[00:34] <artillerytx> Hey guys for some reason i keep getting this error sent to my e-mail - http://ubuntu.pastebin.com/m167e03fe
[00:34] <artillerytx> how can i get rid of it
[01:10] <sanmarcos> is there a tool to redetect hardware ?
[01:20] <joejc> i have a stupid question: if u all have servers why is no one seeding the torrent
[01:23] <sanmarcos> joejc: bandwith isnt free
[01:23] <sanmarcos> most poeple dont seed
[01:24] <joejc> im not asking for 100% but maybe like 1kbps from everyone
[01:25] <joejc> most torrents with no seeders suk is that true for ubuntu server?
[01:30] <joejc> if ubuntu server isnt worth ur time(to respond to me) or ur bandwidth(to seed) why tf are 187 people here?
[01:32] <qman__> the torrents are most effective right around release time
[01:33] <qman__> the rest of the time, the http mirrors are fast, so most people don't seed the torrents
[01:34] <joejc> torrents = 100% guarantee of perfection, i dont like checking md5s
[01:40] <sanmarcos> joejc: yep
[01:40] <sanmarcos> but burning a DVD gave me failed verification don't know why
[01:40] <sanmarcos> stilled used the DVD
[01:40] <joejc> i use usb
[01:40] <joejc> i like perfection
[01:41] <joejc> failure is for the week
[01:42] <joejc> perfection is for the weekend and my time off XD
[02:22] <PsycoGeek> Hi. Do some of you got errors like these on Ubuntu 9.04 AMD64 ? [13653.179309] cron[24381]: segfault at 23fffffff0 ip 00007f715cddda7a sp 00007fff665e7ae0 error 4 in libtalloc.so.1.2.0[7f715cdd8000+8000]
[02:24] <twb> So has anybody moved their Windows AD server into KVM?
[02:27] <PsycoGeek> sad
[03:41] <uvirtbot`> New bug: #406698 in samba (main) "package samba-common 2:3.3.2-1ubuntu3.1 failed to install/upgrade: subprocess post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/406698
[04:26] <uvirtbot`> New bug: #406707 in lsb (main) "lsb_release closed unexpectedly (dup-of: 383697)" [Undecided,New] https://launchpad.net/bugs/406707
[05:58] <Sandra_> hi...
[05:58]  * genii makes more coffee
[05:59] <Sandra_> i have problems to run dpkg-reconfigure libpam-ldap ...doesn't do nothing...somebody know about this problem?
[06:00] <ball> I mustn't drink coffee right now.
[06:00] <ball> That would be a recipe for disaster ;-)
[06:22] <twb> Sandra_: you need ldap-auth-config
[06:24] <twb> aptitude install ldap-auth-config ldap-utils
[06:24] <twb> auth-client-config -a -p lac_ldap
[06:25] <Sandra_> thanks, i got that, but i dont know if nss and pam is running...when i try to config with dpkg-reconfigure libpam-ldap, does not show me nothing
[06:27] <Sandra_> how i can know if is working and configurate again?
[06:31] <Sandra_> hi twb, the second command hwt is for?. I got error Error in creating temporary file
[06:33] <twb> Sandra_: are you root?
[06:35] <Sandra_> yes
[06:35]  * genii shivers
[06:48] <twb> I dunno, then
[07:25] <Sandra_> hi, i delete pam and nss, but when i try to install again it look like is installed but does not create the files and forldes pam.d...How I can reinstall completelly pam and nss?
[07:27] <ScottK> purge, don't remove.  If you just remove a package it leaves the config files.  If you remove them by hand, it assumes you don't want them anymore.
[07:28] <Sandra_> tahnks, i will try again with purge and then re-install
[07:35] <artillerytx> someone was telling me that there are some UPC management tools built into ubuntu server already
[07:35] <artillerytx> UPS
[07:35] <artillerytx> whichever
[07:36] <Sandra_> whow...now i got something worst ...i try to access like root or my account and i got PAM critical error
[07:36] <Sandra_> i can not access to the server now, maybe is because i deleted the pam.d folder?
[07:36] <Sandra_> how i can recover it now?
[07:37] <twb> You deleted pam.d?
[07:37] <twb> That's about the dumbest thing you can do
[07:37] <Sandra_> can I recover a folder deleted with rm?
[07:38] <twb> By reinstalling your OS, or by restoring from the backups that you make every morning
[07:38] <twb> artillerytx: nut is the tool typically used to talk to a UPS from unix
[07:38] <artillerytx> twb: oh okay a
[07:39] <Sandra_> i dont have backup from the OS
[07:39] <twb> Sandra_: then you have learnt an important lesson about the value of backups
[07:42] <Sandra_> but i suppose there is some way to recover it or i can access to the sistem in someway, i have problme with the AM password
[07:42] <Sandra_> PAM
[07:50] <mattt> Sandra_: what you trying to do?  why did you remove pam?
[07:52] <Sandra_> because I tried to reconfigure with dpkg and i couln't so i though if i desinstall that I could install and confgure properly :S
[07:59] <Sandra_> sorry i was disconected
[08:07] <Timmy2Tall> waddup bitches
[08:25] <Sandra_> im getting this error pam_start() failed, error 26 ..somebody kow how recover it?
[09:21] <_21h_> hi all
[09:21] <_21h_> i have trouble with dhcp3-server
[09:22] <_21h_> client discovering dhcp server, server offering ip, client requesting and no DHCPACK as reply
[09:27] <twb> You mean DHCPDISCOVER (from client), DHCPRESPONSE (from server), followed by a DHCPREQUEST (from client)?
[09:28] <_21h_> yes
[09:28] <_21h_> there no DHCPACK after DHCPREQUEST
[09:29] <_21h_> client just requesting ip again
[09:30] <_21h_> DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 8
[09:30] <_21h_> DHCPOFFER of 172.16.0.99 from 172.16.0.1
[09:30] <_21h_> DHCPREQUEST of 172.16.0.99 on eth0 to 255.255.255.255 port 67
[09:30] <_21h_> DHCPREQUEST of 172.16.0.99 on eth0 to 255.255.255.255 port 67
[09:30] <_21h_> and this again again again...
[09:36] <twb> There is not supposed to be a DHCPREQUEST
[09:36] <twb> It indicates that your client is refusing to use the IP allocated to it by DHCPOFFER
[09:39] <_21h_> Client-Ethernet-Address 00:14:2a:aa:25:44 (oui Unknown) [|bootp]
[09:39] <_21h_> 17:33:38.009594 IP (tos 0x10, ttl 64, id 0, offset 0, flags [none], proto UDP (17), length 328) 172.16.0.1.bootps > 172.16.0.95.bootpc: BOOTP/DHCP, Reply, length 300, xid 0x3b5f3bbd, Flags [none]
[09:39] <_21h_>    Your-IP 172.16.0.95
[09:39] <_21h_>    Server-IP 172.16.0.1
[09:39] <_21h_> i think its reply to client
[09:40] <twb> Actually... the fact that it's DHCPREQUESTING with .99 is weird.
[09:40] <twb> Maybe your lease time is ridiculously small?
[09:41] <_21h_> 7000
[09:41] <twb> seconds?
[09:41] <_21h_> its max-lease-time
[09:42] <_21h_> default-lease-time 600
[09:42] <twb> Well, I'm out of idea
[09:42] <twb> *ideas.
[09:42] <twb> Try hassling the ISC people in their channel?
[09:43] <_21h_> say their channel
[09:44] <twb> I've no idea what it is
[09:44] <_21h_> :)
[10:10] <_21h_> i found problem
[10:10] <_21h_> disk full
[10:13] <ttx> heh
[10:14] <twb> Ha
[10:14] <twb> I hate that
[10:15] <twb> Like one time I had a RHEL box take 30 minutes to boot -- turned out that the actual problem was that /etc/sysconfig/network/thingy-eth0 was owned by a user that didn't exist anymore
[10:16] <twb> And the start scripts tried to use ls programmatically (stupid), which resulted in it being stat(2)ed, which resulted in libnss trying to talk to an LDAP server on the network that wasn't up yet, and taking ages to time out (due to hard binding).
[12:23] <quizme> i am using an image from Canonical, which replaces vi with vim.  How can i change back to classic vi ?
[12:24] <pmatulis> quizme: an 'image'?
[12:24] <Viltapi> try removing vi and vim with apt-get and then install vi again?
[12:25] <quizme> pmatulis nm
[12:25] <quizme> apt-get remove vi vim; apt-get install vi  ?
[12:27] <Viltapi> quizme: yeah that's what i would do
[12:28] <quizme> viltapi: it's not finding vi as package
[12:29] <Viltapi> quizme: what about that vim which you installed?
[12:29] <twb> quizme: vim is the default vi implementation on Ubuntu
[12:29] <twb> quizme: you can install nvi, and make vi a symlink to it using update-alternatives(8)
[12:29] <quizme> i hate vim
[12:29] <twb> You could also install busybox and alias vi to "busybox vi"
[12:29] <quizme> nvi...
[12:30] <quizme> what's busybox ?
[12:30] <twb> quizme: use "apt-cache show busybox" to find out.
[12:31] <Bilge> Imagine using vi to edit documents
[12:31] <twb> I use ed.
[12:31] <Bilge> The year is 2009
[12:31] <quizme> twb: that did it.  .  nvi rules
[12:31] <twb> vi is for wankers with expensive glass ttys
[12:32] <Bilge> Cool story bro
[12:32] <quizme> hehe
[12:33] <quizme> I use vi to edit Microsoft Word documents
[12:33] <quizme> j/k
[12:36] <pmatulis> quizme: vi can do more than vim?  why do you want to remove it?
[12:36] <quizme> pmatulis: i don't like learning new things.
[12:37] <quizme> pmatulis: actually, i was having trouble with indentation with vim.... i'm just not used to it.
[12:37]  * pmatulis thought vim is entirely backwards compatible with vi
[12:40] <quizme> pmatulis: when i pasted text into vim, it added an extra couple of spaces each line.  so my text ended up being indented along a diagonal...
[12:41] <pmatulis> k, i've had that happen too
[12:41] <pmatulis> but i didn't think it was vim-specific
[12:42] <pmatulis> quizme: http://vim.wikia.com/wiki/Toggle_auto-indenting_for_code_paste
[12:43] <pmatulis> just a simple matter really, thanks for the tip!
[12:44] <quizme> :pmatulis: thanks
[12:46] <stas> hi guys, anyone's using zfs on fuse?
[13:45] <juxta> I'm seeing some weird modprobe lines in my syslog periodically, I'm trying to work out what might be causing them. I've pastebinned them here http://pastebin.com/d36f2ca54. I don't use xen at all, so I'm trying to work out what might be trying to load that module...
[15:36] <refnumzx> i am getting a [ 3660.842831] dansguardian[16216]: segfault at a7 ip b7d5c5eb sp bfec58bc error 4 in libc-2.9.so[b7ce5000+15c000] from dansguardian in jaunty. anyone know how to fix this?
[15:47] <ScottK> refnumzx: What version of dansguardian and what version of libclamav do you have installed?
[15:48] <refnumzx> i have the version that shipped with jaunty. nothing custom.
[15:50] <refnumzx> dansguardian 2.9.9.7 and liclamav5? not sure how to check for that
[15:51] <refnumzx> libclamav6.
[15:59] <refnumzx> i'd really like to try and run the 2.10 series. but i can't get it to compile, nor can i find deb for it.
[16:17] <\sh> hmm...which version of drbd is being used 0.7 or 0.8? (regarding the kernel modul)
[16:18] <\sh> (add: jaunty here)
[16:19] <\sh> ok...version 8
[16:29] <refnumzx> so i am trying to compile dansguardian.  libpcre doesn't seem to be available. during configure. and apt-cache search doesn't revewal anything promising. ideas?
[16:34] <mathiaz> ttx: hey
[16:35] <mathiaz> ttx: thanks for reviewing my cim-schema merge
[16:35] <ttx> mathiaz: np
[16:35] <Kira> Has a recent security update disabled dynamic SSH tunnel by default?
[16:35] <mathiaz> ttx: I'd like to do that kind of practive more often.
[16:36] <mathiaz> ttx: I've got another package lined up (sblim-cmpi-base) - can I ask you to review the update too?
[16:39] <Kira> I have a Jaunty server at a data center and I connect to it though SSH, using the -D option so I can use it as a SOCKS5 proxy. It's been working almost flawlessly in the past, but I keep getting "open failed: administratively prohibited: open failed" today, since I applied some updates.
[16:54] <Kira> I'm starting to suspect that the Great Firewall of China has defeated SSH tunneling.
[16:57] <refnumzx> heh. al i need is for dansguardian to work on jaunty. other then that. china can firewall all it wants.
[16:58] <Kira> What's dansguardian?
[16:58] <refnumzx> a web content filter.
[16:59] <Kira> seriously, something is messed up.
[16:59] <refnumzx> i know it keeps segfaulting on me. like crazy
[16:59] <Kira> Has my SSH keys been cracked or what...
[17:00] <refnumzx> what is happening with your connection?
[17:01] <Kira> My SSH session works find as long as I don't make a tunneled http request.
[17:01] <refnumzx> what happens then?
[17:03] <Kira> as soon as my web browser makes an http request, which is supposed to be tunneled through my SSH session, my SSH session becomes unresponsive to keyboard input, until the http request is declared timed-out by the web browser + about one minute of idle time.
[17:05] <Kira> at which point I see messages like "channel 3: open failed: administratively prohibited: open failed" in my SSH session.
[17:08] <Kira> Which makes me wonder if the connection is being prohibited at my local machine or at the Ubuntu server.
[17:12] <refnumzx> kira: have you changed hostnames or usernames or something. might be a possibiity.
[17:14] <Kira> refnumzx: switching hostname seems to work
[17:15] <Kira> (there are two domain names pointing to my server)
[17:15] <Kira> work... not
[17:15] <refnumzx> ah. well, glad it worked.
[17:16] <Kira> I get SOME responsiveness and tunneled traffic.
[17:16] <Kira> but it's really slow.
[17:17] <refnumzx> i duno about speed. if the connections works, ssh tunneling works.
[17:17] <Kira> Has the Great Firewall gained some deep packet inspection capability so advanced that it can guess tunneled http traffic without cracking the keys?
[17:18] <Kira> (theoretically, it should be nonsense, I know)
[17:26] <refnumzx> uh. no idea.
[17:44] <refnumzx> i saw a bug report about dansguardian, and a release was supposedly fixed.  release. but i can't find the appropriate deb files.
[17:45] <refnumzx> ideas? compiling latest dansguardian is beyond my skil.
[18:21] <uvirtbot`> New bug: #406987 in lsb (main) "lsb_release crashed with ImportError in <module>() (dup-of: 383697)" [Undecided,New] https://launchpad.net/bugs/406987
[19:03] <clusty> hey
[19:04] <clusty> i was curious about the perfromence I could expect of a NFS server over gigabit ethernet
[19:04] <clusty> sustained transfer not bursts
[19:04] <clusty> is it coming even close to the gigabit throughput?
[19:16] <uvirtbot`> New bug: #399443 in apache2 (main) "package apache2.2-common 2.2.11-2ubuntu2.2 failed to install/upgrade: package apache2.2-common is already installed and configured" [Undecided,Incomplete] https://launchpad.net/bugs/399443
[19:31] <john_doe> I need to "move" a mail server (including users and their mails) from one ubuntu install to another. Do you know a guide to this? I must not lose data or emails sent in the meantime...
[19:35] <clusty> hey
[19:35] <clusty> any1 has some experience with FAI and ubuntu?
[19:35] <clusty> i can not find any help regarding this
[19:38] <mm_work> Hey guys, on 8.04, how do I enable lastb?  I've googled around and cant seem to find anything useful.
[19:40] <KillMeNow> john_doe, i am assuming that your users are local accounts on the box>?
[19:43] <clusty> mm_work, what is that if i dare ask?
[19:48] <mathiaz> zul: could you write up an MIR for puppet?
[19:53] <\sh> mathiaz: puppet in main? sounds google^Wpromising ;)
[19:54] <mathiaz> \sh: puppet is used in a number of places and google is maintaining in debian/ubuntu
[19:54] <mathiaz> \sh: so it seems that it's a good candidate to push in main
[19:55] <mm_work> clusty: it shows bad login attempts.  Like 'last' shows valid logins.
[19:55] <\sh> mathiaz: yepp...I've chosen puppet, too, over cfengine
[19:55] <mm_work> I would just like to see if someone is trying to ssh into my box or not.
[19:55] <mm_work> Im just surprised that its not enabled by default.
[19:55] <clusty> mm_work, ohh ok. why not grep /var/log/auth ?
[19:56] <clusty> you get there all possible logins
[19:56] <mathiaz> \sh: how long have you been using puppet?
[19:57] <mm_work> clusty: thanks, I forgot about that.  And yep, looks like someone is trying a dictionary attack.
[19:57] <\sh> mathiaz: started last year...I'm integrating it now with FAI
[19:57] <mathiaz> \sh: isn't there an overlap between FAI and puppet?
[19:57] <clusty> mm_work, i used blacklist for that
[19:57] <clusty> mm_work, i think :D
[19:58] <clusty> mm_work, the ipchains one seemd very complicated. I used the one that write stuff in hosts/deny
[19:58] <clusty> hosts.deny
[19:58] <clusty> bad news is that i dont think you can kill apache access for example (stuff has to support tcp-wrappers)
[19:59] <\sh> mathiaz: not in the general meaning of configuration management...FAI has a softupdate task, but this is not what you really want when you have different network areas for deployment and production (like native vlan on a trunk for deployment, and your production vlans can't reach your deployment vlan)..you would need a second FAI instance..and there is puppet or cfengine much better...FAI e.g. uses cfengine, too, softupdate is something special for 
[20:00] <mathiaz> \sh: forgive the simple question, but what is the role of FAI in all that then?
[20:00] <clusty> mathiaz, my man :D. I was looking into FAI just now :D
[20:01] <\sh> mathiaz: fully automatic deployment...
[20:01] <clusty> mathiaz, works ok for ubuntu server?
[20:01] <mathiaz> \sh: right - so FAI is used up to the point where the puppet client can be run?
[20:01] <mathiaz> \sh: IOW FAI provides the same functionilities as cobbler?
[20:01] <\sh> mathiaz: that's the idea
[20:02] <\sh> mathiaz: dunno what cobbler is...but FAI is more solaris jumpstart...much faster d-i preseed or kickstart/autoyast
[20:03] <mathiaz> \sh: faster *than* d-i preseed?
[20:03] <mathiaz> \sh: ok - cobbler is something similar than FAI - https://fedorahosted.org/cobbler/
[20:04] <\sh> mathiaz: yes...I d-i preseed jaunty, and it took on a normal server 10 minutes...the same install with fai 3 mins
[20:04] <\sh> (jaunty server)
[20:04] <mathiaz> \sh: does FAI skip the installer?
[20:04] <mathiaz> \sh: and uses debootstrap directly?
[20:05] <\sh> mathiaz: yes...it untars a debootstrap directly on $target partitions (which os partitioned before that)
[20:05] <\sh> mathiaz: and it works with more then debian alike OS...
[20:05] <\sh> mathiaz: SLES 9 -> via autoyast deployed: 30 mins...with FAI: 10 mins (including YUM package manager and YUM repos)
[20:05] <mathiaz> \sh: does FAI provide the same step than the installer?
[20:07] <\sh> mathiaz: means? it boots a fully functional linux..starts a perl process, which runs some tasks, where you can hook in
[20:07] <\sh> mathiaz: it's more a deployment framework...ever seen empirum for windows installations?
[20:08] <mathiaz> \sh: the installer takes care of setting a lot of different part of the system (language, boot, etc..)
[20:08] <mathiaz> \sh: I wonder if FAI covers all of the quirks that the installer support
[20:08] <\sh> mathiaz: ah in that meaning...no that you do for yourself...like writing the preseed file...but more easier or more complex then preseed files...
[20:09] <mathiaz> \sh: and none of the installer component are called.
[20:09] <mathiaz> \sh: does FAI take care of PXE/dhcp, etc...?
[20:09] <\sh> mathiaz: none..well I use some debconf magic to get some e.g. non-interactive sun-java* license agreement, but that's only defining a file with the debconf preseeds for the package and a class file for the host
[20:10] <\sh> mathiaz: you install your own dhcp and your own tftp...and no, it doesn't deal with the management, that's what I'm writing now for FAI with django
[20:10] <\sh> http://www.sourcecode.de/content/fai-djangofied
[20:10] <mathiaz> \sh: ok - you may wanna have a look at cobbler - it does cover that part as well
[20:11] <mathiaz> \sh: cobbler can "import" an release and configure pxe/tftp/dhcp automatically
[20:11] <mathiaz> \sh: and then you add host to it and it will bootstrap the installation with the right files.
[20:12] <mathiaz> \sh: I'm not sure if cobbler has a web interface though
[20:12] <\sh> mathiaz: well...can you add your asset management as cmdb to it?
[20:12] <RoAkSoAx> mathiaz, it does have a web interface
[20:12] <RoAkSoAx> though I don't know its complete functionality
[20:12] <\sh> or better to say: your cmdb is the only base of knowledge ... and you need to tell your deployment tool to use this data
[20:12] <\sh> and that's easy to accomplish with FAI, as a framework for it...
[20:13] <mathiaz> \sh: right - it depends which cmdb you're using
[20:13] <mathiaz> \sh: I don't know how pluggable cobbler is
[20:13] <\sh> mathiaz: HP openviews e.g. or Idoit, or CAs
[20:14] <\sh> mathiaz: right:) I couldn't do it with RHN or kickstart or d-i preseed or autoyast :) that's why I choose FAI, which is more debian
[20:15] <\sh> mathiaz: e.g. the city of munich uses FAI extensively for their school it infrastructure deployment :) gonicus implemented a web ui for it (for this special case, it's named gosa) and they do OS deployment all over their area..
[20:16] <genii> I used to hate debian-installer preseed because of lack in documentation. But I don't mind it so much now.
[20:16] <mathiaz> \sh: right - I saw the FAI module when I looked at gosa
[20:17] <mathiaz> \sh: my main issue with gosa is that the ACL are stored in the objects rather than using the native openldap ACL
[20:17] <\sh> mathiaz: lycos had all their servers deployed with it...many cluster guys are doing it..just check out http://www.informatik.uni-koeln.de/fai/q-answers.html ...
[20:18] <\sh> mathiaz: yes..but this is special to gosa...nothing to do with FAI :) but you are free to do what you want with it:) you can plug it into any infrastructure...my hobby fai project is to install windows ;)
[20:18] <\sh> just like empirum is doing it or altiris
[20:19] <mathiaz> \sh: right - I was disgressing from the topic of FAI above ;)
[20:19] <mathiaz> \sh: it has nothing to do with FAI
[20:21] <\sh> mathiaz: right :) anyways..it's a matter of your taste...there are many solutions out there...and SysADmin should use what she thinks is right for their infrastructure...
[20:24] <Authority> Is it just me, or does cups needlessly include a conversion to PDF in it's processing now?
[21:05] <T-Hawk> hey ppl... recently amavis started sending me updates every 3 hours like this: bayes: synced databases from journal in 0 seconds: 298 unique entries (717 total entries)
[21:05] <T-Hawk> i've looked in the cron job, and in amavisd-new-cronjob and it should pipe stdout to /dev/null, anybody know how i can get rid of these messages?
[21:05] <T-Hawk> i don't get any output if i run the cronjob by hand either, it's only when cron is running it
[23:06] <Xlrr8> Hey Guys,  I have a question... I just came from #ubuntu and a really helpful guy there suggested that i ask my question here.  I've got a dell gx620 64bit box with 8gb of ram.  Its running ubuntu 9.04 64bit.  Can someone tell me why it's only addressing 3.4 gb of ram according to gnome system monitor and also according to /proc/meminfo ? I found a thread http://ubuntuforums.org/showthread.php?t=1061673 however, it doesn't look like t
[23:20] <infinity> Xlrr8: Looks like that thread describes your problem exactly.
[23:20] <infinity> Xlrr8: dmesg | grep MTRR
[23:21] <Xlrr8> Yes i agree.   The gentelman from #ubuntu suggested that i pipe out dmesg and i do in fact have this line in there :
[23:21] <Xlrr8> [    0.000000] WARNING: BIOS bug: CPU MTRRs don't cover all of memory, losing 4096MB of RAM.
[23:22] <infinity> Xlrr8: Right, well.  Not much I can say about that except "complain to Dell". :/
[23:22] <Xlrr8> I guess i'm wondering if this is a bug ?  Or a hardware limitation that the bios is recognizing the ram, but not passing it along to ubuntu ?  Is that even a possibility ?
[23:22] <infinity> It's a hardware/BIOS bug.
[23:23] <Xlrr8> even though the bios recognizes all 8gb ?
[23:24] <infinity> Yes.  It's not actually setting it all up.
[23:26] <Xlrr8> Ok, i was afraid of that. So i removed 4gb and left in 4gb, The system still only addressed 3.4.  Would it be safe to say that this "bios bug" is simply ignoring anything over 3.4gb ram ?
[23:26] <infinity> Found a few other people with GS620s with the same issue.  *shrug*
[23:26] <infinity> And yeah, that would be a safe assumption.
[23:26] <Xlrr8> bummer ! :(
[23:26] <Xlrr8> Thanks for your help/confirmation ! :)