[00:16] <Jeeves_Moss> is there a image program that'll make hashes out of image files, then use the hash to compair them?  I'm looking for a way of looking for duplicates in ~200,000 images and removing them
[00:20] <SpamapS> find . -type f -print0 | xargs -0 md5sum | sort | uniq -w32 -d --all-repeated=separate | cut -c35-
[00:20] <SpamapS> from http://ubuntuforums.org/archive/index.php/t-97701.html
[00:23] <SpamapS> its unfortunate though..
[00:23] <SpamapS> that it md5's the whole image
[00:23] <SpamapS> would be cooler if you could just find all of the files with the same # of bytes, and then md5 only those.
[00:24] <SpamapS> Jeeves_Moss: thats parallelizable too.. add  -P2 to the xargs and it will run two md5's at a time. :)
[00:24] <Jeeves_Moss> SpamapS, this will be a scrape site eventually.  I'd like to be able to make a hash that's semi smart,toss it into a database, then when new files are d-loaded, it hases them, then compares them
[00:26] <SpamapS> well thats pretty specialized ;)
[00:26] <SpamapS> but very easy
[00:26] <SpamapS> if you're already writing code
[00:27] <SpamapS> Jeeves_Moss: you might want to use a collapsing hash algorithm then. size first, then crc32, then hash
[00:27] <Jeeves_Moss> SpamapS, lol, yea.  the "code" will be a cheap/dirty (kinda like my ex-wife) script that scrapes a set of URLs to pull the files to the local box.  Then hash them, and do a compairison of what's in the databse.
[00:28] <Jeeves_Moss> SpamapS, if it finds comething clise, it'll flag it for human attention.
[00:28] <SpamapS> "close" is a pretty tough thing for a computer to determine. ;)
[00:28] <Jeeves_Moss> SpamapS, eventually, I'm going to figure out how to make the script run times to be dynamic so they'll throtle their scrape runs baised on the number of new pics posted
[00:28] <SpamapS> Jeeves_Moss: good luck on your porn search.. I have to be going. ;)
[00:29] <Jeeves_Moss> SpamapS,  lol, how'd you know it was porn?  LOL
[00:29]  * SpamapS just wasn't going to admit that he already wrote the same thing for the same purposes
[00:29] <SpamapS> ok.. I realy do have to go
[00:29]  * SpamapS runs off
[00:29] <Jeeves_Moss> SpamapS, if you want, e-mail it to me.
[01:23] <mathiaz> hggdh: hi!
[01:23] <mathiaz> hggdh: what's the status of maverick-i386 UEC testing?
[01:24] <mathiaz> hggdh: http://iso.qa.ubuntu.com/qatracker/test/4609 <- UEC install (minimal topology) is not started yet
[01:34] <MrDowntempo> Hi
[01:35] <MrDowntempo> I'm trying to install ubuntu-server on a machine with no swap. Is this possible or simply the acme of foolishness?
[01:36] <Patrickdk> how much ram will you have?
[01:36] <Patrickdk> and what will it be doing?
[01:36] <Patrickdk> it's not really foolish
[01:37] <Patrickdk> but you just have to be careful, cause oom will attempt to protect the system, and won't care what it kills (like ssh)
[01:37] <MrDowntempo> I should clarify. I want to install the OS on an SSD. I have a 3TB raid 5 hardware Raid card
[01:38] <MrDowntempo> I have a gig and a half of RAM
[01:38] <MrDowntempo> But should be able to up it to 3gig soon
[01:38] <Patrickdk> I don't see how that clairifies anything
[01:39] <MrDowntempo> It will be a fileserver local and online
[01:39] <MrDowntempo> Well u usually don't want swap on SSDs or raid
[01:39] <Patrickdk> what kind of fileserver? samba? ftp? apache? lighttpd? nginx?
[01:39] <MrDowntempo> Samba and FTP
[01:39] <Patrickdk> heh? swap on raid is awsome :)
[01:39] <MrDowntempo> Lol
[01:39] <Patrickdk> swap on ssd is great also
[01:40] <Patrickdk> life of ssd with swap, depends on how much use change swap pages
[01:40] <Patrickdk> I never payed attention to samba ram usage, other than it never mattered to me
[01:41] <Patrickdk> but I would say, as long as you keep good amounts of cache/buffer space you should be fine
[01:41] <MrDowntempo> Should I just put a swap partition on the SSD and be done with it?
[01:41] <Patrickdk> but it only takes a program to go nuts once, to cause issues
[01:41] <Patrickdk> I would be tempted to put a swap partition on it, and set swappiness to 0
[01:41] <MrDowntempo> The install always fails when it sees no RAID
[01:42] <MrDowntempo> Another option would be to pick up a small fast HDD. Like the old 74gb velociraptor and use that in its entirety
[01:43] <MrDowntempo> But that means spendin mo money. And I donwanna
[01:45] <Patrickdk> I was looking at a pcie ssd solution for my new server
[01:45] <Patrickdk> but the ibis just came out, and that looks interesting as hell
[01:45] <MrDowntempo> I'm sorry it fails when it sees no swap
[01:45] <MrDowntempo> Not raid
[01:46] <Patrickdk> what fails?
[01:46] <Patrickdk> I've installed ubuntu server without swap before
[01:46] <MrDowntempo> The install says it can't continue without swap
[01:46] <MrDowntempo> Hmmmmnnnrrggh
[01:46] <Patrickdk> you using a cd?
[01:46] <Patrickdk> I always pxe boot into the installer, might be different
[01:46] <MrDowntempo> Actually USB
[01:53] <ScottK> mathiaz: Since it seems puppet is the main user of Ruby in main, would you please have a look at Bug 638164 and make a recommendation?
[02:17] <mathiaz> ScottK: I'd ask the security team about the sync
[02:17] <|rt|> anyone know why the server installation cd uses the desktop kernel?  Makes dealing with supplying a driver for a disk controller a total pita
[02:17] <mathiaz> ScottK: I don't think we'd need to sync from Debian given where we are in the release cycle
[02:18] <mathiaz> ScottK: I don't have enough knowledge about ruby to evaluate all the changes introduced (which are minor BTW)
[02:18] <|rt|> b/c you need one driver built for the install process and another to get the installed system to survive a reboot
[02:18] <mathiaz> ScottK: so I'd seek the security team opinion on the bug
[02:48] <pmatulis> for network installs using the alternate installation, where do i get the kernel and initrd for tftp booting?
[02:53] <zul> pmatulis: the same directory where your image is
[02:53] <zul> pmatulis: or you should be able to get it from the alternate cd
[02:55] <pmatulis> zul: yeah, from the cd.  but in the online archive there is just
[02:55] <pmatulis> zul: http://archive.ubuntu.com/ubuntu/dists/lucid/main/installer-amd64/current/images/netboot/ubuntu-installer/amd64/
[02:56] <pmatulis> zul: so above is for server?
[02:56] <zul> i think so
[02:56] <twb> pmatulis: those d-i images don't understand the distinction between "desktop" and "server"
[02:57] <twb> That distinction basically boils down to different preseed files (and different packages in the on-CD pool).
[02:57] <pmatulis> twb: that's what i don't understand
[02:57] <pmatulis> twb: if i'm setting up netbooting and i want to use the "alternate cd" installation, what files do i use?
[02:57] <twb> In other words, if you only use the kernel and ramdisk, it won't give you any server- or desktop-specific questions.
[02:58] <twb> What you probably want to do is grab the preseed file from the CD and put it on an HTTP server
[02:58] <twb> Then when you bootstrap the netinst kernel, pass it url=http://path/to/preseed/file
[02:59] <twb> This information is mainly explained in the installation-guide-<arch> appendices, which you can apt-get install.
[02:59] <pmatulis> twb: so with netbooting, the install process always looks the same?
[02:59] <pmatulis> twb: it's just the preseed that determines the outcome?
[03:00] <twb> From my own experience, basically the only difference is 1) different tasksel choices; and 2) language packages are installed by default.
[03:00] <twb> Right, it's just the preseed file
[03:00] <twb> I suggest you read the preseed files of the server and alternate CDs to get an idea of the differences
[03:00] <pmatulis> twb: and if i don't use a preseed file?  what do i end up with?  :)
[03:00] <twb> If you don't use a preseed file you get12:00 <twb> From my own experience, basically the only difference is 1) different tasksel choices; and 2) language packages are installed by default.
[03:01] <twb> With a preseed file, behaviour should be identical to the CD you got the file from.  The only difference would be that it asks about network configuration before udebs (in expert mode).
[03:02] <pmatulis> twb: alright, that clarifies things a bit
[03:02] <pmatulis> twb: need to try
[03:04] <pmatulis> twb: thanks
[03:15]  * twb knows all!  Tell you friends!
[03:17] <hggdh> mathiaz: My bad, then, I ran it earlier today. Will update the entry
[03:18] <mathiaz> hggdh: great - so the test ran
[03:18] <mathiaz> hggdh: it's just the result that wasn't reported to iso.qa.?
[03:18] <hggdh> oh
[03:18] <hggdh> mathiaz: sorry, i386 was not run, will do it now
[03:22] <hggdh> mathiaz: but I do expect it to run, libvirt was fixed
[03:23] <mathiaz> hggdh: glad to hear that
[03:23] <mathiaz> hggdh: what was the issue with libvirt?
[03:23] <mathiaz> hggdh: was this fixed for RC?
[03:23] <hggdh> mathiaz: it was missing an #include, which caused a failure only on i386
[03:23] <hggdh> this was fixed a week ago, IIRC, by Jamie
[03:24] <mathiaz> hggdh: ah ok
[03:24] <hggdh> but doing it now. Sorry for that
[03:25] <hggdh> mathiaz: BTW -- care to get together (Ara, Ev, you, JamesP, and myself) on UDS to discuss Hudson?
[03:25] <mathiaz> hggdh: not at all - we should definetely meet up
[03:26] <hggdh> thank you
[03:26] <mathiaz> hggdh: I'm currently in the process of setting a Hudson master in EC2
[03:26] <mathiaz> hggdh: so that you can have a look at how things are presented later this week
[03:26] <mathiaz> hggdh: we'll probably test the final maverick -server iso with it
[03:29] <hggdh> mathiaz: cool! Please point me to the right place when it goes live
[03:29] <mathiaz> hggdh: will do!
[03:30] <ScottK> mathiaz: Thanks
[03:54] <w0_> hi.  I installed apache2 and it loads the default page when I try to connect.  when I type 'ps' there's like two things, not including apache. for some reason, 'apache2ctl -k stop' looks at 127.0.1.1 instead of 127.0.0.1.  how can I stop apache2, and how can I keep it from starting at bootup?  much thx
[03:58] <^slacker^> w0_ : to disable services at bootup you can use the update-rc.d tool to remove it from startup
[03:59] <^slacker^> w0_ example: # update-rc.d -f apache2 remove
[03:59] <twb> s/remove/disable/, surely
[04:00] <twb> Except it's probably an upstart job in 10.04, so you need to do something horrible instead
[04:04] <smoser> mathiaz, ping
[04:04] <mathiaz> smoser: hi!
[04:05] <smoser> you're always so chipper
[04:05] <smoser> :)
[04:05] <smoser> i think bug 651698 is more likely related to
[04:05] <smoser> $ ssh-import-lp-id james-page
[04:05] <smoser> WARNING: Invalid keys at [https://launchpad.net/~james-page/+sshkeys]
[04:05] <mathiaz> smoser: hm - ok
[04:19] <hggdh> mathiaz: yay, euca i386 works now :-)
[04:19] <mathiaz> hggdh: \o/
[04:19] <mathiaz> hggdh: thanks for testing this out
[04:20] <hggdh> mathiaz: sorry for having forgotten the i386... I marked it done (and also tested some others, as a consequence)
[04:20] <smoser> mathiaz, i think the patch for ssh-import-id is http://paste.ubuntu.com/502995/
[04:20] <smoser> but the easy fix is to tell james not to use windows
[04:23] <hggdh> heh
[04:27] <hggdh> mathiaz: the only one left is JeOS under ESX, which I cannot do. I also (of course) tested the i386 UEC images
[04:28] <hggdh> and amd64...
[04:28] <mathiaz> hggdh: great
[04:28] <mathiaz> hggdh: usually someone in QA does JeOS under ESX
[04:29] <mathiaz> hggdh: AFAICT noone in the server team has acccess to an ESX server
[04:29] <hggdh> mathiaz: well, I am QA ;-) and I do not know of any ESX server, so must be somebody else
[04:29]  * mathiaz nods
[04:29] <mathiaz> hggdh: may be ask marjo?
[04:30] <hggdh> hum. Will see if he is still logged in
[04:31] <hggdh> he is online, and shot a Q to him. Let's now wait and see if he is also awake
[04:33] <mathiaz> smoser: great - the patch fixes the issue - thnaks
[04:35] <hggdh> mathiaz: no response from marjo, I will check with the QA folks tomorrow morning
[04:36] <mathiaz> hggdh: thanks for helping out!
[04:36] <hggdh> yw
[08:26] <Ichat> [question] is there some way to  upgrade linux (software) raid  from  example   3x 1tb  to    5x  1tb  (without lost data)
[08:27] <qman__> yes, I have done it before
[08:27] <qman__> assuming raid 5?
[08:27] <Ichat> yes raid 5
[08:27] <qman__> yeah, you can simply add the disks to the raid, then expand your filesystem
[08:27] <qman__> if you're using LVM, you have to expand that too
[08:29] <qman__> I don't remember the exact procedure off the top of my head, but it's certainly possible and documented
[08:29] <Ichat> im not using lvm because i couldn't get it to work in hardy - im wanting to upgrade soon, and  add diskspace also i didn't know that  soft-raid was that flexible...  i asume that when i 'add those disks  that the array will  get 'degraded'  for a while?
[08:30] <qman__> it will have to rebuild
[08:30] <Ichat> i see
[08:30] <qman__> which with that much data could take a considerable amount of time
[08:30] <qman__> I had a 6x500GB array, and added two more disks
[08:30] <qman__> with some tweaking to allow it to use the full bandwidth available, it took about a day and a half to rebuild
[08:31] <Ichat> what cpu ?
[08:31] <qman__> CPU isn't really important, it's disks and disk controllers
[08:31] <qman__> but it was an athlon 64 3500+
[08:32] <qman__> mine were mostly limited because I had 4 disks on the PCI bus and 4 on the onboard controller
[08:32] <qman__> as opposed to, for example, a PCIe x4 or x8 controller
[08:33] <qman__> CPU becomes more important with raid 6 though
[08:33] <Ichat> hmmz.  what i ment was,  -  for exaample  a pent 4  will have considerably lower  i/o  than a  sock 940 / am2 opteron  or a    s755 core2 duo
[08:34] <Ichat> btw mine is a  939 opteron 165 (dual 1,8)  so it wont be noticebly faster than what you did i fear :P
[08:34] <qman__> ah, yeah
[08:34] <qman__> good chip though, I have two of them
[08:34] <qman__> one overclocked to 2.7GHz
[08:34] <Ichat> mine runs a 2,3  i cant unlock it :( and couldn't get it stable on higher speeds
[08:35] <Ichat> but most of the time  Cool-N-Quet runs it down to 1,8 again :
[08:35] <qman__> they will take some punishment, that one I have at 1.475v, has hit 75C under load before, it takes it all
[08:36] <qman__> but you need a good board and RAM to get it that high
[08:36] <qman__> since it's only a x9 multiplier
[08:36] <qman__> mine's at 300 fsb
[08:37] <qman__> but anyway, my file server is also a 939
[08:37] <qman__> expect the whole process to take your server out of commission for 2-3 days
[08:38] <Ichat> what pc case (that helps if its a good one)  i recently got a   nzxt thempest    (2x 140mm top)   1x 120mm rear   3x 120mm front
[08:38] <qman__> for the overclocked chip, I used to have a modified $20 cheapo with some 120mm fans
[08:38] <Ichat> i couldn't get my cpu  over  53C  even when room temp was over  31C
[08:38] <qman__> now it's in a 40mm ammo box
[08:38] <qman__> with 3 80mm fans
[08:40] <Ichat> that case ROCK and gets me  12 hdd's  if only i hade € for sata controlers
[08:40] <qman__> pictures here, if you're interested http://qman.strangled.net:8080/pics/ammo-box-pc/
[08:40] <qman__> for the file server, I've got a cooler master 590
[08:42] <Ichat> i wanted an antec 1200  but than this case came allong, only thing was,  that i got damaged in transport  so no on/of button  leds or usb stuff... just the case  thats why i got this   $100 case for less than  $40
[08:42] <qman__> good deal
[08:42] <qman__> my cooler master was about $65
[08:42] <Ichat> i just hade to 'rip appart an old atx case that i had witch was  a sweex ...
[08:42] <qman__> but I had to buy some fans and bay adapters
[08:43] <qman__> my first file server was a heavily modified emachines micro ATX case
[08:43] <Ichat> i just  removed the sweex's  front pannel and  glued that to the tempest case :P
[08:43] <qman__> after drilling some holes and using some pieces of erector set, I managed to cram 7 hard drives in it
[08:43] <eagles0513875> hey guys i am trying to install some public ssh keys where do i put the public keys?
[08:44] <Ichat> btw the mod went terribly bad - and looks BUD-ugly but i does function perfectly so i dont really care
[08:45] <qman__> eagles0513875, I only know how to use ssh-copy-id, not sure if that will do what you want
[08:45] <eagles0513875> well i generated this on a windows machine just not sure where to put the public key once uploaded via sftp
[08:46] <eagles0513875> do i put it in /home/.ssh/?
[08:46] <qman__> I think ssh-copy-id creates ~/.ssh/id_rsa.pub
[08:46] <qman__> so I'd try that
[08:47] <qman__> no, that's wrong
[08:47] <qman__> that's where it generates it in the first place
[08:47] <eagles0513875> ok
[08:47] <eagles0513875> im reading the !ssh
[08:47] <eagles0513875> !ssh
[08:47] <eagles0513875> documentation the first link and its saying /home/.ssh of that user
[08:47] <eagles0513875> does that seem right to you?
[08:48] <qman__> looks like it adds it to ~/.ssh/authorized_keys
[08:48] <qman__> reading through man ssh-copy-id
[08:48] <eagles0513875> qman__: wrong
[08:48] <eagles0513875> thats for Encrypted Home Directory according to the link
[08:49] <eagles0513875> the authorized keys if you read above troubleshooting it says otherwise
[08:49] <qman__> ~/.ssh should have noting to do, directly, with encrypted home
[08:49] <eagles0513875> ok
[08:49] <qman__> though ssh key auth wouldn't work normally if you have encrypted home
[08:50] <qman__> you'd have to put your keys in a different location
[08:50] <eagles0513875> ok
[08:50] <eagles0513875> would ssh-copy-id work even if the key wasnt generated on the machien you want to run it on
[08:50] <qman__> I see what it's saying
[08:50] <qman__> if you don't have encrypted home, add it to ~/.ssh/authorized_keys
[08:51] <qman__> if you have encrypted home, add it to /etc/<username>/authorized_keys
[08:51] <eagles0513875> ahh ok so i just need to add it to the authorized keys file in my home dir
[08:51] <qman__> and yes, you can manually specify a key file on the command line to ssh-copy-id
[08:52] <eagles0513875> ok i can just upload to my home folder then copy it in?
[08:52] <eagles0513875> using ssh sopy
[08:52] <qman__> yes
[08:52] <eagles0513875> ok kool :) thanks
[08:59] <kinygos> hi...i have a remote server running 10.04LTS, it has a second physical disk which was not partitioned or mounted, and i'm trying to install RAID1...there is no data on the server, but i'm not able to control the OS installation (it's done by some imaging software the ISP has)...
[09:00] <kinygos> i've created my raid on the second physical disk, and updated the md config...
[09:01] <kinygos> i'm now trying to copy the contents of the first disk (the one with the OS that is currently running) onto the RAID
[09:01] <kinygos> i started a dd command last night, but it failed (my mistake completely, the RAID is logically smaller than the physical disk)
[09:02] <kinygos> is there another way i can copy all the data from the mounted partitions on the first physical disk to my unmounted RAID device?
[09:05] <qman__> you're not going to be able to do a direct image, you'd either have to use something like partimage which removes the free space, or just do a simple file copy
[09:06] <qman__> I wouldn't recommend doing a file copy online, suggest booting to a live (or otherwise alternative) environment first
[09:07] <spenser> Hi, is it possible to transform a private CA into a public CA by having the private CA signed by the public CA?
[09:08] <kinygos> qman__: i know...i do have remote kvm access through a lights-out board, so can start again if i bugger it up
[09:08] <twb> kinygos: if you don't need the result to be mountable, you can just write something like a cpio or tar archive DIRECTLY onto the RAID array
[09:08] <twb> e.g. find /home | cpio --create >/dev/md0
[09:09] <twb> If you need it to be mountable but don't want to mount the destination filesystem in advance, you could try mksquashfs or genext3fs
[09:09] <kinygos> twb: my plan is to reboot the server, but boot from the second disk (thereby starting the RAID), then add the original disk to the RAID
[09:09] <twb> Shrug
[09:10] <qman__> it's a difficult task any way you look at it
[09:10] <twb> FSVO difficult = tedious
[09:12] <_ruben> mount the target partition at some place, then use cp with a ton of options to copy the data (cp -a might be enough, don't recall)
[09:13] <coder> hi, I've just noticed a process named bond0 in my 10.04.1 server, which I don't see in other 'bonded' servers with, ie, centos
[09:14] <kinygos> awesome, thanks twb, _ruben, qman__
[09:14] <coder> is that normal?
[09:15] <coder> ok, it seems normal, also seen on centos
[09:15] <coder> forget it
[09:18] <twb> _ruben: cp -a will suffice for most everything -- the main exceptions will be /proc and possibly hard links
[09:19] <twb> When I'm copying from an active root filesystem, I usually lvm snapshot it then mount and copy that.  That avoids /proc nastiness, even when the last sysadmin was silly and did things like /var/lib/named/proc
[09:19] <twb> */var/lib/named/chroot/proc, I mean
[09:21] <Riddell> anyone able to sign off on the ubuntu server RC release for lack of jiboumans?
[09:25] <_ruben> twb: there's some directories that need exclusion indeed
[09:26] <kinygos> _ruben, twb: noob question, if i mount my raid at /raid, is there an easy way to exclude it from the cp?
[09:27] <_ruben> kinygos: i tend to go at it the other way around: just specify the dirs you *do* want to copy .. else do smth like find / | grep -v dir1 | grep -v dir2 | xargs cp --target-dir=/raid
[09:27] <_ruben> unless cp doesnt have --target-dir, i know mv does
[09:28] <kinygos> ok, thanks :)
[09:28] <twb> kinygos: you usually would use -xdev in whatever guise the command in question provides it
[10:03] <spenser> Not trying to be impatient but its been about an hour so I'll ask the question again.
[10:03] <spenser> Hi, is it possible to transform a private CA into a public CA by having the private CA signed by the public CA?
[10:14] <qman__> spenser, not really sure, but there's nothing inherently private or public about a CA, the only differentiation is whether a "trusted" company has signed your certificate
[10:14] <qman__> and the only effect that has is that web browsers will trust it by default
[10:17] <twb> the ca infrastructure is a damn rort
[10:19] <spenser> _qman: Specifically say I have a self signed CA and I want to make it signed by Verisign is that possible?
[10:32] <kinygos> i have a partition listed in fdisk -l that doesn't appear to be mounted according to /etc/fstab...the System on this partition is Extended, whereas all the others are either Linux swap / Solaris (one of those) or simply Linux....
[10:32] <kinygos> can that Extended partition be deleted?  i'm struggling with transferring data to my new RAID so want to reduce the amount to copy with dd
[10:34] <kinygos> bugger...i've just noticed that it sits under all other partitions (going by the Start and End values for all partitions)
[10:34] <kinygos> so, different question, is there a command to reduce the size of a partition?  it's mostly empty
[10:40] <kinygos> argh!! i can't seem to install gparted...apt-get fails to fetch many files :(
[10:44] <alvin> kinygos: Try a partedmagic liveCD/USB
[10:45] <kinygos> alvin: i think i've buggered the install up with my various attempts to get it RAID'd...going to rebuild and start again
[10:45] <kinygos> alvin: apt can't read the archive files in /var
[10:45] <alvin> Could be. apt-get failing is pretty new to me :-)
[10:45] <kinygos> arvin: thanks though :)
[10:45] <kinygos> *alvin even
[10:46] <alvin> Why attempts to create a RAID? What's going wrong?
[10:48] <kinygos> alvin: it's a remote server, with no access to external media..server build is automated by some imaging system at the ISP end so i can't configure RAID during install...
[10:48] <kinygos> alvin: after imaging, the server has a second physical drive with no partitions and unmounted
[10:49] <alvin> kinygos: I see. Well, I haven't done it myself, but there are tutorials to create a RAID on such systems. It should work.
[10:49] <kinygos> alvin: my plan is to create RAID on that disk, transfer data from first drive to it, install grub on second drive, then reboot onto the second disk (the RAID) and add the first disk to the RAID
[10:50] <alvin> Transferring data is probably not needed. The rebuild should take care of that. Let me see if I can find something
[10:51] <svnitcloud> need some help on opennebula express
[10:51] <svnitcloud> i have some bridge issues
[10:51] <svnitcloud> and onehost gives STAT err
[10:51] <svnitcloud> what do i do ?
[10:53] <kinygos> alvin: thank you very much :)
[10:56] <alvin> This one looks easy. http://www.cyberciti.biz/faq/linux-creating-software-raid-one-arrays/ There's nothing about grub there, so be careful.
[10:57] <kinygos> alvin: wow, thanks :) can't believe i didn't come across that...
[10:57] <svnitcloud> any help for open nebula ?
[10:57] <alvin> kinygos: As I said, I haven't done this myself. Maybe you need to adapt /etc/fstab. It won't hurt to do a sudo dpkg-reconfigure grub-pc afterwards too
[10:58] <kinygos> alvin: well mdadm did everything i expected, so i'm hopeful :)
[10:58] <kinygos> alvin: thanks for the help...i really appreciate your time
[10:59] <alvin> kinygos: You're welcome
[11:07] <Riddell> smoser: what has to happen to get the ec2 images signed off and published?
[11:24] <jetole> Hey guys. I am auditing my file server to see which iSCSI and NFS is not being used however when it comes to NFS, I don't know how. Does anyone know how I can view a history of which hosts connected to the NFS server, when they did, what NFS mounts they connected to, etc
[11:30] <alvin> jetole: No, but showmount -a will give you some information about connected (or not properly disconnected) hosts
[12:34] <smoser> Riddell, they're ready to go
[12:35] <Thelz> Hello. My ubuntu server installation got corrupted. Is it possible to install ubuntu on a different harddisk and reach the files that were in the original installation from the new installation? (I'm only after the files that were stored remotely using the Samba File Server)
[12:36] <Riddell> smoser: great, so how does that happen?
[12:38] <reggi>  by the way 192.168.0.50 is my linux laptop which i'm performing the ping f
[12:38] <reggi> HEY folks I have a strange problem. When I ping 192.168.0.1 (the gateway) i get the following message http://paste.ubuntu.com/503187/  by the way 192.168.0.50 is my linux laptop which i'm performing the ping from.
[12:41] <reggi> oCean, i'm posting in this channel incase my question is a network related answer.
[12:42] <oCean> one channel at a time for the same question
[12:43] <reggi> oCean, i've asked a question in #ubuntu only to be told to post in this channel due to it's 'network' nature.
[12:43] <oCean> anyway, it means that the ip for the default gw is unreachable
[12:44] <oCean> are you sure the .1 is up, and that the .50 is in the same network (netmask)?
[12:44] <reggi> oCean, it's pointing back to this computer
[12:44] <reggi> my laptop running ubuntu is .50 and the gateway is .1
[12:45] <reggi> by the way, my laptop was configured as a dhcp server
[12:46] <reggi> is there some kind of file that has a map of the network on my laptop?
[12:46] <oCean> so you have multiple interfaces? One being .50, the other one .1
[12:46] <reggi> oh wait. let me explain.
[12:47] <reggi> my linux laptop was setup as the gateway. it had dhcp3 installed and was sharing a 3G mobile internet connection. but now I have the internet connection and gateway from another computer on the network.
[12:48] <reggi> my linux laptop has eth0 set to static (192.168.0.50) and my gateway (the other computer with internet connection) is 192.168.0.1
[12:48] <reggi> when I perform a ping 192.168.0.1 from my laptop i get the output
[12:48] <reggi> as shown here: http://paste.ubuntu.com/503187/
[12:49] <oCean> which brings me back to my original quesion: are you sure the .1 is up, and that the .50 is in the same network (netmask)?
[12:49] <oCean> at which you said it's pointing back to this computer.
[12:49] <oCean> Pointing back??
[12:49] <reggi> hmm good question. let me check.
[12:50] <reggi> ok here is the contents of my ifconfig from my laptop: http://paste.ubuntu.com/503193/
[12:52] <oCean> ok, the /24 netmask is ok, it includes the .1
[12:54] <oCean> 2nd step is to check if there are (correct) routes configured. Check route -n
[12:54] <Thelz> ./24 is the first 3 numbers fixed and the 4th number not fixed, right?
[12:55] <oCean> actually, they're called "octets": an ipv4 address is a group of four octets
[12:57] <reggi> hmm
[12:57] <reggi> yes
[12:58] <reggi> i think you've hit the nail on the head oCean - it's about the routing
[12:58] <reggi> i'll check the route
[12:58] <oCean> Thelz: if you're interested, you can see here how it works: http://jodies.de/ipcalc?host=192.168.0.1. Just play around with the netmask, and see the bits in the octets change :)
[12:58] <reggi> AHUH!
[12:58] <reggi> oCean, how do I edit the route data?
[12:59] <Thelz> oCean: I know. What I don't understand is why IP's aren't written in hex. That would've made stuff a lot easier
[12:59] <oCean> reggi: using the "route" commands.
[12:59] <Thelz> MACs are written in hex
[12:59] <reggi> oh really?
[12:59] <reggi> damn
[13:00] <reggi> so oCean does that mean the route automatically populates itself?
[13:00] <oCean> reggi: partially, when bringing up an interface
[13:00] <oCean> you can add/edit static routes
[13:01] <reggi> i think there are a couple of errors in my route http://paste.ubuntu.com/503198/
[13:01] <reggi> see line 5?
[13:01] <oCean> like "route add -net blah netmask blah gw blah" etc
[13:02] <oCean> nvm line 5
[13:02] <oCean> that's a special/reserved address
[13:03] <oCean> but you seem to have 2 default routes, the lines starting with 0.0.0.0
[13:04] <reggi> ya
[13:04] <oCean> you could remove the default route line on the eth0 interface, and since line 4 is already there
[13:04] <reggi> interesting
[13:04] <reggi> and this is removed by a command line right?
[13:05] <oCean> route del -net default gw 192.168.0.1 should do it
[13:05] <oCean> route del -net default gw 192.168.0.1 dev eth0 <-- or that, with the added 'dev', not sure
[13:06] <reggi> ya i'll try the second option
[13:06] <oCean> oh wait, line 4 should be changed.
[13:06] <oCean> Since that line still says 'everything to 192.168.0.0/24 should go through default gw'
[13:07] <reggi> ahah
[13:08] <oCean> So, following the previous example, you have to run commands like this:
[13:08] <oCean> route del -net 192.168.0.0/24 gw 0.0.0.0
[13:08] <oCean> and add
[13:08] <oCean> route add -net 192.168.0.0/24 gw 192.168.0.1
[13:08] <oCean> capice?
[13:09] <reggi> i get it
[13:09] <reggi> nice one oCean
[13:09] <reggi> peter@peter-laptop:~$ sudo route add -net 192.168.0.0/24 gw 192.168.0.1
[13:09] <reggi> SIOCADDRT: No such process
[13:10] <ewook> the interface on that net is the gw, so change gw blabla to dev eth instead.
[13:10] <oCean> that shows that the default gw (.1) is still not reachable
[13:11] <oCean> reggi: yeah, what ewook said
[13:11] <reggi> ahah
[13:11] <oCean> routingtables are always fun. Anyway, I gotto run...
[13:11] <oCean> bye
[13:11] <reggi> man how do you guys know all this stuff!
[13:11] <reggi> thanks for your help oCean
[13:12] <oCean> ok!
[13:12] <ewook> I've done the same thing you're doing.
[13:12] <ewook> learning by doing.
[13:12] <ewook> or, sometimes breaking :P
[13:12] <reggi> i am a long time microsoft user
[13:12] <reggi> desktop technician
[13:13] <reggi> and decided to go cold turkey to linux
[13:13] <reggi> so i can start learning
[13:14] <ewook> cold turkey way is a hard way :)
[13:14] <ewook> hey ivoks :D
[13:15] <reggi> yeah well the first time I tried to go all linux was over5 years ago and I had a windows machine to fall back on...which mean that I kept relying on the microsoft machine
[13:15] <reggi> so the only way was to go cold turkey
[13:16] <ewook> I'm required to run ms os when working..
[13:16] <reggi> yeah I still have xp but only in a virtualbox
[13:19] <reggi> just a sec i have to test my route settings
[13:19] <ewook> I tried to explain I could do the same, but it was a no go :(
[13:20] <_ruben> is it possible to have a dot in the short, non-fqdn, hostname?
[13:21] <_ruben> does it does
[13:22] <_ruben> s/does/guess/
[13:29] <kinygos> so...i'm back to my original state, a remote server (running 9.04 though), no access to external media, with 2 physical disks, only one partitioned and mounted....
[13:30] <kinygos> i've read a lot about partitions, raid and mdadm...and now have an alternative strategy for implementing RAID1 on that server...
[13:31] <pmatulis> kinygos: what happened to all the advice people poured on you?  was it wrong or what?
[13:31] <kinygos> i had a fundamental misunderstanding of partitions
[13:31] <pmatulis> kinygos: explain?
[13:32] <kinygos> ok, the plan was to implement raid on the second disk, copy the data to that disk, make that disk bootable and reboot from it...
[13:32] <kinygos> then add the original disk to the raid
[13:34] <kinygos> my mistake was that the raid was logically smaller than the physical disk, and i went round in circles trying to copy data to it
[13:34] <kinygos> i've done the typical noob thing which is dive in without really truly understanding what is going on under the hood
[13:35] <pmatulis> kinygos: dunno if that's noob thing, it's what most people do to learn
[13:35] <pmatulis> kinygos: not a great idea on an important system however
[13:36] <kinygos> well, the system is going to host my web application...i'm a developer trying to be everything as i'm under a very limited budget
[13:36] <pmatulis> kinygos: you're saying your array wasn't made large enough to contain all the data?
[13:37] <kinygos> well, i was using dd to copy, and it was pointed out to me that it may not work, but i was incredibly foolish and let it run through the night to find it had failed due to insufficient space
[13:37] <kinygos> ironically, the disk is practically empty :)
[13:38] <kinygos> so, i've been learning about partitions, found out what the Extended partition actually means and how it fits into the grand scheme of things...
[13:39] <kinygos> and i've found an option for mdadm --level=mirror
[13:39] <_ruben> using dd is completely wrong to begin with
[13:40] <kinygos> _ruben: i know...that suggestion did not come from anyone here by the way..it was my mistake alone
[13:40] <_ruben> you want a filebased copy, not a clusterbased copy
[13:40] <kinygos> well, what i'm considering now is creating corresponding partitions on my second disk, and then mirroring each in turn...
[13:43] <kinygos> it may not be the most efficient, as i would've prefered my swap partition to be striped
[13:43] <kinygos> brb in 15mins
[13:44] <_ruben> kinygos: striped swap? so when 1 disk fails, you'd lose half your swap, and your system would crash? sounds like a plan!
[13:46] <pmatulis> heh
[13:54] <_ruben> hm, having a dot in the non-fqdn hostname doesnt work well afterall
[13:55] <kinygos> rofl...and that is why i'm a noob
[14:11] <reggi> hello anyone here?
[14:14] <zoopster> no I don't think anyone is around today, reggi
[14:14] <jpds> reggi: There are 282 people in here.
[14:14] <reggi> har har very funny... it was a little quiet in here
[14:14] <reggi> i've stuffed up my computer
[14:14] <reggi> a little...
[14:15] <reggi> i now no longer have eth0 in my network connection manager
[14:15] <reggi> Wired Network has "device not managed"
[14:16] <reggi> does anyone know how to get my eth0 to show in this network manager?
[14:20] <zoopster> reggi: are you running ubuntu desktop?
[14:20] <reggi> yes
[14:21]  * zul cant believe that people are still using feisty
[14:21] <zoopster> reggi: #ubuntu would be better for those types of q's...
[14:23] <reggi> hey zoopster the reason I thought to ask in ubuntu-server is because my network settings are really wrong
[14:23] <reggi> for example, if I ping 192.168.0.1 i get the following error "From 192.168.0.50 icmp_seq=3 Destination Host Unreachable"...192.168.0.50 is this (my laptop) and 192.168.0.1 is my gateway / router.
[14:25] <reggi> i'm hoping to fix this otherwise i'll have to reinstall ubuntu again :(
[14:29] <raubvogel> Does anyone have an idea of why rsyslog in my 10.04 desktop is using quite a lot of cpu:  1118 root      20   0  330m 7844 1016 S  173  0.1 953:13.17 rsyslogd
[14:31] <zoopster> reggi: a reinstall won't fix incorrect settings
[14:31] <reggi> i hear you zoopster
[14:32] <raubvogel> reggi: try #ubuntu-beginners. I am not being condescending. I had a lot of luck there before. In fact I had more answers about creating a package and building my own repository there than in ubuntu-packaging and ubuntu-mirrors
[14:33] <reggi> hmm
[14:33] <reggi> i'll try that raubvogel
[14:33] <reggi> cheers jolks
[14:33] <reggi> folks
[14:34] <raubvogel> The other thing would be to reinstall the network manager itself
[14:34] <raubvogel> doing a --purge
[14:34] <reggi> raubvogel, what is the full "--purge" command?
[14:35] <raubvogel> In a server, first thing that goes away is network manager IMHO
[14:35] <reggi> oh you mean using the synaptic?
[14:35] <raubvogel> something like apt-get remove --purge network-manager or whatever the package is called
[14:36] <raubvogel> synaptic probably would work too; it has something you can click on to say you want it to completely wipe that package. Normal uninstall leaves the old config files
[14:36] <reggi> I would do away with the network manager because I know how to edit my /etc/network/interfaces
[14:36] <reggi> but I use the network manager to connect to the internet using my mobile phone
[14:36] <reggi> using bluetooth.
[14:36] <raubvogel> don't get me wrong; network manager has come a long way. I use it in my own laptop for that very same reason
[14:36] <reggi> you know, like tethering
[14:37] <raubvogel> But it does things that drives me nuts
[14:37] <reggi> unless you know of a way to connect to my mobile phone using bluetooth and then connect to the 3G internet.
[14:38] <raubvogel> The nicest new thing it has IMHO is that you can specify the mac for the AP you are using
[14:39] <reggi> hmmm
[14:39] <reggi> maybe I should manually add it
[14:39] <reggi> brb
[14:46] <kinygos> hi all...i may be able to get an .iso image mounted so i'll be able to rebuild my server and configure the raid at a sensible time...
[14:47] <kinygos> my question is, is it possible to back the whole system up regularly using rsync or something so that i can recover the whole box without having to reinstall the OS again?
[14:50] <Jeeves_> kinygos: Yes, it it
[14:52] <kinygos> Jeeves_: thanks, i'm feeling way out of my depth
[14:54] <reggi> folks i think the problem with my internet is related to my laptop's network device has stopped working
[15:00] <raubvogel> reggi: You can't start it manually?
[15:01] <reggi> raubvogel, with this command? ifconfig eth0 up
[15:01] <reggi> ?
[15:14] <raubvogel> If you are in the wired network, like ifconfig eth0 192.168.1.24 netmask 255.255.255.0 up (guessing an empty IP here)
[15:15] <reggi> there is something wrong with my ifconfig output, see line 19: http://paste.ubuntu.com/503277/
[15:15] <reggi> what is eth0:avahi?
[15:16] <_ruben> start by forgetting ifconfig ever existed
[15:43] <kalkems> I'm a newbee to IRC. How do I communicate it seems so quiet
[15:53] <pmatulis> kalkems: you're doing fine
[15:57] <kalkems> Questions here deal with server problems - I upgraded my ltsp server this morning and was greated with a message that it has deleted grub. (in chroot) This happened when uppdating latest kernel. Will this affect the server it self or only ltsp.
[16:03] <MTecknology> hm... how long is too long for a php script to run?
[16:03] <MTecknology> I fear this one may time out...
[16:13] <hggdh> Daviey: are you going to propose a SRU for 650446?
[16:14] <Daviey> hggdh, i think so.. Really want some feedback from Ng before doing so.
[16:16] <hggdh> Daviey: my single worry is if this increase takes space from the -Xmx
[16:18] <Daviey> Yeah... I don't know if it takes from there - or is additional..
[16:18] <Daviey> perhaps JamesPage knows?
[16:19] <Daviey> JamesPage, Does -XX:MaxPermSize steal from -Xmx, or in addition to?
[16:19] <JamesPage> Daviey: in addition to
[16:19] <Daviey> great, thanks JamesPage
[16:19] <hggdh> then we are good :-) thanks, JamesPage
[16:19] <JamesPage> heap and perm are separate memory areas in the JVM
[16:19] <Daviey> groovy!
[16:20] <JamesPage> hggdh, daviery:have you guys come across jps and jstat for looking at allocation/utilisation?
[16:21] <Ng> Daviey: unless there's a way to make the JVM spit out memory stats the only feedback we're going to be able to come up with really is "it hasn't died for X days"
[16:22] <hggdh> JamesPage: no, I at least have not (but I left Java a looong time ago, at 1.3 I think)
[16:23] <Daviey> JamesPage, no, sorry. :/
[16:23] <Daviey> Ng, Yeah... i think you are probably correct :(...  How many days uptime did it have before issue?
[16:24] <Ng> Daviey: weirdly it was only a couple of weeks
[16:24] <Ng> perhaps some usage pattern was exercising the system in such a way that it maximised the Permanent memory usage
[16:25] <Daviey> Ng, That is pretty concerning :/
[16:25] <Daviey> I wonder if it was being hammered from a user?
[16:25] <JamesPage> Ng, daviey: jstat would help you monitor memory area utilisation in the JVM which might help diagnose the issue
[16:27] <Ng> JamesPage: interesting
[16:27] <SpamapS> JMX enabled apps tend to give lots of stats.
[16:30] <kalkems> quit
[16:30] <overlord_tm> Is there any way to scrool in intramfs shell? Something like <cmd> | less :)
[16:31] <kalkems> sorry thought that was the way to leave:P
[16:32] <hggdh> kalkems: put a '/' in front: '/quit'
[16:32] <kalkems> hggdh: thanx
[16:35] <dan_maran> Having a weird issue with Aptitude updating behind a Sonicwall. This not only is a ubuntu problem but also happens with debian or any apt based distro.  Yum seems to work fine altough that could be due to the yum fastest mirror.  Essentially what happens is and update is started and it all seems fine for anywhere from 10 seconds to minute, speeds of 600KB then out of no where it will just slow to Bytes like now I have an u
[16:35] <dan_maran> pdate running and it is sitting at 1,446B/s. But if I download something on this notebook it will fly.  So I am thinking it is something with the apt settings triggering some kind of Sonicwall "feature" but I have ensured the QoS is OFF.. Any ideas?
[16:39] <SpamapS> dan_maran: your firewall is probably dropping packets that it shouldn't.
[16:39] <hggdh> dan_maran: maybe you are being throttled by your firewall
[16:40] <SpamapS> I've seen firewalls' state tracking go awry and drop retransmit requests and window adjustments.
[16:40] <SpamapS> dan_maran: do you have access to the firewalls' logs?
[16:40] <dan_maran> @spamaps yes and there is not a thing in there about it.
[16:42] <hggdh> I have had that scenario before... I "solved" it by cancelling the update and restarting it multiple times (it would be fast for a while then slow to a crawl)
[16:43] <dan_maran> @hggdh yes, that is the only way that I get around it
[16:43] <dan_maran> here is one of the errors
[16:44] <dan_maran> 1	09/30/2010 11:34:09.432	Notice	Network Access	TCP connection dropped	91.189.88.46, 80, X1	192.0.100.69, 39326, X0, LINUX-01	TCP Port: 39326
[16:45] <kinygos> SpamapS: i should have listened to you regarding the copy for my RAID problem...dd didn't work as there was just a little less space on the RAID :(
[16:46] <kinygos> SpamapS: i've re-imaged the server, created the RAID on the second disk and now trying to mount it so i can do a file copy with cp -a
[16:46] <hggdh> dan_maran: so your firewall dropped it. Too many errors, excessive use of bandwidth, what? You will need to sniff the connection and/or look at your firewall config
[16:48] <dan_maran> @hggdh - Those drops correspond to my cancellation of the downloads
[16:48] <hggdh> oh darn!
[16:48] <hggdh> dan_maran: then you really have to look at the traffic on the session
[16:49] <zoopster> dan_maran: I'd concur...it will likely be a quick resolution or problem discovery with a packet trace
[16:50] <dan_maran> I will try to get on that ASAP, just was hoping someone else already had this issue and new the fix. oh well. Thanks and I will report back if I find it.
[16:50] <_ruben> dan_maran: could also be various tcp "optimization" features wreaking havoc
[16:51] <hggdh> dan_maran: in my case, when I had it, I had no access to the firewall system
[16:51] <dan_maran> @_ruben that is my thought.... But since YUM works fine on RHEL I thought it would be a quick APT fix.. oh well
[16:52] <SpamapS> kinygos: trying to mount it?
[16:52] <SpamapS> kinygos: You said before you had 5 partitions on the disk
[16:53] <SpamapS> kinygos: I'd suggest creating 5 partitions on /dev/md0, and creating a filesystem/swap on each one of them so they match.. /dev/sda1 -> /dev/md0p1
[16:53] <kinygos> SpamapS: ah, yes, my understanding of partitions was fundamentally flawed yesterday...i've read up and hopefully know a bit more now
[16:53] <_ruben> dan_maran: RHEL/Ubuntu might enable different optimizations by default
[16:53] <SpamapS> kinygos: one good idea would be to save the output of 'fdisk -l /dev/sda' so you can see the size of the non-raid partitions and match them closely.
[16:54] <kinygos> SpamapS: my first disk has 2 primary partions, one Linux swap, the other Extended which itself contains 4 logical partitions...
[16:54] <dan_maran> Right on, thanks for the thoughts. I might install yum on the server and see if I can replicate.
[16:54] <SpamapS> dan_maran: I thought there was nothing in there about it.. looks like the TCP connection was dropped for some reason.
[16:54] <kinygos> SpamapS: ok, sounds like a plan, so create the partitions one at a time using parted?
[16:54] <SpamapS> dan_maran: oh, its logging that you stopped it, right
[16:54] <SpamapS> dan_maran: sorry, I'm catching up ;)
[16:55] <dan_maran> @SpamapS - hhaa ok, yeah those of from me killing it. ;)
[16:55] <SpamapS> kinygos: yes
[16:55] <SpamapS> does anybody else find parted harder to use than fdisk? Maybe I'm just set in my ways.
[16:55] <kinygos> can i do it with fdisk?
[16:56] <kinygos> i'm so out of my depth...learning as i'm going along :(
[16:58] <SpamapS> kinygos: fdisk is, IMO, simpler. ;)
[16:58] <SpamapS> but maybe I'm missing the easier parts of parted
[16:58] <SpamapS> I mean, if I have to resize, parted is where I go. ;)
[16:59] <SpamapS> kinygos: either way, what you need to do is create a corresponding partition for each one from the original disk.
[16:59] <SpamapS> kinygos: then create a matching filesystem on each one (probably ext3)
[17:01] <kinygos> SpamapS: excellent, thanks :) it is indeed the ext3 filesystem, so i'm reading up...i came across parted because i wanted to reduce the size of one partition that was taking most of the disk with only 1% space used
[17:07] <nimrod10> how can I extract the content from a deb package ?
[17:08] <jpds> nimrod10: dpkg --extract package.deb /tmp/location
[17:08] <nimrod10> thanks jpds
[17:09] <oCean> nimrod10: please don't crosspost
[17:10] <nimrod10> oCean, ok I won't do that. I imagined that noone will see my question is a sea of over 1000 users on #ubuntu
[17:11] <oCean> nimrod10: ok
[17:17] <SpamapS> kinygos: You probably have one partition that is swap, not ext3.. so make sure you know which one that is.
[17:19] <kinygos> SpamapS: indeed, it's my first one on the first primary partition...i'm in the process of creating it, but am trying to figure out which cylinder it should end on as the RAID characteristics are completely different to the physical disk
[17:19] <kinygos> SpamapS: eg, physical disk has 255 heads, 63 sectors, 30401 cylinders...RAID has 2 heads, 4 sectors, 61049616 cylinders
[17:19] <kinygos> *eg/ie
[17:20] <SpamapS> kinygos: you can just specify a size.
[17:21] <SpamapS> kinygos: make the sizes about the same, should be fine.
[17:22] <overlord_tm> how can i see which packages were updated and when?
[17:35] <mathiaz> Daviey: o/
[17:35] <mathiaz> Daviey: any news on the release front?
[17:43] <jeiworth> say, quick question, if i want to have byobu autostart on a erver machine when logging in via ssh, where do i configure it? ~/.bashrc?
[17:44] <Daviey> mathiaz, None really.. it's looking good.  I told ttx that i would be happy to sign it off - and he agreed.
[17:44] <mathiaz> Daviey: o^5 - good job!
[17:45] <zoopster> jeiworth: I did it from the screen interface (f9) there is an option to set it
[17:45] <raubvogel> #solaris
[17:45] <raubvogel> ooops
[17:45] <jeiworth> zoopster: oh, well, that sounds obvious, i am so used to mangling with config-files ;)
[17:46] <zoopster> jeiworth: well...call me lazy :-D
[17:49] <jeiworth> zoopster: well, that did the trick, thanks!
[17:50] <zoopster> jeiworth: yw
[18:08] <elb0w> is there a 32bit server edition?
[18:08] <RoyK> yes
[18:08] <elb0w> nvm
[18:08] <elb0w> found it
[18:23] <kinygos> SpamapS: i think i've buggered it up (again)...i have my partitions, but i don't understand how to make the first one a swap partition :(
[18:25] <kinygos> SpamapS: still looking online...sorry..i talk too much
[18:32] <SpamapS> kinygos: in fdisk, you change the type to "linux swap"
[18:32] <SpamapS> kinygos: and then do 'mkswap /dev/md0p0' or whatever the partition is
[18:33] <SpamapS> kinygos: for the ext3 ones, 'mkfs.ext3 /dev/md0pX' should be enough. Then mount, rsync, done.
[18:45] <kinygos> SpamapS: should i have done that before writing the partition table out? :(
[18:46] <SpamapS> kinygos: no
[18:46] <SpamapS> kinygos: you can just go change the types
[18:46] <SpamapS> kinygos: and with swap.. its sort of.. less critical ;)
[18:46] <kinygos> SpamapS: awesome, thank you for your time and continued support :)
[18:48] <SpamapS> np
[19:09] <spajderix> Hi
[19:12] <spajderix> on one of my server every few minutes la jumps very high all of a sudden, and server almost freezes. Situation lasts for a few seconds, and everything gets back to normal. htop shows no cpu usage, no sudden memusage, iotop shows no hdd activity. Is there a way to check what exactly blocks cpu, and what this process does?
[19:13] <RoyK> spajderix: nfs blocking?
[19:17] <kinygos> are directories like /lib and /var/lib related?
[19:19] <RoyK> not really, /lib contains system libraries critical to the OS, /var/lib usually contains stuff like database data, dhcp leases and so on
[19:19] <stbain> Are there any monitoring software packages (such as Zenoss or Hyperic) that are maintained and updated in the Ubuntu server repositories?
[19:19] <RoyK> nagios?
[19:19] <RoyK> that is - if you can say nagios is ever updated (it rarely is)
[19:20] <stbain> RoyK, was looking for something a little more robust. Been a while since I looked at Nagios.
[19:20] <RoyK> nagios is robust, but the gui sucks
[19:20] <stbain> Is it still manual configuration file editing?
[19:21] <RoyK> there are addons to do automatic stuff, but yes, manual config is the standard
[19:21] <RoyK> I somehow like that - a config file doesn't get damaged as easily as a database gets corrupted
[19:21] <RoyK> and a config file is easy to replace if something goes wrong
[19:22] <RoyK> anyway - the configuration of nagios (or Icinga, a fork) is quite easy to learn
[19:25] <stbain> I recall configuring Nagios manually... was hoping perhaps something better had come along by now.
[19:27] <RoyK> stbain: nagios development has almost halted - seems they want to make money on nagios iv instead of keeping up the open development - that was the main reason for the icinga fork, which includes quite a bunch of new stuff
[19:32] <RoyK> Europe according to Estonians http://www.youtube.com/watch?v=WUgqXGu_gTQ :D
[19:33] <kinygos> in a similar vein, i lol'd at http://alphadesigner.com/project-mapping-stereotypes.html
[19:34] <RoyK> hehe
[19:40] <cfairles> what is mysql-dfsg ? vs. mysql-server
[19:40] <RoyK> mysql-wtf?
[19:42] <cfairles> Apparently there's a fix in mysql-dfsg for an apparmor issue preventing the innodb plugin from loading, but no lucid repo i have in my sources has a package called mysql-dfsg...
[19:42] <sbeattie> cfairles: mysql-dfsg is the source package name, mysql-server is the binary package name.
[19:43] <cfairles> hrm
[19:44]  * RoyK doesn't like mysql very well except for tiny web stuff
[19:44] <cfairles> sbeattie, i see, thanks. i think i'll just modify the apparmor config file myself for now.
[19:47] <RoyK> cfairles: keep in mind that there are some rather bad bugs with innodb on high load
[20:01] <illytacos> hello folks, have a bit of a problem and was wondering if someone could help. I am trying to manually configure and connect a SMB file server. I tried smbclient -L //server -U root and entered the server password not sure if this is correct got the error Connection to server failed (Error NT_STATUS_HOST_UNREACHABLE) right now I just want to see this ubuntu box on another windows laptop on the network. Any advice? I'm sure I'm
[20:02] <EvilPhoenix> your message was cut off at "I'm sure I'm..."
[20:02] <illytacos> oh that's ok I was just saying I'm sure I'm going wrong somewhere just not sure where lol
[20:17] <SpamapS> illytacos: are you sure samba is running on the server?
[20:19] <RoyK> illytacos: usually the root account is blocked in samba
[20:19] <illytacos> SpamapS: nope. when I set up ubuntu server edition it gave me the option at the end to select from lamp server, samba server etc so I selected samba assuming it would be installed. I didn't even realize I had to run it. no wonder I'm running into problems
[20:20] <SpamapS> illytacos: well it may or may not be running
[20:20] <SpamapS> illytacos: if its installed, it should be running
[20:20] <illytacos> SpamapS: how would I go about checking that>
[20:22] <illytacos> omg ok sorry I think I understand what happened I have smbfs installed, smb client, samba-common-bin etc but samba itself... not installed -_-' sorry
[20:37] <elb0w> Anyone installed mysql 5.5 on ubuntu 10.04?
[20:37] <scrllock> are you having issues with it?
[20:37] <elb0w> yeah
[20:38] <elb0w> I tried the binary
[20:38] <elb0w> and when I got done it wouldnt start
[20:38] <scrllock> did it give you an error?
[20:39] <elb0w> 100930 15:32:00 [ERROR] Can't find messagefile '/usr/share/errmsg.sys'
[20:42] <elb0w> Anyone else?
[20:44] <_ruben> that path doesnt look very sane to me
[20:45] <elb0w> what?
[20:48] <_ruben> i'd expect something like /usr/share/mysql-server/errmsg.sys instead .. then again, never with any mysql other than the one provided by default
[20:48] <_ruben> and im afk again
[20:52] <illytacos> you guys rock thanks the advice helped. I feel a little silly but I am in new territory here :) thanks for making it easier
[21:00] <elb0w> root@DBS3:/usr/local/mysql# bin/mysqld status --lc-messages-dir="/usr/local/mysql/share/english/"
[21:00] <elb0w> 100930 15:59:40 [ERROR] An old style --language value with language specific part detected: /usr/local/mysql/share/english/
[21:00] <elb0w> 100930 15:59:40 [ERROR] Use --lc-messages-dir without language specific part instead.
[21:00] <elb0w> any clue? sorry for paste in chan, just 3 lines
[21:06] <dankest> Am I visible?
[21:06] <elb0w> WHO SAID THAT
[21:06] <dankest> hahaha appreciate it
[21:07] <RoyK> elb0w: tried #mysql?
[21:07] <elb0w> im sure theyll say they only do queries
[21:07] <elb0w> sec
[21:13] <wirelessdreamer> i'm stuck in a dependency loop, tied into this bug https://bugs.launchpad.net/ubuntu/+source/mysql-cluster-7.0/+bug/629005
[21:13]  * RoyK heaves over his IBM 5155 portable
[21:13] <wirelessdreamer> I can't get apt back into a sane state after trying to install mysqlcluster
[21:14] <wirelessdreamer> this is on lts not the rc
[21:46] <SpamapS> mathiaz: hey, would you say we're done ISO testing for the RC?
[21:46] <mathiaz> SpamapS: definetely
[21:46] <mathiaz> SpamapS: RC has been released
[21:46] <mathiaz> SpamapS: there will be a respin for final
[21:46] <SpamapS> mathiaz: ok, just updating workitems... gotta move that line. ;)
[21:46] <mathiaz> SpamapS: so we'll do more iso testing down the line
[21:46] <mathiaz> SpamapS: ;) - did the same thing this morning!
[21:47] <mathiaz> SpamapS: and I'm working on another WI ATM
[21:47] <SpamapS> mathiaz: *almost* straight http://people.canonical.com/~platform/workitems/maverick/canonical-server-ubuntu-10.10.html
[21:48]  * SpamapS goes to lunch
[21:51]  * ttx waves
[21:52] <mathiaz> ttx: o/^5
[21:53] <ttx> mathiaz: everything good ?
[21:53] <mathiaz> ttx: RC is out of the door
[21:53] <ttx> mathiaz: inet connectivity is particularly bad at this conference
[21:53]  * mathiaz hands a satellite modem to ttx 
[21:54] <ttx> ah! that should be handy, if only the auditorium wasn't in the basement
[21:54]  * mathiaz hands a wide-range satellite antenna to ttx 
[22:18] <ruben23> hi guys whast the command to check the installed packages on an ubuntu-server..
[22:23] <ruben23> anyone have idea please
[23:01] <SpamapS> ruben23: dpkg -l
[23:02] <SpamapS> ruben23: note that it will also show packages which have been installed, but removed and left behind config files. (the ones that start with 'rc')