudoedit echo "# Bridged network interface" >> /etc/network/interfaces [03:13] bash: /etc/network/interfaces: Permission denied [03:13] rutri@Athena:/media/STORAGE/Scripts$ [03:14] im looking at tee now [03:14] Guest49270: Huh? WHat are you doing? sudoedit takes one parameter, the file to be edited. [03:14] hmmm [03:14] ok === dendro-afk is now known as dendrobates [03:15] so how would i use sudoedit in a script file? [03:16] You wouldn't. You can't generally use sudo or sudoedit in an unattended script, it is likely to prompt you for a password... [03:16] yes well i want to automate the edit [03:17] Why would you need to automatically re-edit your network interfaces file?? [03:17] also the tee command will not work for my needs [03:17] because i am making an installation script [03:17] Guest49270: Then it should be run as root and there is no need for sudo at all. [03:18] The installer runs as root... [03:18] yes i have done that using the command "sudo ./myfile.sh" [03:18] i get the same problem [03:18] That's not an installation script... you'd need to already have an installed system to be able to do that. [03:19] hmm [03:19] When and how will this "installation script" be run? be specific. [03:19] * genii makes more coffee, hands out the mugs [03:19] genii: FSVO mug = drip bag [03:19] For automated installations you want to read up on preseeding and perhaps kickstart. [03:19] well do the standard install then i am making a script to install the other applications I want to use like KVM in this case. [03:20] Guest49270: that should be done via preseeding [03:20] Definitely preseeding [03:20] hmm ok [03:20] Guest49270: preseeding is documented in the installation-guide [03:21] will that allow me to make my own iso for ubuntu? [03:21] Guest49270: you don't need to, but that could be done. [03:22] so preseeding is the only way? [03:22] It's the Right Way [03:23] You could pay around using echo foo |sudo tee -a somefile # But it's ugly and you don't seem to know enough to avoid possibly bad things if you make mistakes in that script... [03:23] ok that would be in the ubuntu documentation? [03:23] Guest49270: apt-get installation-guide [03:23] *apt-get install [03:23] if i kill the system i can always renstall [03:24] https://help.ubuntu.com/9.10/installation-guide/powerpc/preseed-using.html [03:25] You are the system administrator of this machine. Learn the tools of your trade. [03:25] on second I am looking at it [03:25] Guest49270: note that he linked to the PPC version [03:26] lol im a programmer this is for my house depending on how much time is involved i may just do it all manually [03:26] apt-get should give you the version appropriate for your Ubuntu release and architecture. [03:26] Ah... just whatever Google found me... [03:27] yes i am googling it as well [03:27] i did find this https://help.ubuntu.com/community/Installation/UnattendedCD [03:28] the document also referencces the PPC version as well [03:29] Come on... https://help.ubuntu.com/9.10/installation-guide/amd64/appendix-preseed.html for the AMD64 version. [03:30] Hm. In https://help.ubuntu.com/9.10/installation-guide/i386/index.html 'Installing Ubuntu 9.04 "Jaunty Jackelope" for i386' When the url should point to Karmic [03:30] yes looking at it [03:30] genii: Good catch, file a documentation bug :) [03:31] jmarsden: Will do [03:31] The URL does point to Karmic, at least. [03:31] genii: installation-guide is maintained in SVN, IIRC. You should submit a patch. [03:31] what i may do is just copy over the existing file as sudo [03:31] Guest49270: suggest you investigate puppet/chef/cfengine [03:32] twb: For one machine? Seems overkill, since he doesn't want to spend the time even to learn preseeding :) [03:32] yes [03:32] jmarsden: true [03:32] I don't understand why he's scripting an install of ONE machine in the first place === dendrobates is now known as dendro-afk [03:33] Guest49270: Now you have at least 3 ways to do this: (a) echo stuff |sudo tee -a /path/to/file (b) sudoedit /path/to/file (c) sudo cp newfile /path/to/file (d) use preseeding. [03:33] i want to automate it and i know a bit out about scripts so i figured i could just make a script file to automate it. [03:34] the reason I am making a script file is because I do not want to do this again the next time i mess up my server [03:34] Guest49270: Oh, five: (e) sudo -s and then run the script in a root shell. [03:34] So pick one and do it. [03:34] Guest49270: the right way to do that would be to make backups of /etc (e.g. etckeeper), and to keep a small notepad listing the changes you make [03:34] Guest49270: it will be FAR more reliable than a scripted process [03:35] jmarsden: sudo -i (usually) beats sudo -s [03:35] sure [03:35] Guest49270: here's an example: http://twb.ath.cx/Preferences/doc/ymir-changelog.txt [03:36] twb: Re -i vs -s, sure... but for editing a few files it doesn't seem likely to make a difference. [03:36] jmarsden: well, it means that root will own your emacs .saveplaces.eld file :-) [03:36] well i figured if i used a script then i could just run the script next time instead of using the notes. since a script would be as easy as writing notes i would save time doing it. [03:36] New bug: #505045 in dhcp3 (main) "DHCP timeout" [Undecided,New] https://launchpad.net/bugs/505045 [03:37] ill look up sudo -i and -s that may work. [03:37] Guest49270: You've spent about 40 minutes so far. How long does writing notes take? [03:37] one question. [03:38] so is the "bash: /etc/network/interfaces: Permission denied" because sudo does not have write access or because the file is already open? [03:39] No, it is because the redirection is happening as you not as root. Sounds like you may understand less about scripting that you think. [03:39] that is verry likely [03:40] hence, this is also a learning experience [03:40] im used to doing direct function calls in C i dont know much about the ubuntu [03:40] When you do somecommand -whatever -options >>somefile # the >>filefile is done in that shell running that command. Even if that command is sudo or su. [03:41] ic [03:41] BTW the same is true in the Windows shell... and in Solaris and whatever other OS you may be familiar with. this is not Ubuntu specific. [03:42] It's not true in Genera [03:43] so the shell does not have access because it is running as my user not truly as root. This must be what they talked about in the ubuntu documentation. [03:43] twb: What fraction of newcomers would be familiar with Genera but not all the other shells? :) [03:43] Guest49270: Yes. [03:43] ok i will try the sudo -i then [03:44] well let me see what it does brb [03:45] jmarsden: Genera isn't a shell, it's an OS [03:45] twb: So if one ported bash to Genera it would suddenly behave differently in this regard? Seems odd... [03:45] Probably people who went to SAIL in the 70s and then switched to, say, a med degree would be familiar with it and not bash [03:45] so sudo -i logs you in as root? [03:46] jmarsden: Genera is a Symbolics Lisp emulator for Alphas [03:46] Just as an example of a place where >>foo won't perform redirection [03:47] Guest49270: In effect. It creates you a root shell and goes through the root login process to set up your environment. It doesn't actually log you in directly as root, you can (for example) suspend that shell and get back to your original shell. [03:47] so would the command "sudo -i -k myfile.sh" would be correct? [03:48] i added -k to force the system to ask for a password after [03:48] how do you set a static ip with ifconfig [03:48] twb: For other examples, CP/M from the 1980s or DOS from before DOS 2.11 also lack such redirection :) [03:48] just one shot type a thing [03:49] sabgenton: sudo /sbin/ifconfig eth0 1.2.3.4 [03:49] jmarsden: will that work on wifi? [03:49] or do i need iwconfig [03:49] sabgenton: I have no idea. [03:49] ok thats a start thanks [03:50] You're welcome. [03:50] jmarsden: why do you give the full path out of interest? [03:50] sabgenton: An abundance of caution, and in case you don't have /sbin in your $PATH. [03:50] ah [03:50] thought so see [03:51] :) [03:51] hmm sudo -i and -s didn't work ether [03:51] the file must be getting used by the os [03:51] Type sudo -i # no other parameters. In the root shell, run the script. [03:52] You are making this about 100x harder than it needs to be. [03:52] sudo -i ./test.sh [03:52] NO! [03:52] sudo -i [03:52] Then, in the shell, ./myscript.sh # or whatever [03:52] ok [03:52] one second [03:54] still didnt work [03:54] the confusing part is that another file is generated [03:55] jmarsden: It may have been more proper to do iwconfig as it let me auto complete but it work with ifconfig [03:55] :) [03:55] Guest49270: Then your script is probably broken. pastebin the script and pastebin the command you typed and what output you got. [03:55] file:///etc/network/interfaces/r/r is what is generated [03:56] Guest49270: BTW you have now spent 1 hour on this... wouldn't writing notes have been simpler and quicker? [03:56] yes [03:56] That's a local URL on your machine. Noone else can use it. [03:56] but i want to know why this is not working [03:56] Then pastebin the stuff I asked for. [03:57] jmarsden:after just fiddling with hostapd config with bit's of peaces i got of the net i got AP mode working! [03:57] sabgenton: Good. [03:57] is there a way i can send you my script jmarsden? [03:57] it's actually better than the depreciated madwifi [03:58] Guest49270: Yes, use pastebin [03:58] !pastebin [03:58] For posting multi-line texts into the channel, please use http://ubuntu.pastebin.com | To post !screenshots use http://tinyurl.com/imagebin | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic [03:58] !pastebinit [03:58] pastebinit is the command-line equivelent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the « pastebinit » package from a package manager - Simple usage: command | pastebinit [03:58] well so far the same card I once used is apsolutley responsive [03:58] madwifi was way more bumpy! [03:58] unless it improved to [03:58] but I doubt it [03:58] ath5k rules! [03:59] :D [03:59] sabgenton: That's quite a change from your crying about the loss of madwifi-tools yesterday :) [03:59] *=erm* *cough* [03:59] http://ubuntu.pastebin.com/d106f3920 [03:59] :j [03:59] Today he knows better [04:00] :D [04:00] well at least I'm still not moaning today :D [04:00] (others might) [04:00] :) [04:00] jmarsden did i do it correctly? [04:01] Guest49270: Yes. Now also pastebin me the command you typed to run it and the output it generated, please. [04:02] Guest49270: Um. Was this script file created in Windows? So it has DOS CRLF line endings in it?? [04:02] well i dont need to pastebin that, it was "sudo ./myscript.sh" [04:02] hmm it should not have been [04:02] maybe thats whats messing it up [04:03] OK. the line spacing looks weird... what does file myscript.sh output? [04:03] Oh, and there is a ! missing from the first line, it should read #!/bin/bash [04:03] i was wondering why i seen the two carrage returns at the end of the filename. [04:03] yes [04:04] well the script was not running with the ! in the file [04:04] Guest49270: But you failed to inform the channel of this... [04:04] i dont think ubuntu has bash in it [04:04] :) [04:04] im sorry for that [04:04] i did not think it was important [04:04] Your crlf line endings broke it. Ubuntu has bash but does not have bash\r :) [04:05] i thought that is used to make sure the script only runs in a bash shell [04:05] Had you mentioned that we could have solved that a LONG time ago. [04:05] ahh [04:05] What does file myscript.sh output? [04:05] that woould eplain the \r in the file name [04:05] interfaces\r\r [04:06] Indeed. [04:06] ok now that i know what is wrong [04:06] how do i conver to unix? [04:06] Fix it with dos2unix and then run it again. [04:06] convert8 [04:06] thank you [04:06] (and put the ! back in there) [04:06] You're welcome. [04:07] And next time use a Linux text editor for writing your scripts :) [04:07] i have been [04:07] WHich one? [04:07] gedit [04:07] Hmmm. I've never seen gedit add CRLF line endings to a newly created file. [04:07] if i were using my normal one it would ask me if i wanted to convert it [04:07] well i can explain that [04:08] i used an old script [04:08] i think the file was started in windows [04:08] Guest49270: And the old script... OK. [04:08] then i took over with gedit [04:08] gedit will use whatever is already there. [04:08] exactly [04:08] however it does not inform me of that [04:09] Correct. Emacs does. [04:09] yea [04:09] i have one more quick question [04:10] is there a way to change my default editor in ubuntu [04:10] jmarsden: "Emacs does" - Isn't that a valid response to pretty much any question? [04:10] erichammond: Good point :) Except for questions like "does your editor crash regularly?" :) [04:11] Guest49270: Yes. export EDITOR=/usr/bin/emacs # is one way :) [04:11] You can stick that or something like it in your ~/.bashrc [04:11] yes [04:11] i have had problems using the bashrc before [04:12] but thats something totally different i guess [04:12] Probably. if you prefer, use the select-editor command instead. [04:12] hi is it possible to create directory service for 2 networks like network1-all linux client and network2 all are windows client.. [04:12] That's newer and Ubuntu-specific, I just set EDITOR=emacs :) [04:13] well I am more familiar with the gui so i tend to used gnome [04:13] i wanted to change what editor gets used when i double click the files and hit display [04:14] Bah, EDITOR is for line editors. VISUAL. [04:14] Also, emacsclient -c -t -A '' is your friend. [04:15] yea i use ultraedit [04:15] ruben23: Two different directory services (such as Fedora Directory Server) on the same server? That might be interesting... why would you need two different directory services? [04:15] Guest49270: This channel is #ubuntu-server. Ubuntu Server does not have a GUI. [04:16] yes [04:16] jmarsden: my network will have windows client on a single network and linux client on another network.. [04:16] and the administration is on one only [04:16] the script is being run on ubuntu server thats why im here [04:16] i mean the admin will adminster all of them..is it possible [04:16] ruben23: Yes, but why can't they share one directory service? [04:17] im making modifications on a pc with ubuntu desktop [04:17] jmarsden: cant it be possible single directory service on windows and linux client..? [04:17] then send it to the server over an nfs share [04:17] like one single server adminstration for them all [04:19] ruben23: As far as I know it can. Ah... sorry, I misread your question, I think. [04:19] jmarsden: what application can i used for it..? [04:19] ruben23: I think you could use Fedora Directory service on both networks at once. [04:20] Fedora Directory service can manage windows client..? [04:20] what do i need to install to use dos2unix? [04:21] nm [04:21] got it, suto apt-get install tofrodos [04:21] sudo* [04:23] ok now i get bash: ./test.sh: /bin/bash/: bad interpreter: Not a directory [04:23] Guest49270: Unwanted extra trailing / on the first line of your script. [04:23] Guest49270: Should be #!/bin/bash [04:24] no i used the program you suggested to convert the script file [04:24] Right, and there is another bug in the scirpt, and unwanted extra trailing / character on the first line. [04:24] Remove it. [04:25] You want: #!/bin/bash You have: #!/bin/bash/ [04:25] See the extra / ? [04:25] ahh sorry [04:25] yes i see now [04:26] thankyou very much [04:27] You're welcome. [04:27] the test works [04:27] i know i spent a lot of time on this butt i have learned much [04:27] one more thing [04:28] "sudo apt-get -y -q install kvm qemu libvirt-bin" will automate the install right? [04:29] Guest49270: Looks reasonable. Uninstall them all, and then try it out :) [04:30] thank you agian [04:31] i know this was painful for you but it was very helpfull [04:32] Guest49270: OK. You might want to try reading something like http://tldp.org/LDP/intro-linux/html/intro-linux.html for general Linux background info [04:33] yea i have had a few references [04:33] most of the stuff i can handle with ebox [04:34] however virtual servers is not one of them [04:35] since i cant get a fuppies working on ubuntu server i have to do my trans-coding on a windows server running on ubuntu using KVM [04:36] Just make the investment of a few hours of your time to really learn Linux and bash (and maybe even Emacs) . Then work on getting a "fuppies" (whatever that is) running on Ubuntu. [04:36] fuppies is a media streaming progra for linux [04:37] as for emacs i have no need to learn it i use nano on the server [04:37] Guest49270: Sapir-Whorf. You lose. [04:38] Then... fuppes should run fine in Ubuntu. And using Emacs on all platforms gives you one free powerful editor across all platforms. [04:39] yes well thats another problem all together [04:40] as for emacs i am already familiar with ultraedit and gedit and nano gets me by for how often i change files on the server. [04:42] but anyway i really appreciate the help, now that the carrage return is removed the script works exactly as expected [04:42] thank you once again [04:43] I thought fuppes was already ITPd [04:43] I remember someone talking about it a year ago [04:44] I guess it's still in RFS [04:45] Yeah, #426048 [04:49] There is a thread about fuppes on Ubuntu 9.10: http://ubuntuforums.org/showthread.php?t=1310511 [04:50] Amazing... he wasted an hour on a script he didn't really need, instead of on getting fuppes to work, and he wants to run Windows in a VM instead! [04:51] Anyone called Guest[[:digit:]]+ ain't too bright [05:56] how can i set up so that ubuntu will automatically re-start all users running processes that were there before reboot.... [06:03] matte: with great difficulty [06:07] New bug: #467480 in clamav (main) "package clamav-freshclam 0.95.2+dfsg-4ubuntu4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 100" [Low,Incomplete] https://launchpad.net/bugs/467480 [06:16] really [06:17] doesn't sound like the ubuntu way [06:17] ;) [06:17] matte: Is there a way to do that in any other OS you know of? [06:17] windows, put the apps in the Startup folder [06:17] or with msconfig [06:17] yes [06:17] No, that won't do it at all. [06:17] actually it does [06:17] That would only run the apps you put there, not the ones running at the timeof the reboot [06:18] Which coul dbve very different [06:18] *could be [06:18] well thats what i mean, apps you select to run after a reboot [06:18] like a startup folder [06:18] for linux [06:18] That's not at all what you said though. [06:18] ok well thanks for making that clear [06:19] matte: Is this for a server (in which case you can put apps into /etc/rc.local or into ~/.bashrc files for each user, etc) ? Or for a desktop with a GUI, in which case I don't know the answer, ask in #ubuntu [06:19] ok awesome [06:19] thank you [06:20] You're welcome. [06:45] does anyone know what the default password is for a newly added user using the command "sudo useradd -m rutri" [06:47] Guest89077: don't use useradd. [06:47] Guest89077: use adduser [06:48] ok what if i have already done the command? [06:50] Guest89077: sudo deluser rutri && sudo adduser rutri [06:52] what is room number for? [06:52] Leave it blank... its for people in universities whose offices have room numbers :) [06:52] I bet you can't put Ogham in there, though [06:52] lol ok [06:53] and now how do i add that user to the administrator group? [06:53] As an argument in the adduser call [06:54] so i can use usermod? [06:54] Guest89077: It's silly to create the user and then modify it right away. But if you must, sudo usermod -a -G admin rutri [06:55] Guest89077: Please read the man pages for adduser, deluser and usermod :) [06:55] ok thanks you answered my other question [06:55] yes the command i knew the syntax allready [06:56] i was going to ask what the admin group is called but you answered taht [07:00] ok one last question, how do i get a dameon to automatically start? ie: i dont want to have to do "/etc/init.d/libvirt start" all the time [07:02] Guest89077: sudo apt-get install sysv-rc-conf -y && sudo sysv-rc-conf # is one way. [07:03] thank you ill research into that [07:03] i appreciate the help [07:03] Surely libvirt registers itself by default [07:03] thats what i thought [07:03] i did use the apt-get to install it [07:05] Guest89077: So it should already run at boot time... did you actually test this? [07:06] on second [07:06] one* [07:06] jmarsden: it should also invoke-rc.d itself on install [07:07] Right... the question suggested he wanted to mess with what daemons started automatically... I thought he'd already accidentally prevented it from starting at boot or something like that... [07:07] Anything is possible [07:11] ok i see it running [07:12] "ps -aux | grep libvirt" showed it [07:12] Guest89077: In general, try to avoid asking the channel for info you can determine for yourself using man pages or by trying something out. BTW, /etc/init.d/libvirt-bin status # is the more conventional check on a daemon's status :) [07:13] Guest89077: pgrep [07:14] jmarsden: unfortunately policy doesn't require init.d's to provide a status target :-( [07:14] IME only RH-originating stuff like libvirt provides status [07:14] twb: True, but most do, including libvirt-bin :) [07:15] IIRC even dh_make's example doesn't, despite it being trivial due to s-s-d --make-pidfile [07:16] Hmmm, looks like about 50% on this machine. [07:16] unfortunately I have spent the better part of an hour researching my problem with no luck [07:17] I tell a lie, dh_make does provide status [07:17] Guest89077: Testing whether something starts at boot takes an hour? Booting a modern machine should not take that long! [07:17] jmarsden: but filing the change request and getting it approved takes WEEKS [07:17] :-) [07:18] :) [07:18] well i thought it was because the daemon was not starting but thats not the case. [07:18] Guest89077: Do not guess what your problem is and ask questions based on your guesswork. Instead, ask about the actual problem. [07:19] Guest89077: See http://catb.org/~esr/faqs/smart-questions.html for more on this. [07:19] ok i am trying to access a virtual server remotely using cirt-manager [07:19] jmarsden: in a test here, I get 17 with and 43 without [07:19] Using x=0 y=0; for i in /etc/init.d/*; do if $i status; then ((x++)); else ((y++)); fi; done; echo $x $y [07:20] Some of those are from base-files or whatever it is, though [07:20] sorry virt-manager* [07:21] hi, how to know which package does a binary belongs too ? [07:21] Goddammit I hate it when init scripts don't correctly work when stdio isn't a tty [07:21] This just hangs: x=0 y=0; for i in /etc/init.d/*; do if $i status |& grep -qi usage; then ((x++)); else ((y++)); fi; done; echo $x $y [07:21] maxagaz: dpkg -S /path/to/binary [07:21] twb: I did: grep -c status /etc/init.d/* | grep -c :0$ ; ls -1 /etc/init.d/* |wc -l # to get 70 don't have it out of 156 total. [07:22] the command i use is "virt-manager -c qemu+ssh://192.168.198.1/system" and it asks for a password but then i get a message saying that libvirt-bin is not running, however it is running. [07:23] jmarsden: I was worried about false positives :-) [07:23] jmarsden, thanks [07:23] maxagaz: You're welcome. [07:24] jmarsden: e.g. alsa-utils [07:24] twb: Yes, your way is probably more accurate than mine. [07:24] x=0 y=0; for i in /etc/init.d/*; do test $i = /etc/init.d/rcS && break; echo $i >&2; if $i status |& grep -qi usage; then ((x++)); else ((y++)); fi; done; echo $x $y [07:24] Your way is also about twice as long :) [07:25] ...that one (checks for "Usage:" output) gives 24 with, 15 without [07:25] Oops, s/break/continue/ and I get 36 with, 23 without [07:28] twb: That's more in line with my percentage. [07:28] Guest89077: That sounds strange, but I'm not enough of a virtualization expert to really help... did you ask in #ubuntu-virt ? === root is now known as Guest77812 [07:31] Oops, those numbers are backwards [07:32] I have a Kerberos gss-api query... [07:32] how do I start the gss-server from the samples? [07:34] knock knock... [07:35] Who's there? [07:35] your mom. [07:35] Your mom who? [07:35] [07:36] Kindly leave the stage! [07:36] * twb fills Ziber's pants with whitewash [07:37] need some help with kerberos... [07:37] how do I start the gss-server from the samples? [08:26] morning [08:27] is there a way to set the stride option when installing ubuntu server with a raid 5? [08:30] Dunno. You can with raw mdadm === Hawkeye is now known as drchrishawker [09:41] hello all [09:43] i need some help [09:44] drchrishawker: just ask your question [09:45] I am installing server 9.1 [09:46] i need to set it up to run irc bots (pirc and python). is there a link which gives me instructions to do this. [09:46] also, is it possible to get a desktop on server 9.1? [09:49] drchrishawker: you'll have to google the ircbot howto. I'm sure you're not the first to do it though. If you want a desktop on the server, consider installing the desktop edition and just the ircbot packages? [09:49] is there a difference between the two distro's? [09:52] drchrishawker: yes, but for you the biggest difference will be the packages installed from scratch [09:52] drchrishawker: if you want a desktop environment, starting from the desktop edition is probably easiest for you [09:53] then i can install the packages from scratch? [09:54] And just for correctness sake, there's no such thing as Ubuntu 9.1 - I guess you mean 9.10 [09:56] yes, maxb [09:57] drchrishawker: correct; you only seem to want the irc related packages on your machine which should be a simple 'apt-get install' [09:58] i also want web server and openssh server to remote-access [10:25] I can not get phpmyadmin to work with Ubuntu [10:31] hey! trying to install ubuntu server 9.10 386 from a usb drive using unebootin to extract the iso. when i start the installation the instruction does not seem to apply. [10:31] "On Ubuntu 9.10 server edition the install menu will be shown right after reboot. Chose "Help" and then press F6. At the boot prompt type "install cdrom-detect/try-usb=true" and hit enter." [10:32] Help goes to the usual installation.. and I can't use the shell there for "install" === ogra_ is now known as ogra === ogra_ is now known as ogra [12:14] hi [12:15] which file i need to look for NFS Server for logging in Ubuntu ? [12:15] i have issues in my NFS Server [12:16] kaushal: most log go to /var/log/syslog [12:17] kaushal: either /var/log/*nfs* or /var/log/syslog [12:25] I don't think /var/log/*nfs* is standard [12:44] mornin [13:21] yo [13:28] oy [13:42] Has anyone else had experience installing Ubuntu 9.10 Server onto a setup like this?: /dev/sd[a-c]1 => md0 => lvm, which branches off to these LV's: root and swap === oru_work_ is now known as oru_work === dendro-afk is now known as dendrobates [13:58] I am trying to install an ssl cert and when I go to start apache it tells me that it failed. When I look in the apache access.log it says "Init: Unable to read server certificate from file" What are some common causes for this error? [13:58] it point to the correct location [13:59] Elad: does the www-data user have rights to read the file? [14:00] no, just root [14:00] let me update the permissions [14:02] I chowned the certificates to www-data and I still get the same error [14:03] good morning. [14:03] any other ideas? [14:04] Elad: what are the permissions on the directory above the cert? [14:04] does it need permissions other than these "-rw-r--r--"? [14:05] the permissions I just posted are for the actual cert itself [14:05] folder permissions are "drwxr-xr-x" [14:06] Elad: looks right to me, just to double check the cert file is good? [14:06] sommer: How do I do that? [14:09] Elad: try "openssl verify /path/to/cert" === oru_work is now known as techsupport [14:10] oh, looks like that may be my problem, I have it pointed to a crt, and it is looking for a pem [14:11] is there a way to generate a pem from the crt? I will start googling it, because I'm not sure what the difference is [14:12] Elad: there's a way with the openssl command... don't remember how exactly though, should be able to google it [14:12] I use a crt though [14:12] Elad: Is crt a specific format? [14:13] I thought it was just a generic sort of extension for either a DER or PEM formatted certificate. [14:14] ya, that sounds right === techsupport is now known as boevik === RoyK^ is now known as RoyK [14:18] ah pem is just der with a text header http://gagravarr.org/writing/openssl-certs/general.shtml [14:32] i am having issues with my LDAP user auth on my ubuntu server [14:32] the pam-ldap that is generated by the ubuntu is giving me grief and would not let me log in or change passwords (my own tweaked version of pam confs allows me to log in) [14:32] could somebody please give me a hand? [14:33] omh, I think I just figured out what my problem is - the key that I generated on my server is the private key? The key given to me by verisign is my public key? [14:35] Elad, the key given to you by verisign? usually you'd generate a key and a certificate signing request (CSR), which only gets signed by Verisign... you'll want to use the key you generated at the beginning, along with the signed certificate that verisign sent you [14:37] I am stuck, I can't seem to get past this error when running openssl verify mycert.crt: [14:37] unable to load certificate [14:37] 29250:error:0906D06C:PEM routines:PEM_read_bio:no start line:pem_lib.c:647:Expecting: TRUSTED CERTIFICATE [14:40] is the first line in you .crt file "-----BEGIN CERTIFICATE-----" [14:40] hi all [14:40] hey hippster [14:41] smoser: ping [14:41] I'm trying to find a very quick booting LAMP system, just wondering how ubuntu does? [14:41] hey [14:41] smoser: pm ? [14:42] hippy: you tick LAMP while installing the OS [14:42] and it's there [14:43] I had a space between the ----- Begin . . . . says it is okay now [14:43] thank you for posting that [14:44] thanks but that doesn't really answer my question.. am already up to my eyeballs in isos (debian, centos, slackware, netbsd)... all very slow boot times - am running from qemu [14:44] Elad, np :) [14:49] hippy, it always depends what you mean by quick, and it depends on the system on which it will run as well. [14:51] hippy, for instance, my desktop boots quickly, so do my server VMs... but by quickly I mean it takes around 30-40 seconds to boot perhaps? [14:52] i'm looking more like 6 seconds max.. netbsd came very close but failed in other areas [14:57] not to worry, i knew it would be a long shot finding something super quick... i'll go back to tinkering with qemu's snapshot feature.. thanks anyway :D [14:57] Hi there.. trying to set up 2 ipv6 addresses on the same nic @ boot but unsuccessfully [14:57] anyone know how to? [14:59] now it is telling me that my key doesn't match I have the SSLCertificateFile /site.crt AND my SSLCertificateKeyFile /site.key . . . . Is that how they are suppose to be? [15:00] because I only have 1 set on this machine, and I haven't regenerated them since I originally generated them a couple days ago [15:06] New bug: #505938 in samba (main) "package samba-common 2:3.4.0-3ubuntu5.3 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/505938 [15:29] Elad, they are supposed to match, but you do have the right Apache directives [15:30] cyphermox: what do you mean they are suppose to match? [15:31] the two files, the .key and .crt, are supposed to be for the same "certificate" [15:31] one is the private key, the other is the signed certificate [15:32] gotcha, well, I generated the crt from the key, so I am assuming they match [15:32] so I don't know what the problem is [15:32] Elad, as for the crt file, check that your .key starts with "-----BEGIN RSA PRIVATE KEY-----" [15:33] -----BEGIN RSA PRIVATE KEY----- [15:33] Proc-Type: 4,ENCRYPTED [15:33] DEK-Info: DES-EDE3-CBC,297188752A5F22E6 [15:33] and it really needs to be the same key you used to generate the CSR that you sent verisign [15:33] it is [15:34] ok. then the following commands should give you MD5 sums that should match: [15:34] 'openssl rsa -noout -modulus -in site.key | openssl md5' and 'openssl x509 -noout -modulus -in site.crt | openssl md5' [15:35] if they don't, I'm not sure what you could do past starting over, generating a new key/CSR and getting it signed by Verisign [15:36] no LDAP ppl around to give me a hand? [15:37] cyphermox: after I enter in my passphrase it goes back to the command prompt, doesn't give me any output - is this typical? [15:38] Elad, I don't think so [15:38] are you certain you have the right passphrase? [15:39] yes, looking further up the screen it looks like the command has an error in it [15:39] where? [15:40] openssl:Error: 'md5 and openssl' is an invalid command. [15:40] ah no [15:40] openssl rsa -noout -modulus -in site.key | openssl md5 [15:40] is one command [15:40] it generates a sum for your key file [15:41] to check your crt file, it's this one: openssl x509 -noout -modulus -in site.crt | openssl md5 [15:41] and the output from both of these should match exactly [15:41] ah I see, I did both of the commands as 1 command [15:43] shoot, they don't match [15:43] so I need to send another csr to verisign? === boevik is now known as techsupport [15:44] I think that would be the easiest thing. I believe you can do this without paying again, but I don't usually do the purchases myself :) [15:45] so for future reference I should be able to generate the site.key and site.crt and do the md5 on them, make sure they match, THEN send verisign the csr? === techsupport is now known as boevik [15:49] New bug: #237775 in spamassassin (main) "spamassassin perl conflicts with razor package" [Low,Incomplete] https://launchpad.net/bugs/237775 [15:49] Elad, no [15:50] cyphermox: how come? [15:50] the md5 is to make sure the key and the certificate that verisign sends you back match. I did this because I run scripts that generate apache config, and wanted to make sure the certificates weren't in the middle of being renewed (e.g. the key and cert would not match) before pushing files and breaking things [15:51] cyphermox: the .crt is suppose to be the one that verisign sent me back? [15:51] yes [15:51] ah, let me try that one [15:52] I was wondering when that one would come into play [15:52] what you send them is a different file that gets created at the same time as the key, and is technically a CSR (certificate signing request) [15:56] Elad, all the information you need is here: http://www.verisign.com/support/tlc/csr/modssl/v00.html [15:57] that is what I went through and did, but now I am trying to get it to work with the real cert === dantaliz1ng is now known as dantalizing [16:38] New bug: #312493 in php5 (main) "Not possible to run PHP in a multiuser and secure way" [Wishlist,Triaged] https://launchpad.net/bugs/312493 [16:38] Hi, has anyone started or plans to test ivoks new packages for the cluster stack in lucid? (see announce on the mailling list) [16:40] mathiaz, hi. I'm curious your setup for the ISO testing with preseeding and scripts... is that in a bzr branch somewhere? [16:41] cyphermox: https://code.launchpad.net/~mathiaz/+junk/iso_testing_scripts [16:45] thanks [16:46] cyphermox: do I need to generate an intermediate.crt for the SSLCACertificateFile? or can i just reuse the public.crt from the SSLCertificateFile? [16:50] Elad, no, that's for something different [16:50] SSLCACertificate is a certificate you can download from Verisign (check their support pages) which helps identifying them as a valid signer for your certificate [17:05] What's the prefered way to get something like lvm on a shared iscsi disk in ubuntu? [17:05] Is it just to use clustered lvm? [17:05] Are there any alternatives that don't require redhat-cluster suite? [17:06] ivoks is our expert on that. He's not around at the moment. [17:07] ah, ok, what is the prefered mailing list to ask such a question on then? Is it the server team list, or is there another? [17:08] Server team list is good. [17:08] ok, thanks! [17:22] is there a tui for setting up a network interface in ubuntu server a la system-config-network in rhel? [17:27] hello, i am trying to set bind up but i'm having some difficulties [17:31] my zone file is here: http://pastebin.com/d537d11f1 [17:40] i have a static route setup so that 192.168.0.x traffic is routed to eth1. I have a web server on 192.168.0.4, and can ping it. but firefox will not connect to it (i know the web server works). How do i fix this? [17:41] I am using firefox on the same system on which i have the static route setup on [17:42] mrmojo: what's the difficulty? [17:43] marks256: you can ping 0.4 from the same computer on which firefox is running? [17:44] cemc, yes. i actually just sshed into 0.4 [17:44] marks256: what does firefox say ? [17:44] cemc, unable to connect. i don't think firefox is using the route [17:46] here is the output of route: 192.168.0.0 * 255.255.255.0 U 0 0 0 eth1 [17:47] what ip does the firefox machine have? [17:47] eth0 or eth1? [17:48] eth1 is 192.168.0.1 [17:49] marks256: can you pastebin output of 'ip r' command ? [17:49] cemc, sure [17:50] cemc: i've requested your IM friendship on gtalk and icq :) [17:50] cemc, http://pastebin.com/d3538eff0 [17:50] ivoks: I don't see anything here :) [17:50] heh [17:50] gaim was much better :/ [17:51] pidgin works too [17:51] or that :D [17:51] ;) [17:53] marks256: is the webserver set up correctly? [17:53] cemc, yes [17:53] cemc, i was using it the other day on a different network [17:54] that doesn't necessarily mean it will work here ;) [17:54] marks256: try this: "telnet 192.168.0.4 80" [17:54] from 0.1 [17:54] cemc, i don't see why it wouldn't. the networks were setup the same [17:55] cemc, connection refused, but it found it. [17:55] marks256: connection refused could mean: 1. there's some firewall on the webserver, 2. webserver not listening on 0.4 / not running at all [17:56] marks256: log on to 0.4 and try: "netstat -nlp | grep 80" [17:56] cemc, there is no firewall that i know of. i will check and make sure apache is running [17:56] cemc: how about now? :D [17:57] ivoks: nothing. lemme relog, pidgin is running for 34days (stress testing it :) ) [17:57] hehe [17:57] ivoks: try again, do you see me online ? [17:57] nope [17:57] @gmail.com is domain, right? [17:57] cemc, no output for "netstat -nlp | grep 80" [17:58] ivoks: yes, that should be it. and I'm logged on and talking to other ppl [17:58] ivoks: try again, relogged there too [17:59] marks256: it's not listening on port 80, or it's not running, do a "ps ax |grep apache" [17:59] cemc, 1512 pts/0 S+ 0:00 grep --color=auto apache [17:59] marks256: apache not running, it's an ubuntu system too? [18:00] cemc, omg. i had to enter my ssl passphrase before apache would start :| [18:00] cemc, it works now [18:02] marks256: note ze commandz: netstat, ps, telnet ip 80 :) [18:02] cemc, noted ;) what exactly is netstat? i've used all the others before [18:04] cemc, thanks for all the help. lunch time now :) [18:04] marks256: man is your friend, always [18:04] man netstat [18:04] first line :) [18:04] cemc, i'll check out the man page. thanks :) [18:06] magnetic__: /usr/sbin/openssl rsa -in server.key.old -out server.key [18:06] eh [18:07] When booting ubuntu server, I get a blank screen unless I use the recovery option. Once booted, modprobe radeonfb gives me a nice high res console. How can I fix that? [18:08] add radeonfb to /etc/modules [18:08] and run update-initramfs -u [18:15] ivoks: thanks! perfect [18:16] np [18:16] Is that a normal problem with radeon based machines in text mode? [18:16] Perhaps a consequence of the 9.10 upgrade [18:16] I didn't really get useful hits on google [18:17] i don't know [18:17] most of my servers have output redirected to serial console and IPMI [18:17] :) [18:18] :) heh [18:18] Using ubuntu server on old laptops is probably not a well tested setup :) [18:18] :) [18:18] Built-in UPS though [18:19] Does nicely as a home mail server [18:22] well, take care people [18:22] see you tomorrow [18:25] VSpike: lower power too :) [18:29] ivoks: Hi, I've started testing on your new rhcs packages [18:30] the_real_dave: very true! [18:30] VSpike: probably the only good use for a laptop with a wrecked screen :) Make it serve you lol [18:32] the_real_dave: heh. This one has a wrecked battery and keyboard. Same result tho [18:36] VSpike_: can I ask why you need to use your own mail server? Do you need your own domain for that? [18:38] the_real_dave: There are other ways to solve the same problem, so it was partly just educational, but I wanted to store all my mail in an IMAP server [18:38] Could have used gmail I guess [18:38] But having it local gives better speed [18:38] So I use fetchmail, postfix, dovecot, maildrop [18:39] You don't need your own domain for that if you don't want one [18:39] VSpike_: ah ok. Do you backup nightly or similar? I'd hate to lose all my email to a bad HDD or the like :S [18:39] Yes, I backup the maildir's to another machine daily [18:40] I like that I can do all the filtering server side, and then use any mail client from any machine [18:40] Handy when workign across lots of machines and trying mail clients to find the one I like best [18:55] actually, what do people use for monitoring bandwidth usage? [18:56] I tried iptraf and bandwidthd, the latter of which creates a zombie process :( Any suggestions? All my server does is file sharing through NFS, Squid and Apt-Cacher proxies. I'd like to be able to see who on the LAN is connected, and what they're connected to :) [19:07] what do people use for monitoring bandwidth usage? I tried iptraf and bandwidthd, the latter of which creates a zombie process :( Any suggestions? All my server does is file sharing through NFS, Squid and Apt-Cacher proxies. I'd like to be able to see who on the LAN is connected, and what they're connected to :) [19:33] the_real_dave, I use vnstat and vnstat-php frontend [19:33] mathiaz: Why would you want to demote radvd? [19:33] Jeeves_: heh, i just responded to that :) [19:33] but that's more like a counter than active monitoring [19:33] Jeeves_: it was discussed during UDS - I'm just asking for feedback [19:33] qman__: a counter would be nice too :) I'll check it out, thank you :) [19:35] mathiaz: Ok, but what were the reasons to even think of demoting it? [19:35] the_real_dave: i've used ntop and cacti [19:35] the_real_dave: dunno if you saw this yet http://www.ubuntugeek.com/bandwidth-monitoring-tools-for-linux.html [19:36] Jeeves_: less maintainance for a LTS release [19:36] Jeeves_: IMO we should keep it [19:36] i dont think radvd would require much maintenance [19:36] it usually Just Works [19:36] Jeeves_: Others suggested otherwise [19:37] Sam-I-Am: Well, it seems to be rather poort code [19:37] Jeeves_: If you have an opinion, I'd suggest to reply to the thread [19:37] Jeeves_: in that case, a lot of stuff needs demotion :) [19:37] mathiaz: Been there, done that :) [19:37] Jeeves_: great! thanks [19:38] Sam-I-Am: Try running radvd on a link with another mtu than 1500, that seems to break stuff [19:39] heh, names on postings != irc nicks [19:39] "who?" [19:40] Jeeves_: i.. think its primarily designed for ethernet. wonder if jumbo frame MTUs confuse it too. [19:40] i remember running it on gig links before, but that was the openbsd variant... rtadvd. [19:40] the T means its uh... better? heh [19:43] * Jeeves_ <- Mark Schouten [19:43] ahh [19:43] <- matt [19:43] Sam-I-Am: Ethernet runs on > 1500 bytes too :) [19:43] also just saw another post to the ipsec thread... i kinda wonder when we might see l2tpv3 on linux [19:43] since it seems to be gaining a lot of steam [19:44] Jeeves_: thats why i said jumbo frames :) [19:45] mathiaz: btw, sorry i kinda took a back seat for a while with server stuff (like openldap) ... lots of employment issues in the last few months which will hopefully be solved soon. [19:45] life's been a little rough [19:47] New bug: #471461 in clamav (main) "package clamav-milter 0.95.2+dfsg-4ubuntu4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,Incomplete] https://launchpad.net/bugs/471461 [19:47] New bug: #482984 in clamav (main) "clamav crash on startup" [Undecided,Incomplete] https://launchpad.net/bugs/482984 [19:47] New bug: #483412 in clamav (main) "package clamav-base 0.95.3+dfsg-1ubuntu0.09.10 failed to install/upgrade: subprocess installed post-installation script returned error exit status 128" [Undecided,Incomplete] https://launchpad.net/bugs/483412 [19:47] New bug: #483731 in clamav (main) "clamd crashed with SIGSEGV in mpool_free()" [Undecided,Incomplete] https://launchpad.net/bugs/483731 [19:47] oh dear [19:54] New bug: #362890 in cyrus-sasl2 (main) "pidgin crashed with SIGSEGV in _sasl_MD5Update()" [Low,Won't fix] https://launchpad.net/bugs/362890 === dendrobates is now known as dendro-afk [20:07] New bug: #496691 in mysql-dfsg-5.0 (universe) "mysql server and client crash during installation (sudo apt-get install mysql-server mysql-client)" [Undecided,Invalid] https://launchpad.net/bugs/496691 [20:07] clam-av has a CLI interface right? So you could use it to scan for Windows viruses on Ubuntu Server? [20:08] the_real_dave: Yes. You just have to be able to mount the partition. [20:08] * ScottK routinely boots Kubuntu live CDs to scan windows boxes. [20:09] ScottK: how effective is it against Avast! for Linux do you know? [20:09] the_real_dave: I don't. [20:09] Seems to work reasonably well as long as you stay current. [20:09] ScottK: ok, thank you :) [20:09] In Ubuntu we put a lot of effort into keeping the clamav version current in all releases as a result. === dendro-afk is now known as dendrobates [21:03] what's the reason for this error: [21:03] nfs: server 10.1.0.2 not responding, still trying [21:04] the nfs server is online [21:14] Aison: random nfs blurbs :) [21:15] Aison: sometimes its just too sensitive [21:15] Aison: unless you actually have network problems or congestion [21:15] or the nfs server is busy.. the linux nfs stuff needs a little work in the multithreaded area [21:21] yeah, i'm using nfs for a longer time now, and it allways sucked somewhere [21:21] hello jeiworth [21:25] Aison: heh, it hasnt improved too much, although you might have some better experience with nfs4 over tcp [21:53] kirkland: o/ [21:53] mathiaz: \o [21:53] kirkland: it seems that NODES is no longer defined in eucalyptus-local.conf [21:53] mathiaz: hrm... must have missed that as part of my changes [21:53] kirkland: I'm running 1.6.2~bzr1120-0ubuntu7 [21:54] mathiaz: fresh install, or upgrade? [21:54] kirkland: in a CLC+Walrus/CC+SC/3xNCs configuration [21:54] kirkland: an fresh install [21:54] mathiaz: okay, let me look ... [21:54] kirkland: well - the eucalyptus package have been *purged* before [21:55] kirkland: which shouldn't play any role [21:55] mathiaz: can you pastebin your eucalyptus-local.conf? [21:56] kirkland: http://paste.ubuntu.com/355194/ [21:56] kirkland: this is *after* euca_conf --register-nodes has been called [21:56] kirkland: by the auto-registration script [21:56] mathiaz: okay ... can you try rerunning that with sh -x ? [21:56] mathiaz: and pastebin the sh -x output? [21:57] kirkland: hm - first I need to figure what's the actual command [21:58] mathiaz: and silly question, but they didn't happen to leak back into /etc/eucalyptus/euclayptus.conf ? [21:58] mathiaz: /usr/sbin/* [21:58] kirkland: the auto-registration log doesn't give the *actual* command [21:58] kirkland: I meant the complete command, with the arguments [21:58] mathiaz: hrm, seems like a good thing to add to the log ;-) [21:58] mathiaz: given all the other useless crud eucalyptus logs :-) [21:58] kirkland: grep -Ri nodes /etc/eucalyptus/ doesn't return anything [21:59] mathiaz: where is this autoregistration code? [22:00] mathiaz: ie, what files/pkg? [22:00] kirkland: IIUC uec_component_listener [22:00] kirkland: ^^ that's the process listening [22:00] kirkland: I'm trying to find the source code of it [22:00] mathiaz: is that in the eucalyptus source code? [22:00] kirkland: ./debian/registration/uec_component_listener.c [22:01] kirkland: IIRC this is is written by ttx [22:01] kirkland: (or may be cjwatson) [22:01] mathiaz: looking ... [22:01] kirkland: as part of the auto-registration spec [22:02] mathiaz: hmm, that file doesn't do anything with euca_conf or eucalyptus.conf [22:02] kirkland: right - it calls node [22:02] kirkland: or any other relevant script [22:02] kirkland: ./debian/registration/node [22:02] kirkland: that one calls euca_conf [22:04] kirkland: http://paste.ubuntu.com/355201/ [22:04] kirkland: ^^ this is the end of sh -x euca_conf .... [22:05] mathiaz: okay ... [22:05] kirkland: I think it's because NODES= is not defined in the default -local.conf template [22:05] kirkland: line 16 won't match [22:06] mathiaz: yeah, that looks like it [22:06] mathiaz: can you throw in NODES= and retry? [22:07] mathiaz: i'll fix it in euca_conf itself [22:07] mathiaz: with a failover, if NODES= is not found [22:15] kirkland: hm - well - things are not working here [22:15] kirkland: however it doesn't seem related to NODES= anymore [22:15] mathiaz: hmm, okay [22:15] kirkland: the CLC cannot talk to the CC apparently [22:17] hi, i have this on ubuntu server---> http://pastebin.com/m238cb5cc i would like to remove it and i got this error---> http://pastebin.com/m3f430a83 [22:17] anyone have idea on it..? why i cant remove it [22:25] ruben23: It says it's open. [22:35] soren: how do i check for remaining free size on my volume group [22:39] vgdisplay [22:41] soren:---> i got this----->http://pastebin.com/m7bd09b75 how do i check my free space [22:42] Free PE / Size 1098 / 34.31 GB [22:42] It's right there, dude. [22:42] Line 20. [22:42] ok so i can create more lvm... [23:06] hello - trying to find out why sysklogd doesnt use logrotate - anyone got a clue? [23:07] with sysklogd there seem to be a cronfile in /etc/cron.weekly to rotate syslog, /var/log/messages and so on... Other syslog servers like rsyslog are rotated by logrotate... In fact if i install rsyslog, I end up hving both the logrotate rotation, and the sysklogd weekly cronjob setup [23:07] just the cronjob will fail as sysklogd is no longer installed [23:24] oh my bad still running 9.04 apparently there was a move to rsyslogd and logrotate in 9.10.. I guess that answers my question [23:29] Folks: https://help.ubuntu.com/community/aufsRootFileSystemOnUsbFlash used to say that AUFS was no longer working in Ubuntu since 9.10, but now it says that maybe it is and the author of that page will investigate whether it is likely to work with 10.04 LTS. [23:30] Any insight from you guys? I really need a union filesystem for having a boot drive that doesn't actually get written to. Because our boot drives are USB flash drives and they seem to have a half-life measured in months. [23:30] So we spend all our time building new ones and replacing broken ones. === robbiew is now known as robbiew_