[00:17] <cclausen> wtfplymouth: I had some trouble with fstab when upgrading from dapper to hardy.  can you be more specific with your problems?  you haven't provided enough info for a response
[00:18] <xperia> hello to all. can somebody tell me how can i clean up my swap space ? i have php scripts that need a lot of memory and becouse of that the swap space is used a lot. the strange thing however is that my swap space is not cleaned up after the sripts finish. what can i do ? after a while i run out of swap space and my server freeze !
[00:19] <domas> hmmmm
[00:19] <domas> first of all, it is usually other processes sitting in swap
[00:19] <domas> so you have to get them to try to run and touch those pages
[00:19] <domas> setting swappiness to 0 might help too
[00:19] <domas> :)
[00:22] <xperia> domas: thanks :-) i have only 128 MB Ram and my scripts use both 160 MB. becouse of that the swap is used a lot but from my understanding the swap space should be given free after the script finish and exactly this is not the case
[00:24] <wtfplymouth> cclausen: i know that my fstab has a problem, but 9.10 booted with a not available disk and 10.04 stops the boot
[00:25] <wtfplymouth> i just want 10.04 to not stop the boot at an mounting error
[00:25] <cclausen> wtfplymouth: you should be able append a root= line from grub
[00:25] <cclausen> did you read the release notes?
[00:25] <cclausen> there is something about specific hardware and a 1 MB disk alignment or something
[00:26] <cclausen> not sure if that is related, but it might be worth a try
[00:34] <jaypur> just installed ubuntu server 10.4, and i cant install openssh or openssh-client, or openssh-server
[00:34] <jaypur> what's going on???
[00:35] <cclausen> is there a network connection at all?
[00:35] <cclausen> what errors do you receive?
[00:35] <jaypur> cclausen, talking to me?
[00:35] <cclausen> yes
[00:36] <jaypur> i've installed apache... installing mysql now...
[00:36] <jaypur> and openssh-server is not working...
[00:36] <jaypur> not installing
[00:36] <cclausen> what command are you running?
[00:36] <cclausen> and you say mysql is installing now.  are you trying to run multiple installs at once?
[00:36] <jaypur> sudo apt-get install openssh-server
[00:36] <jaypur> no...
[00:37] <cclausen> what error does apt-get give you?
[00:37] <jaypur> let me finish installing mysql i'll tell you whats going on
[00:37] <jaypur> thanks for the help
[00:37] <jaypur> just a sec
[00:41] <jaypur> cclausen, it says that it has a newer version
[00:41] <cclausen> apt-get update; apt-get install --reinstall openssh-server
[00:41] <cclausen> what EXACT error message?
[00:42] <jaypur> cclausen, installing now with the --reinstall
[00:42] <jaypur> :)
[00:43] <cclausen> well then, no problem
[00:43] <cclausen> you probably want to apt-get upgrade too and see if anything else has newer versions
[00:45] <jaypur> what's the difference between upgrade and update?
[00:47] <cclausen> update pulls in the list of newer packages
[00:47] <cclausen> upgrade actually installs them
[00:47] <cclausen> I think peopel are encouraged to use aptitude instead of apt-get these days
[00:48] <cclausen> but I am not sure
[00:48] <cclausen> aptitude safe-upgrade is probably what you want for patching most of the time, I think (feel free to correct me)
[00:48] <handheldCar> you can add the -u option to list what upgrades you have.
[00:48] <whaley> cclausen: that's the one
[00:48] <jaypur> cclausen, got it
[00:49] <cclausen> speaking of that, I've been trying to use the automatic upgrades
[00:49] <whaley> and I use aptitude now instead of apt-get... just out of laziness.  I'd rather have one command instead of two (apt-get and apt-cache) for package management
[00:49] <cclausen> but they don't see to automatically install
[00:49] <cclausen> were the betas just not auto-updating b/c it was beta?
[00:50] <handheldCar> whaley: that sounds nice. I've been tired of two commands.
[00:50] <handheldCar> cclausen: you mean unattended-upgrades?
[00:50] <cclausen> I think so
[00:51] <cclausen> whatever the automatica updates from teh ubuntu server installer does
[00:51] <cclausen> I was using cron-apt before
[00:51] <handheldCar> There's a whole ubuntu.org page I haven't read on that.
[00:51] <cclausen> yeah, I think I read that and set the various config files to actually update
[00:51] <cclausen> but whenever Ilog I see that there are updates waiting for me
[00:51] <cclausen> I guess I'll what happens how that things are released
[01:00] <domas> meh, why oh why is linux swapping, when it has few gigs of memory free
[01:00] <domas> :)
[01:00] <cclausen> if you don't think you need it, just swap -a off
[01:00] <domas> swappiness is at 0 already
[01:00] <cclausen> err, swapoff -a
[01:01] <domas> well, kswapd gets stressed when only few gigs of memory available are left
[01:01] <domas> it was spinning at 100% on one core
[01:05] <domas> like here, http://p.defau.lt/?__SOOu0go_iFe9cWMCF29A
[01:06] <domas> swappiness at 0, 1G of cache, 200M of free memory
[01:06] <sbeattie> wtfplymouth: the other option is to add 'nobootwait' as a boot option (see e.g. bug 510415).
[01:09] <owen1> when will the docs for 10.4 server be online? i can't see it here: http://www.ubuntu.com/products/whatisubuntu/serveredition/techspecs/whatsnew
[01:10] <jaypur>  /etc/apache2/sites-enabled/000-default is empty what should i do?
[01:10] <cclausen> that should be a symlink to sites-available
[01:11] <cclausen> or are both empty ?
[01:13] <jaypur> cclausen, i have to edit 000-default, to change the port... and i don't know why is empty
[01:14] <cclausen> jaypur: is this a new install?  You can remove the files and apt-get install --reinstall the various apache2 packages
[01:14] <cclausen> should setup a new default for you
[01:16] <jaypur> hmm i'll do this
[01:16] <owen1> to updgrade my server, should i change karmic to lucid in sources and run dist-upgrade?
[01:16] <jaypur> just installed
[01:16] <cclausen> owen1: http://www.ubuntu.com/getubuntu/upgrading
[01:17] <owen1> cclausen: thanks
[01:17] <owen1> also, are there docs about 10.4?
[01:17] <jaypur> 000-default still empty
[01:18] <jaypur> sudo nano /etc/apache2/sites-enable/000-enable
[01:18] <cclausen> owen1: what specifically did you need docs on?
[01:18] <cclausen> most of the on-system docs should be up to date for the packages installed, look in /usr/share/doc
[01:18] <cclausen> as well as man pages
[01:19] <cclausen> jaypur: I don't know what to tell you.  are you sure you reinstalled the package that had the file you wanted in it?
[01:19] <zorzar> hey i just set up a mailserver on a ubuntu box, i have postfix and dovecot with postgrey configured, what is the best way to include spamasassin? smapd oder amavis?
[01:19] <cclausen> jaypur: apache2.2-common package was reinstalled?
[01:20] <jaypur> i'm rebooting the pc just a sec
[01:20] <owen1> cclausen: i thought there is some kind of docs about running/maintaining/backups/system admin for ubuntu server.
[01:21] <cclausen> jaypur: that nano command was wrong FYI.  its not 000-enable (at least not to my knowledge)
[01:21] <cclausen> owen1: those types of docs won't be too mcuh of a change from current versions.  I'd just read what is out there now
[01:21] <jaypur> cclausen, what should be the command
[01:21] <jaypur> ?
[01:22] <cclausen> jaypur: well, first, try ls -l /etc/apache2/sites-available
[01:22] <cclausen> what files are there?  a default file?
[01:23] <cclausen> owen1: I use a daily rsync snapshot hardlink script for backups myself to conserve disk space on unchanged files
[01:23] <jaypur> default, and default-ssl
[01:24] <owen1> cclausen: thanks for this insight. i found this page, i believe that is what you are refering to - https://help.ubuntu.com/community/Servers
[01:24] <jaypur> got it
[01:24] <jaypur> !!!
[01:24] <jaypur> now it's on!
[01:24] <cclausen> jaypur: sudo nano /etc/apache2/sites-available/default does that work?  or is it empty too?
[01:24] <jaypur> :)
[01:24] <jaypur> now it's on
[01:24] <cclausen> jaypur: I didn't think there was a problem.  good
[01:25] <cclausen> owen1: basically, that, yes
[01:25] <owen1> cclausen: i use rsync as well. what does it mean hardlink?
[01:25] <owen1> i know what symlink is.
[01:25] <jaypur> cclausen, now i'll check the port
[01:25] <cclausen> owen1: a hardlink is a directory entry that points to a file already on disk
[01:25] <jaypur> http://tuxserver.no-ip.org/
[01:25] <jaypur> :)
[01:26] <cclausen> owen1: read the docs on the --link-dest option to rsync
[01:26] <owen1> cclausen: great. thanks
[01:27] <cclausen> owen1: there is an option to cp to copy using hardlinks as well.  I sued this to create the initial directory to link-dest against.
[01:27] <cclausen> owen1: and of course remember that you really want an off-site backup too, not just a local version
[01:28] <owen1> cclausen: right now i rsync to external usb drive
[01:29] <cclausen> owen1: ah, ok, that works too.  I have actual TSM backups for production servers do just do the rsync for my users to get data without needing to ask me to pull it down from TSM.
[01:30] <cclausen> zorzar: I'd look for a guide online and try following the instructions.  I haven't used postfix, but I had spamassassin working with exim4 on Debian sarge back in the day
[01:31] <owen1> cclausen: what made switch from debian server to ubuntu?
[01:31] <owen1> i am also curious about the differences between the two.
[01:32] <cclausen> owen1: ubuntu has a stable release cycle and 5 year support on LTS releases
[01:33] <cclausen> owen1: debian thinks "when it is ready" is a fine release schedule
[01:34] <cclausen> debian is the base for ubuntu and many packages come directly from there, but debian doesn't have LTS releases and I'm pretrt lazy and don't like to upgrade that often.  I still have some dapper servers running too.
[01:34] <jaypur> cclausen, i have this error for ssh http://ubuntuforums.org/showthread.php?t=33236
[01:34] <jaypur> i don't know why i'm getting it... didn't do anything that never done b4
[01:34] <cclausen> jaypur: did you just reinstall?  that error just means that youe ssh key has changed
[01:35] <cclausen> it should tell you the line number in your known hosts file
[01:35] <jaypur> cclausen, i got this error, removed everything and reinstalled and it's still
[01:35] <cclausen> did you remove your known_hosts file?
[01:35] <jaypur> yes but i can't find this file
[01:35] <zorzar> cclausen: there are guides for everything out there, i wonder what the best solution is
[01:35] <cclausen> should be ~/.ssh/known_hosts
[01:36] <owen1> cclausen: are u working for someone or independent? also, do u pay canonical for support?
[01:36] <cclausen> zorzar: there probably is no "best" solution.  There are just "works for me" solutions
[01:36] <cclausen> owen1: I do not pay.  I fix problems myself
[01:36] <cclausen> had to manually fix a bad nagios2 package... wasn;t sending emails out due to a path being wrong in teh default config
[01:36] <owen1> cclausen: so why do u care about the LTS? isn't it for companies that need tech-support?
[01:37] <jaypur> cclausen, it's empty...
[01:37] <jaypur> grrr
[01:37] <jaypur> arrgh
[01:37] <cclausen> owen1: our campus policy says I must have "supported" OSes on the network
[01:37] <cclausen> jaypur: ren your .ssh directory and try again
[01:37] <cclausen> (obviously that could be bad if you use SSH keys)
[01:38] <owen1> cclausen: so u'r the sysadmin for the campus. cool.  when r u planning to uprade to 10.4?
[01:38] <cclausen> owen1: and yes, we might need to call for support one day.  good to know that I can if needed.  Much better than paying redhat upfront $50 per server.
[01:40] <owen1> cclausen: great. so ubuntu is on a different route than red-hat, in that aspect.
[01:40] <cclausen> owen1: in my opinion, yes
[01:40] <owen1> cclausen: i just read that canonical are experimenting with admin certificates.
[01:41] <cclausen> owen1: I also REALLY hate RPMs.  apt-get is much better about dependancies and there are more packages available in the repos
[01:41] <jaypur> cclausen, why is this happening to me... never happened this before... how can i remove all things from ssh and reinstall it?
[01:41] <cclausen> jaypur: its a client issue.  has nothing to do with the ssh packages themselves
[01:41] <cclausen> jaypur: one whatever ssh client you are using, change the known_hosts file and remove the offending key
[01:42] <Jota> Hello all!!   I've tried to upgrade my 8.04 server to 10.4, but it no sucess:   sudo do-release-upgradeChecking for a new ubuntu releaseNo new release found
[01:43] <Jota> What is wrong??
[01:43] <jaypur> cclausen, sorry if i'm a little bit dumb today, i'm having some family issues right here... so can you tell me please what should i do right now?
[01:44] <zorzar> cclausen: ok thanks
[01:45] <owen1> cclausen: how many machines do u manage and what is the strategy for installing/updating them? do u use some kind of a centeralized offline repository?
[01:47] <cclausen> owen1: I just have about 13 ubuntu servers, most of them are old laptops in closets to act as DHCP servers
[01:47] <cclausen> I do not currently have an automated deployment system
[01:47] <cclausen> but I think I'll setup fai now that 10.04 is released
[01:47] <jaypur> cclausen, i does not have the file .ssh
[01:48] <jaypur> at ~/.ssh
[01:48] <cclausen> jaypur: did you cd ~/; mv .ssh dot_ssh ?
[01:48] <jaypur> at my home,... .ssh does not exist
[01:48] <cclausen> jaypur: its hidden by default you are using ls -lA to look for it?
[01:48] <jaypur> ls -lA
[01:49] <cclausen> and you are doing this on your ssh client system were you are getting the error?  not the ssh server?
[01:49] <jaypur> does not exist
[01:49] <jaypur> i'm doing it directly to my server
[01:49] <jaypur> physical
[01:50] <cclausen> jaypur: pastebin your exact error message somewhere
[01:51] <jaypur> i type ls -lA .ssh
[01:51] <jaypur> is it correct?
[01:51] <jaypur> sorry
[01:51] <jaypur> =/
[01:52] <cclausen> jaypur: please goto http://paste.ubuntu.com/  and paste the full error message
[01:52] <jaypur> cclausen, i'm in my laptop, the error is in the pc.... the error is impossible  to access or does not extist
[01:52] <Jota> Hello all!!
[01:52] <Jota> I've tried to upgrade my 8.04 server to 10.4, but have no sucess:
[01:52] <Jota> "sudo do-release-upgradeChecking for a new ubuntu releaseNo new release found"What is wrong??
[01:53] <owen1> cclausen: can't find a lot of info about FAI but i'll look into it.
[01:53] <Jota> The 10.4 LTS server is not available??
[01:54] <Jota> Anyone has already done the upgrade from 8.04 to 10.4 ???
[01:54]  * cclausen attempt to upgrade from 8.04 to 10.04 right now on a web server
[01:55] <jaypur> cclausen, lol cd /home/user/.ssh does not exist lol
[01:55] <cclausen> oh, jota. I just had your problem
[01:56] <jaypur> ls -a is showing me
[01:56] <cclausen> Jota: you need to read more carefully.  need to append --devel-release to that command
[01:56] <cclausen> jaypur: paste your error message as I requested before.  I can't help you without seeing the real error message
[01:56] <cclausen> Jota: ERR, SORRY
[01:56] <jaypur> cclausen, the error message is: does now exist
[01:56] <jaypur> not*
[01:56] <cclausen> Jota: its --devel-release
[01:57] <cclausen> jaypur: try the ssh command again
[01:57] <Jota> ok...  but if I append -d, will I get the final release, or the RC (or beta) release???
[01:58] <jaypur> aaaaah
[01:58] <cclausen> Jota: the beta becomes the final release so you get both
[01:58] <cclausen> Jota: I thinl everything is release right now though
[01:58] <jaypur> why did i need to reinstall it... the last version of ubuntu all things worked right...
[01:58] <cclausen> gah!  got not enough free disk space error...
[01:58] <cclausen> this is not good...
[02:00] <Jota> cclausen: are you sure about that?? I will get the final 10.4 release, as relesed yesterday??
[02:00] <jaypur> cclausen, the file just does not exist, ~/.ssh
[02:01] <Jota> or should I wait a few more days to do the upgrade??
[02:02] <cclausen> Jota: you get whatever is in the repo now.  should be release by now
[02:02] <cclausen> jaypur: ignore the file does not exist message. re-try the ssh command.  what error does ssh give you?
[02:02] <jaypur> cclausen, what ssh command???
[02:04] <cclausen> jaypur: the one you originally asked about
[02:04] <cclausen> jaypur: or am I really confused?
[02:05] <jaypur> cclausen, i'm tired and having some personal problems here, i'm really sorry to bother you
[02:05] <jaypur> cclausen, i think i'll reinstall the ubuntu server
[02:05] <jetole> Hey guys. I'm new to software raid. I just installed a new hard drive on a server and did a ubuntu install I told it to use raid1 for 2 disks but only one disk was connected. I want ro connect my old disk, boot the server and copy data from the old drive to the new one before I make the old one part of the raid. If I re connect the old drive and boot off the new one will it auto attach that disk to raid? I don't want to lose the data of the old disk ...
[02:05] <jetole> ... before I copy it
[02:05] <jaypur> cclausen, it's easier for me i think and i won't bother you... and at the instalation... i'll select to install the open ssh....
[02:06] <Jota> ok...
[02:06] <Jota> and, another question: has anyone done the "do-release-upgrade" via ssh??
[02:08] <cclausen> Jota: I did that before from dapper to hardy and it mostly worked...  I wouldn't do this from ssh on anything important
[02:10] <cclausen> jetole: I think you want to make an offline backup before you do anything else
[02:10] <cclausen> jetole: in general, all RAID solutions I've seen want to try and mirror the data immediately, clobering whatever was there previously
[02:13] <jaypur> cclausen, I'll reinstall everything thanks for your help man
[02:14] <jetole> cclausen: that sucks but ok
[02:14] <jetole> cclausen: thats also piss poor programming since I didn't tell it which disk to use as the second one but I'm not hear to argue it. I just wanted to know
[02:15] <cclausen> jetole: it might work fine, but I think you are better safe to make a backup before attempting anything
[02:15] <cclausen> jetole: always err on the side of caution
[02:15] <jetole> I don't have that option since it's a home server and not one at the office or data center but I will try booting off a live CD
[02:16] <cclausen> jetole: you don't have backups ?  that is problem no matter where the server is at
[02:16] <cclausen> and RAID is NOT a backup mechanism
[02:16] <jetole> all our servers at work have 3ware cards except for some old Dell servers that have a dell rebranding of an adaptec I think
[02:16] <cclausen> the Dells I use have PERC cards which are re-branded LSI controllers.  I have a custom built server with a nice areca controller too.
[02:16] <jetole> cclausen: I know but I'm not rish
[02:17] <jetole> these are PERC too but I thought it was adaptec
[02:17] <jetole> poweredge 2950 III
[02:17] <cclausen> 2950 II uses PERC6
[02:17] <jetole> I think perc 5i but not sure
[02:17] <jetole> oh
[02:17] <cclausen> oh, you got the perc5?
[02:17] <jetole> not if you have 2950 II with perc6 I don't
[02:17] <cclausen> I have two 2950s with PERC5, but they are original, not the III version
[02:18] <cclausen> our 2900 IIIs have PERC6
[02:18] <jetole> then yeah I really doubt I have perc5 if you have perc6 in II
[02:18] <cclausen> but I don;t have linux installed on any of the good server hardware, just in VMs and on old laptops
[02:18] <cclausen> lspci would tell you for sure though
[02:19] <jetole> cclausen: linux is our VM. Used to be ESXi 3.5 but I was disapointed in a lot of features and we migrated a year or so ago
[02:19] <jetole> maybe 18 months
[02:19]  * jetole checks
[02:19] <cclausen> I use Hyper-V here.  don't like VMWare at all.
[02:19] <aetaric> jetole: what features was it missing?
[02:20] <jetole> 01:00.0 RAID bus controller: LSI Logic / Symbios Logic MegaRAID SAS 1078 (rev 04)
[02:20] <cclausen> yep, LSI controller
[02:20] <jetole> aetaric: I don't remember now
[02:20] <aetaric> i've yet to find anything that esx(i) can't do.
[02:21] <jetole> yeah I remember back when I made the change I had that argument with a bunch of guys in the vmware room
[02:21] <cclausen> aetaric: Hyper-V has maximum CPU limits per VM as well as priority settings.  I didn't see that in VMWare
[02:21] <jetole> I don't want to have it again
[02:21] <jetole> glad it works for you
[02:21] <aetaric> cclausen: it is in there. it is a more advanced feature.
[02:22] <cclausen> ah, ok.  I figured they would add it somewhere.  I haven't looked at VMWare stuff since 2007
[02:22] <aetaric> i just use it for vmotion....so...
[02:22] <jetole> cclausen: well I know since I left they came out with vmware ESXi 4
[02:22] <cclausen> and I still don't like the crazy network adapters that VMWare adds on teh guests
[02:22] <cclausen> we don't have SAN gear here, so vmotion wouldn't work anyway
[02:22] <aetaric> yeah... it is a bit complex...
[02:23] <cclausen> I've not found an affordable multiple VLAN redundant SAN solution
[02:23] <jetole> Ah I did our first SAN setup ~8 months ago and another one since
[02:23] <jetole> one in the office and one in the DC
[02:23] <aetaric> cclausen: you don't need a san.... iscsi would work, or nas
[02:23] <cclausen> everything is designed to run on the same subnet in the same datacenter and I need cross-campus redundancy
[02:23] <jetole> aetaric: iscsi is san
[02:23] <cclausen> aetaric: yeah, but what is the point if its a single point of failure?
[02:24] <cclausen> jetole: technically, iSCSI is just block-level storage.  we have software iSCSI targets on a few systems
[02:24] <jetole> san is block level storage over networking
[02:24] <jetole> san is almost more but iscsi is san
[02:24] <cclausen> well, if you put it that way, yes, that is iSCSI
[02:24] <jetole> *almost always
[02:24] <jetole> as nas is filesystem over the net san is block level which covers iscsi and ata over eth
[02:25] <jetole> I know it's rediulously simple san but it is san
[02:25] <jetole> *rediculously
[02:25] <jetole> still don't know if I spelled that right
[02:25] <cclausen> jetole: I'm thinking more about the actual storage on the backend, not the protocol used
[02:26] <jetole> you mean like clusters of DAS etc
[02:26] <cclausen> sort-of
[02:26] <cclausen> I mean the ability to migrate volumes betweens disks in different physical locations while the data is online
[02:26] <cclausen> actual storage management
[02:27] <cclausen> was looking at nexenta stuff, but it still cost a bit too much
[02:27] <jetole> ah, for management we use linux
[02:27] <jetole> we have some expensive file servers and maybe getting some DAS extensions later this year
[02:28] <jetole> but linux does what we need
[02:28] <cclausen> I use Microsoft Dfs for end-user network CIFS shares.  works very well for Windows clients
[02:28] <jetole> oh god know
[02:28] <jetole> I use DFS too
[02:28] <cclausen> I guess I need to test if Linux has Dfs client support yet...
[02:28] <jetole> I hate it
[02:28] <jetole> no
[02:28] <jetole> no it doesn't
[02:28] <jetole> use openafs
[02:28] <cclausen> haha
[02:28] <cclausen> I'm a chanop on #openafs
[02:28] <cclausen> :-)
[02:28] <jetole> I hate dfs and when I get the time I will be replacing it with openafs
[02:29] <jetole> I know very little about openafs but I use DFSr and I despise it
[02:29] <cclausen> openafs doesn't have read-write replication
[02:29] <jetole> thats what I was refering too, I forgot DFS does not mean DFSr
[02:29] <cclausen> be aware of that
[02:29] <cclausen> I do use the newer Dfs namespace and Dfs-R replication in 2003 R2
[02:29] <cclausen> the older windows 2000 stuff kind of sucked bigtime
[02:29] <jetole> I use DFSr on 2008
[02:30] <jetole> I used it on 03
[02:30] <jetole> not 2000
[02:30] <jetole> brb
[02:30] <cclausen> yeah, I need to upgrade to 2008
[02:30] <cclausen> err, 2008 R2
[02:31] <jetole> back
[02:31] <jetole> yeah we don't have any R2 yet
[02:31] <jetole> anyways, doesn't openafs do what DFSr does?
[02:31] <cclausen> I'm doing the lucid upgrade on a 2008 R2 hyper-V VM right now...
[02:32] <cclausen> no, openafs doesn't have ANY read-write replication at all
[02:32] <cclausen> just read-only replication
[02:32] <cclausen> and its not realtime
[02:32] <cclausen> you need to manually "vos release"
[02:32] <jetole> oh that sucks
[02:32] <cclausen> yeah, pretty much
[02:32] <cclausen> its on the roadmap for afs, but won't be production ready for a couple of years
[02:33] <cclausen> so until I see something else, DfsR seems to be the best available for fairly low cost
[02:33] <jetole> I use drbd mostly which does real time read/write at the block level which works great with ocfs and also you can do realtime io on glusterfs to more then two hosts but these are not windows options
[02:33] <jetole> well glusterfs may be
[02:33] <jetole> drbd definitly is not
[02:34] <jetole> so with openafs you can only write on the master server?
[02:36] <cclausen> sort-of
[02:36] <cclausen> its worse than that
[02:36] <cclausen> the afs client prefers the read-only volume be default
[02:36] <cclausen> so the RW one has a different path
[02:43] <jetole> your downplaying afs a lot for someone whose an op in that room
[02:44] <cclausen> well, its useful, but not as a Dfs replacement
[02:44] <cclausen> especially for read-write replication
[02:44] <cclausen> and its worse than that, I'm hosting the afs conference this year: http://workshop.openafs.org
[02:45] <jetole> sounds to me like the only real perk is cross platform deployment but what situation would say openafs is good for. Like what type of real world usage would you recommend openafs for?
[02:45] <cclausen> well, I just tried an upgrade from 8.04 to 10.04 and that failed catastrophically.  apaprently the installer didn't like my old kernel...
[02:45] <cclausen> jetole: if you are using it for read-only distribution
[02:45] <jetole> meh, I gave up on the upgrade option years ago. I always do fresh installs now
[02:46] <cclausen> I know a company that uses AFS to globally deploy new software to a large number of servers in 2 seconds
[02:46] <cclausen> AFS callbacks ensure that new data is read at the same time across all of the servers
[02:46] <jetole> what about say, web content on a server farm?
[02:46] <cclausen> that too
[02:46] <cclausen> pictage does that
[02:46] <jetole> and I would run openafs server on the same host as ftp
[02:47] <cclausen> you should probably ask more in #openafs
[02:47] <cclausen> we're probably way off topic for this channel
[02:47] <jetole> another night. I'm going to bed in 10
[03:04] <jaypur> cclausen, re installing :D almost done
[03:04] <cclausen> jaypur: cool
[03:04] <cclausen> I also re-started the upgrade process
[03:05] <cclausen> I first installed the default linux-image-generic kernel in 8.04.  I think I'll have better luck with this than the self-built xen kernel I was using before...
[03:05] <jaypur> hm
[03:05] <cclausen> (luckily going backwards is easywith a VM)
[03:06] <jaypur> the 10.4 is with a new resolution i think
[03:06] <cclausen> probably different video drivers
[03:12] <jaypur> cclausen, finished....
[03:12] <jaypur> now i need to see the ssh if it's working...
[03:12] <cclausen> excellent.  I'm getting farther in this install too.  I think that the installer crashed on me last time and my system was half upgraded
[03:15] <jaypur> now i'll change the ssh port
[03:15] <cclausen> jaypur: why are you changing the ssh port?
[03:16] <jaypur> because my internet service blocks ports like 22, 80....
[03:16] <jaypur> sux BRAZILIAN INTERNET lol
[03:16] <cclausen> ah, ok
[03:16] <cclausen> that is a good reason :-)
[03:17] <jaypur> changed
[03:17] <jaypur> do i need to restart ssh?
[03:17] <cclausen> I would
[03:17] <jaypur> how do i restart the app:?
[03:17] <cclausen> invoke-rc2.d ssh restart
[03:18] <cclausen> restarts the sshd server
[03:18] <cclausen> err, hmm. might not be a "2" in the command
[03:18] <jaypur> why the invoke-rc2.d?
[03:18] <cclausen> its invoke-rc.d ssh restart
[03:18] <cclausen> same as running /etc/init.d/ssh restart
[03:19] <jaypur> damn i can't connect
[03:19] <cclausen> what error ?
[03:20] <jaypur> ssh: connect to host IP port 1010: Connection refused
[03:20] <cclausen> firewall ?
[03:20] <jaypur> aaaaaaaw noooo
[03:20] <jaypur> not again
[03:20] <jaypur> @    WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED!     @
[03:20] <jaypur> now i restarted the app right
[03:20] <cclausen> what client are you sshing from?
[03:20] <jaypur> ...
[03:21] <jaypur> and again
[03:21] <cclausen> yes, that is an expected error
[03:21] <jaypur> from ubuntu the normal ssh
[03:21] <cclausen> you are sshing from your ssh to your server ?
[03:21] <jaypur> here we go again
[03:21] <cclausen> (e.g. to localhost ?)
[03:21] <jaypur> im sshing LOL from my laptop to my server
[03:21] <jaypur> same internet
[03:21] <cclausen> you need to change the known_hosts file on the laptop
[03:21] <cclausen> this error has nothing to do with the server side at all
[03:22] <jaypur> aaaaaaaaaaaaaaaah
[03:22] <jaypur> at the laptop
[03:22] <jaypur> lol
[03:22] <jaypur> ommfg
[03:22] <jaypur> YTEAAAH
[03:22] <jaypur> YEAH
[03:22] <jaypur> IT WORKED
[03:23] <jaypur> cclausen, i love you man ahahhaa
[03:23] <cclausen> this is what I was trying to say before
[03:23] <cclausen> :-)
[03:23] <jaypur> aaah sorry man i was at a bad time
[03:23] <jaypur> now i took a bath and everything is cool at my house
[03:23] <jaypur> i was having some problems
[03:23] <jaypur> sorry
[03:23] <jaypur> now i'll configurate apache yeaahh
[03:24] <owen1> cclausen: thanks for helping us here
[03:24] <cclausen> no problem
[03:24] <cclausen> I'm watching do-release-upgrade do its thing...
[03:25] <owen1> yeah, i am installing it as well now-
[03:25] <owen1> do u use vbox to manage vms?
[03:25] <owen1> i want to learn to use the text-based client of vbox
[03:26] <cclausen> I only run VMs in Microsoft products myself, Hyper-V and Virtual Server
[03:26] <owen1> so i can manage it with screen
[03:26] <cclausen> yeah, that sounds like a good plan
[03:26] <owen1> ok
[03:26] <cclausen> I wish I had a text interface to hyper-v that I knew how to use
[03:26] <cclausen> (there is powershell, but I haven't learned all that stuff yet...)
[03:27] <owen1> if u'll swith to vbox i'll be happy to blog my progress so others will learn it.
[03:28] <cclausen> write a blog post or something
[03:28] <cclausen> or write something on the community forums
[03:28] <jaypur> it works
[03:29] <jaypur> now should i install mysql, phpmyadmin, php5 and no-ip?
[03:29] <jaypur> >D
[03:31] <cclausen> hmm.. does anyone know how to get rid of graphical 10.04 boot screen and get a text log?
[03:31] <jaypur> hmmm
[03:32] <jaypur> i think you can do that at start up manager
[03:32] <jaypur> or editing the grub dunno
[03:32] <cclausen> ok
[03:32] <cclausen> whatever graphical mode is in use now is basically useless in Hyper-V
[03:33] <owen1> %s/s/php5/ruby
[03:33] <owen1> cclausen: i did it in 9.10.
[03:34] <jaypur> sudo apt-get install mysql?
[03:34] <owen1> cclausen: i deleted a file from /etc/init.d
[03:34] <jaypur> is it ok?
[03:34] <cclausen> jaypur: yeah, that should work
[03:34] <owen1> cclausen: but i can't remember it's name
[03:34] <cclausen> owen1: I wouldn't do that
[03:34] <cclausen> it can end badly
[03:35] <cclausen> if you want to prevent something from running, uninstall the package
[03:35] <jaypur> mysql php5, phpmyadmin... i think thats all i need
[03:35] <owen1> cclausen: maybe the package is called gdm? not sure..
[03:35] <cclausen> or update-rc.d -f remove <filename>
[03:35] <cclausen> gdm is the graphical login screen
[03:36] <jaypur> scp comes with ssh right?
[03:37] <cclausen> should
[03:37] <cclausen> openssh-client I think
[03:38] <jaypur> i need to transfer some files
[03:38] <jaypur> when i was "noob" at the server, i used the ubuntu desktop to manage the files
[03:38] <jaypur> now i can do it remotly...
[03:38] <cclausen> dpkg -S `which scp` -> openssh-client: /usr/bin/scp
[03:38] <owen1> cclausen: didn't u asked to get rid of the graphical login screen?
[03:38] <jaypur> and i want to transfer again remotly... and use all text based
[03:38] <cclausen> owen1: yes
[03:39] <cclausen> oh, I see, you were trying to answer
[03:39] <cclausen> not that screen
[03:39] <jaypur> i thing is...
[03:39] <cclausen> its the boot one
[03:39] <cclausen> before gdm loads
[03:40] <owen1> cclausen: r u talking about desktop or server?
[03:40] <cclausen> owen1: server
[03:40] <owen1> cclausen: is the server coming with gdm?
[03:41] <cclausen> not that I can tell
[03:41] <cclausen> its before gdm would load
[03:41] <cclausen> its a screen with the words Ubuntu 10.04 in the middle
[03:41] <cclausen> maybe it is grub
[03:41] <cclausen> I will have to look around
[03:41] <owen1> cclausen: why do u want to get rid of it?
[03:42] <cclausen> the screen updates at like one line in 5 seonds
[03:42] <cclausen> very annoying to try and use
[03:42] <cclausen> e.g. the video mode is too slow to use
[03:45] <jaypur> to run a complete web server...
[03:45] <jaypur> what do i need? mysql, phpmyadmin, php5.... anything else????
[03:45] <owen1> real web sevrer runs on ruby (-:
[03:46] <cclausen> jaypur: apache2 is a real webserver
[03:46] <cclausen> you don't need php or mysql
[03:46] <owen1> jaypur: u might need nginx/apache
[03:46] <cclausen> well, ht grub2 install rendered my vm unbootable
[03:46] <cclausen> time to find an iso to boot from...
[03:46] <jaypur> because some applications need mysql... but when i look at the tutorial they tell to install phpmyql some sh*t like that
[03:47] <owen1> cclausen: u use alternace CD, correct?
[03:47] <owen1> alternate
[03:50] <handheldCar> jaypur: http://suds.blogsite.org/content/installing-drupal-run-home
[03:50] <jaypur> handheldCar, thanks
[03:50] <cclausen> I use the server cd
[03:51] <cclausen> oh no.  not drupal!
[03:51] <cclausen> that is the way to much pain
[03:51] <cclausen> vi and html files are all you need
[03:51] <cclausen> much safer that way
[03:51] <handheldCar> well it's good to learn cms
[03:52] <cclausen> maybe, but its a pain to keep up to date
[03:52] <cclausen> and can quickly turn into a security problem
[03:53] <handheldCar> it's praised as easy, but it becomes a pita every now and then.
[03:56] <owen1> jaypur: might be a good read for you - https://help.ubuntu.com/community/SSH?action=show&redirect=SSHHowto
[03:56] <jaypur> owen1, bookmarked!
[03:57] <jaypur> scp: .: not a regular file
[03:57] <jaypur> everything is going wrong and i do all the same HAHAHA
[03:59] <handheldCar> What is scp?
[03:59] <jaypur> it's an app to transfer files from your pc to your server by ssh
[04:00] <handheldCar> what are you copying? I just download things to my server.
[04:00] <jaypur> a .zip
[04:01] <jaypur> i really need to copy it now because i have to go to sleep and start the server lol
[04:01] <jaypur> counter strike server lol
[04:02] <owen1> handheldCar: short for secure copy
[04:03] <owen1> jaypur: you can also use it to get stuff from your server.
[04:03] <handheldCar> i use sftp username@server.ip.address
[04:03] <jaypur> yea i now
[04:03] <handheldCar> put filename.zip
[04:03] <jaypur> handheldCar, but you need to configure the ports for ftp? or get a ftp client?
[04:03] <owen1> scp -r <folder> username@server.ip
[04:04] <jaypur> i want to transfer a file from my pc to the server
[04:04] <handheldCar> sftp is part of ssh
[04:04] <jaypur> scp -P 1010 /home/users/file.zip user@ipadress: /home/user
[04:05] <handheldCar> u shouldn't have to specify either home directory path.
[04:06] <jaypur> what should i do?
[04:06] <handheldCar> one of the above and report back
[04:08] <jaypur> lol it's working now
[04:08] <handheldCar> :)
[04:08] <jaypur> but didn't put any path to the server, it will tranfer to what folder?
[04:08] <handheldCar> ~
[04:08] <owen1> jaypur: home i believe
[04:09] <jaypur> thanks
[04:09] <owen1> it's so simple, ah?
[04:10] <jaypur> i've already done it... but i forgot... hahaha
[04:10] <jaypur> i love linux
[04:10] <jaypur> :)
[04:10] <jaypur> tomorrow i'll set the mysql...
[04:10] <jaypur> i'm too tired...
[04:10] <jaypur> i'll just set the counter-strike server on
[04:13] <owen1> cclausen: i found this link to automate deployment of many ubuntus. might be helpful insteald of FAI - https://help.ubuntu.com/community/AutomatedNodeDeployment
[04:18] <jaypur> now i'll sleep
[04:18] <jaypur> thanks a lot guys
[04:18] <jaypur> you made my day
[04:24] <owen1> come back tomorrow!
[04:26] <jaypur> yeah
[04:26] <jaypur> just got back
[04:26] <jaypur> i was brushing my teeth
[04:26] <jaypur> i think my server is faster now
[04:26] <jaypur> i put
[04:26] <jaypur> nice -n -20 to counter-strike server
[04:27] <jaypur> now it's aaaaaaaaall riteee
[04:27] <jaypur> thanks a lot everyone
[04:27] <owen1> come back after flossing
[04:27] <jaypur> have a good nite, and weekend....
[04:28] <jaypur> hahahaha
[04:28] <owen1> it's actually double meaning - floss = free libra and open source
[04:29] <jaypur> lol
[04:29] <jaypur> nice
[04:29] <jaypur> now i'll go
[04:29] <jaypur> bye man
[04:29] <jaypur> see ya
[04:29] <owen1> take core
[04:29] <owen1> care
[04:36] <cclausen> ok, not a fan of grub2
[04:49] <owen1> cclausen: what's going on?
[04:49] <cclausen> got grub1 reinstalled
[04:49] <cclausen> and system is booted
[04:49] <cclausen> seeing why apache isn't starting now
[04:50] <cclausen> something about suexec requiring a wrapper
[04:51] <owen1> you uninstalled grub2?
[04:51] <cclausen> no
[04:52] <cclausen> I booted to a boot cd to fix grub
[04:52] <cclausen> the chain load works
[04:52] <cclausen> the remove legacy part does not
[04:52] <cclausen> apparently there is a new apache2-suexec package
[04:52] <cclausen> I'd consider having suexec in config files and the upgrade not installing the package a bug
[04:54] <owen1> why do u use apache forL
[04:54] <owen1> ?
[05:05] <cclausen> why do I use apache?
[05:05] <cclausen> its a web server
[05:05] <cclausen> I need one
[05:06] <owen1> nginx is better in terms of cpu usage - http://www.joeandmotorboat.com/2008/02/28/apache-vs-nginx-web-server-performance-deathmatch/
[05:07] <owen1> and it's very easy to configure
[05:07] <cclausen> no, its not
[05:07] <cclausen> I need mod_auth_kerb
[05:08] <cclausen> and I need suexec functionality
[05:08] <owen1> got it
[05:08] <cclausen> and does that handle .htaccess files?
[05:08] <cclausen> I've seen people use nginx as a proxty, but I don't think it is a good general web server for a multi-user hosting environment
[05:11] <owen1> i want to install it but not sure about multi-user hosting
[05:12] <mikelifeguard> Why is allowing PHP in ~/public_html when mod_userdir is enabled considered a security risk?
[05:13] <cclausen> b/c any user can access any others user's code
[05:13] <cclausen> unless you take care to really lock down php
[05:14] <cclausen> and you can't use suexec with mod_php
[05:14] <cclausen> (I use mod_fcgid and suexec for php)
[05:15] <mikelifeguard> How could one user access another user's PHP code? (you mean to edit right?)
[05:16] <mikelifeguard> I suppose they might allow that by accident, or not knowing what chmod o+w does
[05:17] <cclausen> no
[05:18] <cclausen> if the web server can read it, I can read it
[05:18] <cclausen> I can grab files that list things like mysql database passwords
[05:18] <cclausen> don't need write acces to cause problems
[05:18] <cclausen> and ifthe web server has write access its a problem
[05:19] <mikelifeguard> ah, right
[05:28] <AdamDV> Any one got any guides on compiling apache2 on 10.04?
[05:29] <AdamDV> (Or any version)
[06:02] <ScottK> Why not just use the packaged version?
[10:04] <Bilge_> Assuming that `host` gets its information from the locally running `named`, how can I flush the cache so that it picks up an entry that every other DNS server is seeing
[10:04] <Bilge_> Also I don't know why it would cache a not found response so aggressively, but that's a separate issue
[10:09] <Schmidt> Bilge_: Shouldn't a caching DNS go out and fetch unknown DNS entrie ?
[10:10] <Schmidt> from its DNS :)
[10:13] <Bilge_> What
[10:19] <owen1>  i usually delete /etc/init/gdm.conf so i will not see the graphical login manager and have a text-based login. is this the correct way?
[10:20] <Schmidt> owen1: You could just disable the gdm and/or xdm from the init system
[10:24] <johnlvpl_> Question: can Ubuntu Server feed an internet style cafe?
[10:35] <johnlvpl_> Johnlvpl
[10:44] <owen1> Schmidt: can u elaborate?
[10:45] <Schmidt> owen1: there is a tool called update-rc.d which one can use to disable for example gdm
[10:46] <Schmidt> used like "$ update-rc.d gdm remove"
[10:46] <owen1> Schmidt: awesome
[10:47] <Schmidt> owen1: and I should upgrade the server to 10.04 soon :)
[11:08] <johnlvpl> Guys is ubuntu server be appropriate to run an internet café
[11:11] <wathek> hello all
[11:11] <wathek> from where can I get the pxe files for the lucid ?
[11:14] <owen1> johnlvpl: yes. the internet cafe module was added since 9.04
[11:16] <owen1> lucid was installed, i am restarting and hopefully will u see u all in 40 seconds (-:
[11:20] <Pirate_Hunter> has anyone tried lvm install with lucid because i've tried twice and /root doesn't get recognised, not sure if it is a problem with lucid or grub considering in 8.04 I had lilo as my loader, anyone had the same issue?
[11:26] <owen1> works!
[11:27] <johnlvpl> Hi guys can ubuntu server run an internet cafe
[11:28] <RoyK> johnlvpl: don't see why not - even win95 could do that :)
[11:29] <Pirate_Hunter> has anyone tried lvm install with lucid because i've tried twice and /root doesn't get recognised, not sure if it is a problem with lucid or grub considering in 8.04 I had lilo as my loader, anyone had the same issue?
[11:30] <RoyK> Pirate_Hunter: it doesn't help very much to repeat the question - people usually can read
[11:30] <johnlvpl> royk: I want to steer away from a windows environment and have something that is secure; reliable and has the appropriate monitoring tools
[11:30] <RoyK> if they can't, it matters even less :þ
[11:31] <RoyK> johnlvpl: do you want the server as the router/proxy or what?
[11:31] <RoyK> thin terminals?
[11:31] <johnlvpl> royk: was looking at a solution that offers desktops, thin clients and a wifi access
[11:32] <RoyK> look into LTSP
[11:32] <RoyK> if wifi access is something you're gonna sell, dunno any standard solution, but it should be trivial to write one
[11:33] <johnlvpl> looking at ltsp now thanks royk
[11:33] <owen1> Pirate_Hunter: i just installed xubuntu lucid
[11:35] <RoyK> Pirate_Hunter: /root or /?
[11:36] <RoyK> Pirate_Hunter: is /boot on a separate filesystem?
[11:36] <Pirate_Hunter> RoyK, yup
[11:36] <RoyK> yup what?
[11:38] <Pirate_Hunter> RoyK, im doing a raid 0 with lvm, /boot is on raid1, the rest on 0, on top of that I've got lvm as i.e. /, /home, /var... etc but /root is the only one not being recognised on boot
[11:38] <RoyK> out of interest - why would you have /root on a separate filesystem?
[11:39] <RoyK> also, raid 0 is playing with matches and petrol
[11:41] <RoyK> disks die, it's in their nature
[11:42] <Pirate_Hunter> RoyK,, why wouldn't I? It is the setup I chose on top of that the only thing I can find for my problem on google is this link http://www.mail-archive.com/ubuntu-bugs@lists.ubuntu.com/msg2183075.html but I aint using a powerpc. On top of that raid 0 works quite well under 8.04 wouldn't know why lucid would be any different aand this are new disks
[11:43] <RoyK> Pirate_Hunter: /root should be on / since you want it to exist in case of a non-root-filesystem failure
[11:43] <RoyK> raid0 surely works, I'm not saying that, but it's not really safe
[11:45] <Pirate_Hunter> RoyK, /root is mounted as / just typed it as /root as its simpler to understand in my opinion, I understand 0 aint the most safest but for what I am doing it is a good choice, just wondering if you would know why / isn't being picked up instead sending me to busybox intiframs
[11:46] <RoyK> might be a grub2 failure
[11:46] <RoyK> there have been boot problems with grub2, at least in the alpha/beta stages of lucid
[11:47] <owen1> cclausen had some issues with grub2 a few hours ago
[11:48] <jaypur> owen1, :D
[11:49] <jaypur> cclausen, :D
[11:49] <Pirate_Hunter> I'm suspecting it could be grub could never get it working on 8.04 so went by the default choice of lilo as I gave up installing grub manually, however lucid doesn't give an option for lilo. Well im wiping the disks again and will try to the procedure again hopefully I'll be able to identify the problem
[12:07] <Database> I've come across a very strange problem with my server. I just upgraded it from 9.10 to 10.04, and the server is now no longer booting - can't SSH in, no video is being displayed on the physical monitor, not responding to pings. The hard drive is active, as I can hear it, and there appears to be some sort of network activity, but I cannot interact with the machine at all.
[12:13] <mcas> Database: du you have a serial port on the server?
[12:13] <Database> mcas: Not that I know of.
[12:14] <Database> The server's an old IBM NetVista, see.
[12:14] <mcas> can you boot into a live cd
[12:15] <Database> I would have to check - the CD drive wasn't responding, but I think that's simply because it wasn't plugged in. Hold on, I'll go try.
[12:20] <Database> Hmm. It looks like there's not even any video output at POST.
[12:23] <mcas> but POST is still readable?
[12:23] <Database> No.
[12:23] <Database> There's no output at all.
[12:24] <Database> I turn the server on, everything spins up - the CD drive, the hard drive - but there's no indication as to what it's actually trying to do.
[12:24] <mcas> strange idea but ... is it possible that the video card is broken?
[12:24] <Database> I hope not - it's onboard!
[12:25] <Database> But it is a distant possibility that it's somehow been fried without me noticing.
[12:27] <Database> I suppose the logical next step would be to dig up a graphics card and see what happens.
[13:51] <martinjh99> I have enabled mod_rewrite with a2enmod rewrite and restarted apache...  Is there anything else I need to do to get rewrite working...?
[13:52] <martinjh99> I seem to remember I had to change a setting somewhere but can't remember where...
[14:13] <Falc> change AllowOverride in sites-available/<config file>
[14:20] <martinjh99> Have done still working - Druapl is saying it is not configured...
[14:20] <martinjh99> not working
[14:20] <Database> have you restarted again?
[14:20] <martinjh99> yup
[14:22] <wathek> hello all I'm configuring a PDC but when I try to add a Windows machine to the Domain I put the username/password of the domain admin and it says the username is unavailble
[14:27] <RoyK> wathek: I guess #samba would be of more help
[14:31] <wathek> RoyK, ther's no samba 3 for lucid ?
[14:35] <PirosB3__> Hi all, i am a happy user of ubuntu server, currently 8.04 , which benefits are there of upgrAding to 10.4? Thanks
[14:42] <joschi> Pirate_Hunter: https://help.ubuntu.com/community/Server/TechSpecs/new plus the things mentioned in http://www.ubuntu.com/getubuntu/releasenotes/1004
[14:43] <joschi> meh, I meant Piros83__ but I was obviously too late
[14:56] <wathek> I got a problem I couldn't login to my ubuntu-server it says  Login tamed out after 60 seconds
[15:07] <Pupeno> Any recommendations on how to backup Amazon EC2 instances?
[15:28] <bluethundr_> I need to purge courier-authdaemon so I can rebuild it from source so I can work around an incompatibility in an old packaged version. but when I try to do that I get this error: http://pastebin.com/wiBbvUC8
[15:28] <bluethundr_> any thoughts/hints?
[15:34] <marshall> i just installed dovecot-postfix, how do i create email users?
[15:51] <PirosB3_> i all, i am a happy user of ubuntu server, currently 8.04 , which benefits are there of upgrAding to 10.4? Thanks
[15:53] <jaypur> PirosB3_, the main benefit i saw was the screen resolution, and i think it's faster
[15:55] <PirosB3_> Great :) although i dont hAve X installed, only ssh. Do you think upgrade could screw something? Thanks :D jaypur
[15:58] <jaypur> PirosB3_,  look my friend...
[15:58] <ukev> Hi, I've installed ubuntu 10.04 and migrated my kvm guests from gentoo, but now they are not working anymore
[15:58] <ukev> the error is: internal error cannot parse QEMU version number in ''
[15:58] <ukev> what does that mean?
[15:58] <jaypur> i don't know, i can't give you a word that would screw anything... but... dependes how you upgrade it i think... i use X with ssh.... i installed nautilus at my server so i start ssh -X ip... and sudo nautilus :)
[15:58] <jaypur> PirosB3_,
[15:58] <ukev> I could not found anything, also I could not found a bug in launchpad
[15:58] <RoyK> ukev: it probably means "I'm allegic to gentoo! go away!"
[15:59] <ukev> the guests are ubuntu...
[16:00] <PirosB3_> Thanks jaypur
[16:00] <jaypur> PirosB3_, you're welcome :D
[16:02] <RoyK> ukev: sorry - I just don't know - don't use kvm
[16:02] <RoyK> s/don/I don/
[16:03] <ukev> RoyK, I thought kvm is the preffered way for virtual machines in ubuntu since it dropped support for xen and openvz and a few more..
[16:06] <Jeeves_> ukev: It is
[16:06] <ukev> (btw. I'm using kvm with libvirt)
[16:06] <Jeeves_> RoyK just isn't using it
[16:06] <ukev> and it seems like a libvirt error message
[16:09] <ukev> I tried to start the virtual machine with virsh and through virt-manager - both times same error message, I also created a new virtual machine because I thought the config file syntax has changed but there is the same error message
[16:11] <AdamDV> ukev: https://bugs.launchpad.net/bugs/490370 http://bugs.gentoo.org/show_bug.cgi?id=294169 http://osdir.com/ml/libvir-list/2009-05/msg00192.html http://www.mail-archive.com/libvirt-users@redhat.com/msg00107.html
[16:11] <AdamDV> http://lmgtfy.com/?q=internal+error+cannot+parse+QEMU+version+number+in+ubuntu
[16:12] <RoyK> is kvm really as good as xen these days?
[16:13] <AdamDV> RoyK: Not from my expierence
[16:13] <AdamDV> Xen is far superior for 99% of things.
[16:13] <AdamDV> Especially for VPS's.
[16:15] <ukev> AdamDV, thanks, I found these ressources, too, but I could not find a solution in it. I don't want to pass usb inside a machine and I don't need a wrapper script
[16:16] <AdamDV> Heh.
[16:25] <ukev> ok, I figured out that it will start if I disable apparmor like the one in the launchpad bugreport, but I don't use any custom wrapper script
[16:25] <ukev> the emulator entry points to /usr/bin/kvm
[16:30] <ukev> ok it works now after removing the <emulator> tag
[16:30] <ukev> thanks
[16:34] <AdamDV> Anyone know what Ubuntus apache2 --prefix is?
[16:39] <PhoenixSTF> hello i need to change the colors of the Server terminal any leeds?
[16:40] <AdamDV> PhoenixSTF: Your $PS1 variable, and your ~/.bashrc
[16:40] <AdamDV> And your terminal profile ;)
[16:41] <PhoenixSTF> so i gotta change the .bashrc file
[16:43] <AdamDV> yES
[16:43] <AdamDV> *Yes :|
[16:43] <PhoenixSTF> any simpler way?? its a work my brother gotta do for school so there is no simpler way?
[16:46] <AdamDV> No.
[16:46] <AdamDV> Unless you change your terminal profile.
[16:46] <PhoenixSTF> and everytime i reboot the server the collors will stay?
[16:47] <AdamDV> Yes
[16:47] <AdamDV> But, only you will see them.
[16:48] <PhoenixSTF> ya but thats on command screen not on gui
[16:48] <PhoenixSTF> so no problem right
[16:49] <PhoenixSTF> well ive give a try
[16:49] <PhoenixSTF> thanks a lot adam
[16:49] <PhoenixSTF> :D
[16:49] <PhoenixSTF> cheers m8
[17:10] <AdamDV> I should check my screen more often :/
[17:19] <wtfplymouth> after nearly 2 days of trying still no sucess of automatically skipping mount errors on boot with 10.04
[17:22] <cclausen> AdamDV: where you asking for apache2 --prefix at compile time?
[17:22] <AdamDV> cclausen: Nevermind, I've decided to rebuild the deb instead of compiling
[17:22] <cclausen> AdamDV: you can apt-get source for the packages and look at the rules file
[17:23] <cclausen> I don't think there is a single --prefix.  more likely various places where specificed to match debian policy
[17:24] <AdamDV> cclausen: Thats what I did :)
[17:24] <cclausen> what did you need to change?
[17:25] <cclausen> b/c just rebuilding the deb straight up is the same as just installing the provided package
[17:25] <AdamDV> --suexec-docroot and --suexec-html
[17:25] <AdamDV> I changed those two in the debian/rules
[17:25] <cclausen> there is now a custom suexec package
[17:25] <AdamDV> And incremented the version and wrote my changes in the changelog.
[17:25] <cclausen> that I thought let you do things like that
[17:25] <AdamDV> If there is, its too late now.
[17:26] <AdamDV> But, Its still a good idea to rebuild the deb, because I'm going to put it on around 100 or so servers.
[17:26] <cclausen> http://packages.ubuntu.com/lucid/apache2-suexec-custom
[17:26] <AdamDV> (Were setting up a hosting infrastructure)
[17:26]  * AdamDV clicks
[17:26] <cclausen> using it in more places I'd say its better to NOT rebuild it yourself so that normal security updates will apply
[17:28] <AdamDV> Hmm
[17:29] <AdamDV> How would I use apache2-suexec-custom ?
[17:41] <Jeeves_Moss> just wondering how to restart authdaemond.  I'm getting this error.  http://pastebin.com/f9vTy7V7
[17:49] <AdamDV> cclausen: If I install apache2-suexec-custom do I also have to install apache2-suexec
[17:54] <failover> I'm tryng to install Lucid 10.04 64b from pendrive in a hp proliant DL 145 G3, but after choose a keyboard layaut, i can't go ahead because the instaler can't find the packages in a "cd-rom"
[17:54] <failover> Am i missing someting ?
[17:55] <failover> i'm think that is impossible for the instaler detect a cd-rom if the machine dosen't have one...
[17:55] <failover> i build the bootable usb using unetbootin, any tips ?
[18:09] <cclausen> AdamDV: I am not sure.  I assume it would pull it in as a dependancy if it was needed also
[18:13] <AdamDV> W00T!
[18:14] <AdamDV> nano on Lucid has syntax highlighting :D
[18:28] <cclausen> failover: is there a netinstall package that you can put on the usb drive instead of using a cd-rom based one?
[18:31] <failover> cclausen, i already got a solution,  https://help.ubuntu.com/community/Installation/FromUSBStick
[18:32] <failover> cclausen, i start the install with this command: "install cdrom-detect/try-usb=true
[18:40] <amstan> hey, can you guys help me with a noob question, i tought i understanded this, but apparently i don't
[18:41] <amstan> I can't write stuff on a folder in my server for some reason
[18:41] <amstan> hold on, i'll post it on pastebin
[18:41] <cclausen> amstan: that isn't very descriptive at all
[18:43] <amstan> http://pastebin.com/RxkgVCPs
[18:43] <amstan> cclausen: that link should provide more info
[18:43] <cclausen> amstan: the "+" in the ls output means there are ACLs on the files
[18:43] <amstan> cclausen: yes, i want to add that too
[18:44] <amstan> cclausen: http://pastebin.com/AGcEkhHX
[18:44] <amstan> i just did a chmod 770 on all of those files
[18:44] <amstan> i tought they should of changed the acls
[18:45] <cclausen> man acl
[18:45] <cclausen> man setfacl
[18:45] <cclausen> man getfacl
[18:45] <cclausen> and see if that helps
[18:45] <cclausen> chmod won't change the ACLs
[18:45] <cclausen> that is kind of the point of the ACLs
[18:46] <amstan> cclausen: so i'm to understand that they'll complement eachother and unless both give permissions for something, the action is denied
[18:46] <cclausen> yeah, basically
[18:46] <amstan> ok
[18:46] <cclausen> you can setfacl --remove-all to get rid of the ACLs, but this can have complications.
[18:47] <cclausen> you might need to run that as root, assuming you have root access
[18:47] <amstan> yes
[18:47] <cclausen> (or your computer may use a different ACL scheme entirely)
[18:47] <cclausen> looks like your ACLs only allow write to the owner
[18:48] <amstan> another question, is it possible that some sftp clients will ignore the acl settings?
[18:48] <cclausen> what do you mean ignore?
[18:48] <cclausen> most sftp clients won't know that ACLs exist
[18:48] <amstan> ok, nvm that
[18:48] <cclausen> but they won't allow you to just write there anyway
[18:48] <amstan> i still have the same issue with the denied
[18:48] <amstan> i just did a --remove-all
[18:49] <cclausen> are you sure it worked?
[18:49] <amstan> getfacl still reports group::r-x
[18:49] <amstan> wait.. what?
[18:49] <amstan> even the chmod is wrong now
[18:49] <cclausen> try with setfacl -d --remove-all
[18:50] <cclausen> let me ask, why were acls setup in the first place?
[18:50] <cclausen> if you don't want them, don't use them
[18:50] <amstan> cclausen: because i have multiple users writing files to places
[18:50] <cclausen> you can remove "acl" from the mount options in fstab and reboot
[18:51] <amstan> it's usually through stuff like samba
[18:51] <cclausen> normal unix permissions should work for most cases
[18:51] <cclausen> unless you have a specific reason to use them
[18:51] <cclausen> I mean they work, but you have to know what you are doing
[18:51] <amstan> well.. the problem i have is that when a user creates a new file, none of the other users can edit it
[18:51] <amstan> and it's supposed to be a public folder
[18:51] <cclausen> so create a group for the users, put everyone in it
[18:51] <cclausen> and have th user chmod g+w
[18:51] <cclausen> and possibly chgrp it if needed
[18:52] <amstan> they usually do the file stuff through samba
[18:52] <cclausen> oh, well, that is a different problem if its through samba
[18:52] <amstan> it's a family server, i can't expect my mom or dad to chmod stuff
[18:52] <cclausen> erad the samba docs to figure out how to set the acls correctly
[18:55] <vraa_> i think cclausen has it right, if your family is on windows, you take their usernames, and create the same usernames on the solaris, put them in a group, and chmod/chgrp the folder with the group's goahead
[18:56] <vraa_> i am on that part right now setting up a zfs fileserver, i am going to try to get osol added to a win2k8r2 active directory and get idmap to do all the stuff
[18:57] <vraa_> oops sorry i thought you were using zfs
[18:57] <vraa_> although the same might apply
[18:57] <amstan> here's a problem i had: i uploaded photos on the server with samba, and all the files would just get alex:alex 775
[18:57] <amstan> same with sftp
[18:58] <amstan> so nobody else could edit them, even if the parent dir was writable by others
[18:59] <amstan> acls seem to be working
[19:11] <cclausen> vraa_: unless I'm missing something, there is no native zfs support for linux
[19:11] <vraa_> ya sorry i thought we were talking about osol
[19:12] <cclausen> vraa_: this channel is titled ubuntu-server
[19:21] <cjs> Is there an image from which I can install that VM-optimized kernel, rather than the standard one?
[19:23] <cclausen> there is jeos version of ubuntu
[19:23] <cclausen> not sure if its been updated for 10.04 yet or not though
[19:23] <cjs> Ah, I think that is the one, isn't it?
[19:24] <kklimonda> 2/b 18
[19:24] <kklimonda> heh
[19:24] <cclausen> I'm using the linux-image-generic under Hyper-V and it seems to work
[19:25] <cjs> Oh, it does work. I'm just looking for something efficient.
[19:25]  * amstan is installing zoneminder
[19:25] <cjs> This whole mess happened because I decided to take the "easy" way out and use vmbuilder to build VMs. Which turns out, if you're not using bog-standard options, to just completely ignore what you asked for half the time. Of course you only find out ten or fifteen minutes later.
[19:28] <AdamDV> amstan: whats zoneminder?
[19:29] <scar> so jeos is better to use for a VM than standard ubuntu-server?
[19:30] <amstan> AdamDV: this cool video security program
[19:30] <amstan> it has a web interface
[19:30] <AdamDV> nise
[19:30] <amstan> so you can install cameras around the house and so on
[19:30] <amstan> yeah, look it on wikipedia
[19:31] <cjs> Yeah, the jeos kernel (which is just a standard package, if I understand correctly) is leaner and uses virtual devices in preference to emulated "real" ones.
[19:33] <scar> you add '--flavour virtual' to the vmbuilder's options
[19:33] <amstan> btw, anyone know how to fix the "'Access denied for user 'debian-sys-maint'@'localhost' (using password: YES)'" error?
[19:33] <amstan> i reinstalled the os, but copied over the /var/lib/mysql/ folder
[19:33]  * scar wonders if --flavor works
[19:33] <amstan> so the user account has a different password, where can i find what password ubuntu is trying to use?
[19:34] <cjs> scar: Yes, but vmbuilder has been my Big Pain In The Ass for like a day now, which is why I'm wanting to avoid it. Though if you can make it work for me, I will send you money for beer.
[19:34] <RoyK> --add-colourful-flowers
[19:34] <scar> cjs, what else has been a problem?
[19:35] <cjs> https://bugs.launchpad.net/ubuntu/+source/vm-builder/+bug/573171
[19:36] <cjs> It's simple, I would think. I want to build on a block device (in this case, an LVM volume). I want it to use the whole darn thing.
[19:37] <cjs> But I think that vmbuilder's command line parsing (and I only got to this after misery with using an options file) is so buggy that it's rather hopeless for me to try to get it to build what I want.
[19:38]  * amstan has fixed his mysql problem
[19:38] <scar> hmmm... well i will be setting up a server with several VM's so i'll see what happens.  i want to figure out how to use the free space in my volume group as a hard drive and have the VM directly make partitions in it
[19:41] <cjs> scar: You don't want the *VM* to make partitions in it. Just make a logical volume for each "drive" you want to use in a VM, and pass those in as the "drives" for VMs.
[19:41] <cjs> Does that make sense?
[19:41] <gsfs> what would be the natural steps to make sound work after install a fresh ubuntu server edition?  I did install alsa-utils. turned the sound up, but get no sound
[19:42] <gsfs> using mocp for testing sound.. server will become mpd server if I get the sound to work
[19:46] <gsfs> btw  the package speaker-test would be nice to have in 10.04 too :)
[19:46] <scar> cjs, i want the VM to make partitions on it, then i can easily resize. i've done what you said but, on the host, the logical volumes still show up as one big unallocated volume
[19:46] <scar> i think iwant to use a logical volume pool, but i can get it defined right.
[19:47] <scar> http://libvirt.org/storage.html#StorageBackendLogical
[19:47] <AdamDV> msantos: Hey! I thought I was the only one in Toronto to use Ubuntu.
[19:47] <cjs> scar: Oh, there's tricks with a loopback mount to actually get the partitions.
[19:48] <cjs> The issue here is that the Linux loopback concept, unlike the BSD one, does not treat the new device as a real disk with a partition table, but as just a partition. But of course when you pass that partition in as a "disk" to KVM, it treats it like a disk.
[19:49] <cjs> But you can confirm pretty easily that you really do have a partition table there. Do an install on to an LVM volume, and then "fdisk -l" it. You'll see a partition table.
[19:51] <scar> so, when i want to increase the size of a partition in a VM, i can still increase the size of the logical volume. then i can resize within the VM?
[19:54] <gsfs> would sound work out of the box normally on a ubunut server??
[19:54] <gsfs> or do I have to install and tweak to make thing workl??
[19:55] <cjs> scar: Right.
[19:55]  * scar still wonders what a logical volume <pool> would be for
[19:56] <DanC> should "Upgrading Using the Alternate CD/DVD" work as documented in https://help.ubuntu.com/community/LucidUpgrades ?
[19:56] <cjs> I'm contemplating that as a strategy for my next try. Install the standard 5GB partition on the logical volume, and then expand it to fill the rest of the logical volume in some way. Clearly not while the LVM is running, given that it's the root....
[19:56] <DanC> it starts downloading from the net no matter what I do
[19:56] <cjs> scar: do you have a reference for the idea of a logical volume pool?
[19:57] <cjs> (I know RAID and disk crap pretty well after 20 years of sysadmin....)
[19:57] <scar> cjs, just that libvirt.org link above...
[19:57] <cjs> DanC: It will, I believe, always download even newer stuff from the net after it installs the versions from the CD.
[19:58] <DanC> frustrating, since there's a dialog that asks if I want that and promises not to use the net at all if I say no
[19:59] <cjs> scar: Ah!
[19:59] <cjs> DanC: have you considered unplugging the network cable? :-)
[20:00] <DanC> I did 'ifdown eth0'
[20:00] <DanC> and it said "failed to download [zillions of packages]"
[20:00] <cjs> DanC: Let me clarify here: you were using the installer?
[20:01] <DanC> I was using: gksu "sh /cdrom/cdromupgrade"
[20:01] <cjs> scar: That's brill. Libvert can just as easily allocate chunks of an LVM VG as files in a directory. Great.
[20:02] <DanC> is there a more direct path, using aptitude or the like?
[20:02] <scar> cjs, that's what i thought but i can't figure how to use that <pool> block
[20:02] <cjs> DanC: Not sure what that means. What I was thinking of is basically that if you tell your installer that you have network connectivity, it will try to download stuff. Otherwise it won't.
[20:03] <DanC> gksu "sh /cdrom/cdromupgrade" is the command given under https://help.ubuntu.com/community/LucidUpgrades#Upgrading%20Using%20the%20Alternate%20CD/DVD
[20:03] <DanC> i.e. "Upgrading Using the Alternate CD/DVD"
[20:04] <cjs> scar: You're working with a VG that's already defined and up and running?
[20:05] <cjs> scar: if that's the case, just <pool type="logical"><name>MyVG</name></pool> should be all you need....
[20:05] <gsfs> used hours and hours moving music into my new mpd server, and now sound doesnt want to work :\  I dont really know how to start troubleshoot, but alsamixer finds soundcard (irq 17) and everything looks fine,  but no sound.  testing with mocp .. anyone who's good at this here who can give a hint?
[20:06] <DanC> I can only sympathize, gsfs
[20:06] <DanC> sound is working for me these days, but I've had plenty of trouble like that in the past
[20:07] <scar> cjs, yes. i tried that... i put it into my vm's XML file in /etc/libvirt/qemu/vm.xml but it disappears when i redefine the xml with virsh
[20:08] <cjs> scar: Doesn't that go in that storage.xml or whatever file?
[20:08] <scar> cjs, i don't know
[20:08] <scar> i couldn't find any info where it actually goes
[20:09] <cjs> As in, /etc/libvirt/storage/default.xml?
[20:09] <axisys> how do I upgrade from 9.04 to 10.04 in one jump ?
[20:09] <cjs> Yes, I'm almost certain.
[20:09] <DanC> trying cdromupgrade again... here's the  "Include latest updates from the Internet?" dialog text: http://paste.pocoo.org/show/208565/ . note "If you answer 'no' here, the network is not used at all."
[20:09] <cjs> (How did I find out about that file? etckeeper is your friend. :-))
[20:10] <scar> cjs, so there is already a <pool type=dir> in there... should i just delete that whole thing? or add another <pool> block underneath?
[20:11] <scar> i don't store images in directories...
[20:11] <cjs> I'd probably add another. Is there any reason to kill the current one?
[20:11] <axisys> sudo do-release-upgrade --devel-release downloads the karmic.tar.gz .. i was expecting it dowload lucid.tar.gz
[20:11] <DanC> so I pick no, but then I get a "Do you want to start the upgrade?" dialog that says "You have to download a total of 1,209M."
[20:12] <cjs> I'd guess that the GUI gives you choices and so on...is that what you're using?
[20:12] <cjs> I'm actually tempted to give this a go right now, but I really need to get to bed soon.
[20:12] <cjs> It may only be 2 a.m. in India, but it's 4 a.m. in Japan.....
[20:14] <cjs> DanC: And this isn't just a brain-fart from someone who didn't realize that you "download" upgrades from the CD as well?
[20:14] <DanC> no... my network definitely goes busy
[20:14] <cjs> Hm.
[20:14] <DanC> and as I said earlier, when I "ifdown eth0", it failed
[20:14] <cjs> Errr...."Welcome to the f**ked by Ubuntu club?" :-)
[20:15] <cjs> (More seriously, I can't see where you could be going wrong here.)
[20:15] <DanC> what is this "do-release-upgrade" you speak of? should that work?
[20:17] <axisys> DanC: are you talking to me?
[20:17] <DanC> I guess so.
[20:17]  * DanC sees it documented as a server upgrade deely on https://help.ubuntu.com/community/LucidUpgrades ... gives it a whirl...
[20:17] <axisys> DanC: do-release-upgrade should work for upgrade .. yes
[20:17] <cjs> Well, thanks for the help. Night.
[20:18] <axisys> DanC: my question was if i can do the upgrae from 9.04 to 10.04 in one jump w/ do-r-up
[20:18] <DanC> can I expect do-release-upgrade to find packages on /cdrom ?
[20:18] <axisys> DanC: it upgrades over the network
[20:18] <axisys> u can add your cdrom as a repo later
[20:18] <DanC> felgercarb: "You have to download a total of 1748M. This download will take about
[20:18] <DanC> 34 days 1 hour with your connection. "
[20:18] <axisys> DanC: ^
[20:19] <DanC> I'll be old later
[20:19] <axisys> DanC: that number changes a lot.. i saw it jumped from 20 hrs to 1 hr.. it is dynamic calculation.. not perfect..
[20:19] <DanC> is there seriously no way to upgrade from an .iso image?
[20:20] <axisys> DanC: if you have flaky network.. that number will be changing a lot
[20:20] <axisys> DanC: there is using the alternate iso image
[20:20] <DanC> do tell... I have the alternate iso image
[20:20] <axisys> DanC: according to the upgrade url
[20:20] <axisys> !upgrade
[20:21] <axisys> DanC: https://help.ubuntu.com/community/LucidUpgrades
[20:21] <axisys> DanC: Upgrading Using the Alternate CD/DVD
[20:21] <Pirate_Hunter> clean install of ubuntu 10.04 on boot im sent to busybox with initramfs due to the fact it doesn't pick up / on lvm, how do I go about correcting this?
[20:21] <DanC> yeah... that's what I've been trying to do
[20:22] <DanC> I run cdromupgrade and say "no, I don't want stuff from the net" but then it tries to grab stuff from the net anyway
[20:22] <axisys> DanC: if that fails .. you could just do do-r-u
[20:22] <DanC> round and round we go... will do-r-u get packages from the mounted .iso image?
[20:23] <axisys> DanC: no.. you have to add the cdrom iso after you finish upgrade.
[20:23] <axisys> DanC: add cdrom as your repo.. after you finish upgrade
[20:23] <DanC> then there's no way to upgrade from an .iso image. QED. crap
[20:24] <axisys> DanC: i already told you everything you need to know
[20:24] <axisys> DanC: don't expect different answer by asking it again
[20:25] <DanC> do you acknowledge that I have followed the "Upgrading Using the Alternate CD/DVD" instructions to the letter and it doesn't work?
[20:25] <axisys> DanC: does not work does not say much.. state what you experiencing exactly
[20:26] <axisys> DanC: if someone experience the same issue .. then you probably can get more help
[20:26]  * DanC looks to see if axisys joined since the 1st time I said exactly what I'm experiencing... doesn't look like it
[20:27] <axisys> DanC: if you explained and got no help yet, then you might want to post it in mailing list
[20:28] <axisys> DanC: usually people are very helpful there as well and you get a broader audience
[20:29] <DanC> re your 9.04 question, axisys, while researching my problem, I see: "Please note that you cannot upgrade directly from a version that is earlier than 9.10." in http://hubpages.com/hub/Ubuntu-Offline-Upgrade
[20:35] <axisys> DanC: thanks
[20:35] <axisys> nice page ;-)
[20:35] <DanC> unfortunately, that article gives exactly the same instructions for my case... but yeah... with nicer decoration ;-)
[20:36]  * DanC considers a clean 64bit install
[20:37] <axisys> DanC: for desktop or server? that 64 bit
[20:37] <DanC> desktop
[20:37] <axisys> DanC: i would highly recommend otherwise.. and highly recommend 64bit for servers
[20:37] <axisys> DanC: lots of plugin wont work
[20:38] <axisys> DanC: i have been using ubuntu for 4/5 years.. and every time i go that path i suffered
[20:38] <DanC> I mostly use my mac for consumer-mode plug-in stuff... but... I do like to watch hulu and use skype on linux
[20:39] <axisys> DanC: my unsolicited advise :-)
[20:39] <axisys> DanC: hulu / skype both works with 32bit.. somecase performance is bad with 64bit on desktop
[20:40] <DanC> do you mean "works with 64bit"? (of course it works with 32bit)
[20:41] <axisys> DanC: no i thought you are doing 64bit for hulu.. nm
[20:42] <DanC> I'm considering 64bit and wondering whether hulu will still work
[20:42]  * DanC finds a bug that might be relevant to my situation... https://bugs.launchpad.net/ubuntu/+source/update-manager/+bug/571968
[20:47] <DanC> that's probably not exactly my problem, but it does tell me where to look for some log files, in which I find:
[20:48] <DanC> 2010-05-01 14:10:23,159 INFO ignoring missing '/cdrom/dists/lucid/restricted/debian-installer/binary-i386/Packages'
[20:48] <DanC> I have a Packages.gz there instead. that should work.
[20:49] <DanC> 2010-05-01 14:10:23,245 DEBUG AptCdrom.add() returned: True
[21:02] <metalf88011> does anyone know how I access cherokee-doc after I install them? I know I can read them using nano but they look like I should be able to access them as web pages
[21:15] <metalf88011> is anyone using a cherokee web server?
[21:32] <alienseer23> I upgraded my home server to lucid, and now almost every php script (drupal, ampache, gallery2, and so on) are all jacked... is there a major change in the way php behaves in lucid I need to be aware of to correct potential problems in upgrading?
[21:34] <marshall> i've changed the hostname on my jaunty mail server but postfix seems to still be sending mail as the old hostname. how do i fix this?
[22:08] <cclausen> marshall: I often have to edit /etc/mailname to fix that
[22:08] <cclausen> alienseer23: can you be more specific?  What exactly doesn't work?
[22:10] <alienseer23> cclausen: in ttrss, I get this "Sanity check: invalid RPC reply" which from my understanding means that the backend is not, for some reason, giving the frontend valid xml
[22:10] <alienseer23> cclausen: for starters :D
[22:10] <cclausen> alienseer23: just create a simple file that calls phpinfo();  does that work?
[22:10] <artisan> hi, does the installer for the server branch come with lvm and crypt options?
[22:10] <alienseer23> I cannot even log into the reader to export my feeds
[22:11] <alienseer23> cclausen: yes
[22:11] <cclausen> alienseer23: did you over-write your php.ini when prompted during the upgrade?  Or are you using the original one that you had?
[22:11] <alienseer23> cclausen: original
[22:11] <cclausen> yeah, well, I'd look into that
[22:12] <cclausen> there should be a new in teh same directory with a different extention
[22:12] <cclausen> also, does running php5 -v produce any warnings?
[22:12] <cclausen> or running php5-cgi -v ?
[22:12] <alienseer23> cclausen: I have always retained the original, let me go look into diffs and be back
[22:12] <alienseer23> hang on
[22:12] <cclausen> (are you using mod_php?)
[22:13] <alienseer23> yes there are a few warnings
[22:13] <cclausen> artisan: I think that lvm and crypto are there in server too.  I don't use those options though.  I'd say to just boot it and see
[22:13] <cclausen> alienseer23: yeah, fix the warnings and I bet most ofthe problems will go away
[22:13] <alienseer23> "PHP Deprecated:  Comments starting with '#' are deprecated in /etc/php5/cli/conf.d/idn.ini on line 1 in Unkn
[22:13] <alienseer23> "
[22:14] <cclausen> alienseer23: just edit that file, find the "#" and replace with ";"
[22:14] <alienseer23> ok, hang on :)
[22:14] <cclausen> its likely the top line of the file
[22:14]  * cclausen just did that last night
[22:15] <artisan> cclausen: yeah, i am just downloading the server is onow, but wanted to save bandwidth, as I had to download alternate iso 4 times, to realize that it just won't boot from usb stick, so now going for server :)
[22:17] <alienseer23> cclausen: um... "PHP Warning:  PHP Startup: Unable to load dynamic library '/usr/lib/php5/20090626+lfs/pcntl.so' - /usr/lib/php5/20090626+lfs/pcntl.so: cannot open shared object file: No such file or directory in Unknown on line 0"
[22:18] <alienseer23> How does one go about resolving this?
[22:18] <cclausen> alienseer23: you'll have to comment that out.  that means you custom compiled something, probably via PEAR.  you'll need to recompile for this system
[22:19] <alienseer23> indeed
[22:19] <cclausen> I had to do that with the Oracle oci8 driver
[22:19] <alienseer23> I do remember
[22:19] <alienseer23> where is that to comment out tho? I am feeling a bit like a blind man looking at that...I see nothing
[22:20] <cclausen> cd /etc/php5/conf.d
[22:20] <cclausen> grep pcntl.so *
[22:20] <cclausen> (I assume anyway)
[22:20] <alienseer23> lemme go check that out
[22:20] <cclausen> there is likely an extention=/path/to/file somewhere that loads that module
[22:21] <lamont> marshall: /etc/mailname or /etc/postfix/main.cf
[22:23] <alienseer23> now I have to try and remember why i needed that extention to begin with
[22:23] <alienseer23> heh, notes...must take notes
[22:27] <alienseer23> cclausen: ok, no more php warnings when I do the php5 -v command, but I still get the tt-rss error
[22:28] <cclausen> alienseer23: can you run that specific script at the command line?
[22:28] <alienseer23> i don't think so
[22:29] <cclausen> alienseer23: can you use wget -S to pull down the erroring url?  and pastebin it?
[22:32] <alienseer23> http://paste.ubuntu.com/426087/     tt-rss spits this out as well
[22:32] <cclausen> hmm
[22:33] <cclausen> alienseer23: maybe some internal problem with it?  Does it need that other php module you commented out?  Did you try reinstalling it? or checking on the config?  writing a simple test for it?
[22:34] <jherrlin> hallo! i have a small problem with a full disk (/dev/sda1), i havnt put any files in  that partition, but i have deleted alot of files from external disks, i guess that all the files have moved to the trashbin. but i cant find it, iam running Ubuntu Server 8.04. my googlesearches dont give me much of help
[22:34] <alienseer23> i did a reinstall, completely went over the config file.... the other php module was for another project entirely, and I tried some simple tests, but nothing gives me anything different than what is in the pastbin
[22:35] <guntbert> !trash
[22:35] <guntbert> alienseer23: ^
[22:36] <alienseer23> yeah
[22:37] <alienseer23> may have found something in  tail /var/log/apache2/error.log
[22:37] <jherrlin> thx alot ubotto
[22:41] <electro_> what you run dpkg -l , what does the "rc" at the beginning mean?
[22:43] <cclausen> electro_: means removed / uninstalled but that the config files are probably still there (e.g. package was not purged)
[22:43] <electro_> how do you purge it?
[22:43] <lil_cain> apt-get remove --purge <file>
[22:43] <lil_cain> (assuming you have apt)
[22:43] <cclausen> that won't work if its already removed though
[22:44] <cclausen> or at least it didn't work before maybe it does now
[22:45] <lil_cain> hmm. That is true
[22:45] <lil_cain> try dpkg purge <package> ?
[22:53] <MTecknology> I have dnsmasq set to enforce but when I startup the system dnsmasq runs unconfined...
[22:53] <MTecknology> any ideas if this is just a bug?
[23:30] <osmosis> I have a ubuntu hardy KVM libvirt server. When I run win2003 guest instances with multiple CPU cores, it seems to be causing the node to fail and need a hard reboot.  Does this sound familiar to anyone?
[23:30] <cclausen> osmosis: are you changing this on win2k3 systems that were running on one cpu?
[23:30] <cclausen> osmosis: or are these fresh installs?
[23:31] <osmosis> cclausen, it could very well have been changed after install.
[23:32] <osmosis> cclausen, a bit funny that it takes out the entire node server though.
[23:32] <osmosis> cclausen, im forcing all win2k3 instance to 1 vcpu for now.
[23:33] <cclausen> osmosis: sounds like you need to change the HAL in the windows systems
[23:33] <osmosis> cclausen, i could see that making sense if the win2k3 wasnt recognizing multiple cpus ... but my main concern is stopping the node from going down.
[23:34] <cclausen> osmosis: look in device manager.  are the systesm using the ACPI HAL or the "ACPI Multiprocessor" HAL?
[23:35] <cclausen> osmosis: or just use a single cpu and live with it
[23:35] <osmosis> cclausen, well..im adminstering the node and dont have Admin access to my clients guest instances. So i can easily answer that.
[23:36] <cclausen> ah, ok
[23:36] <osmosis> cclausen, is there a known instability ?
[23:36] <osmosis> meant...can't easily answer that.
[23:36] <cclausen> osmosis: windows 2003 is known to have problems if you reconfigure number of CPUs on an installed system
[23:37] <osmosis> cclausen, by problems do you mean things that could take out the whole node as well ?
[23:37] <osmosis> cclausen, my best hope right now is to migrate off hardy and move to a clean install on different hardware with lucid.
[23:39] <osmosis> maybe related to      <acpi/>  <apic/>   <pae/>    settings as well
[23:40] <cclausen> yep, could be
[23:56] <alienseer231> when attempting to click on links firefox is attempting to download the index.php file instead of going to the appropriate pages, how do I remedy this?
[23:57] <cclausen> alienseer231: use wget -S is the content type set to text/html ?
[23:58] <alienseer231> cclausen: where do I check that?