[00:14] <KillMeNow> http://www.linuxforums.org/forum/ubuntu-help/104116-apt-get-error-e-method-http-has-died-unexpectedly.html
[00:14] <KillMeNow> check that out Zende
[00:14] <KillMeNow> hope it helps
[00:15] <zende> KillMeNow: I saw that thread, but it doesn't have enough information to help
[00:15] <zende> apt-get segfaults when writing gpg keys
[00:16] <zende> [pid  1911] write(4, "-----BEGIN PGP SIGNATURE-----\nVer"..., 189) = 189
[00:16] <zende> [pid  1911] +++ killed by SIGSEGV +++
[00:39] <KillMeNow> hrm...  you can always try to truss the gpgkey creation
[00:39] <KillMeNow> see exactly where it's barfing
[00:40] <zende> KillMeNow: how do I do that?
[00:41] <KillMeNow> http://www.cyberciti.biz/faq/howto-use-linux-truss-strace-command/
[00:44] <KillMeNow> you should be able to monitor all the system calls etc using that
[00:45] <Bookman> I just installed denyhosts in 9.04 desktop and I modified the .conf file.  Now when I start it up and examine the denyhosts.log file it does not seem to be picking up the settings that I've changed in .conf even though it says that it is pointing to the right .conf file.
[00:46] <Bookman> *denyhosts.log file is just denyhosts
[01:00] <dinger2006> anyone used imagemagick?
[01:01] <KillMeNow> only used it for gallery2 long time ago
[01:02] <dinger2006> ah ok, noone ever used it for conversions?
[01:42] <Bookman> Anyone online that can lend a hand with a denyhost installation in 9.04 desktop?
[05:00] <teddy_> running ubuntu from a usb flash drive...apt-get update and apt-get upgrade...The update is very slow...973 B/sec.  I was wondering if this was because it is a flash drive, or that mirror is very slow
[05:02] <teddy_> how do i change mirrors to find out if its the mirror or the faslh usb ?
[05:02] <twb> teddy_: it's more likely to be the mirror.
[05:09] <jmarsden> teddy_: On Ubuntu server, change the mirror you use by editing /etc/apt/sources.list .  You may also want to check your general Internet connection speed is decent -- maybe it is not the mirror but is your network connection??
[05:10] <twb> If I saw 973B/sec, I would assume that someone has kicked out the network cable
[05:10] <twb> And the reason it doesn't say 0B/s yet is because it's averaging against a much faster initial speed (before the cable was yanked).
[05:11] <teddy_> i got Karrrazy bandwidth
[05:12] <twb> teddy_: you're in the caucasus?
[05:12] <teddy_> whats that?
[05:12] <twb> teddy_: a mountain range.
[05:12] <teddy_> no
[05:13] <twb> Because "Karrrazy" sounds like an eastern European name :-)
[05:14] <twb> http://en.wikipedia.org/wiki/Caucasus
[05:20] <teddy_> usb flash drive system, apt-gets now much faster after changing mirrors
[05:21] <nick125> Good evening. What virtualization method would you guys suggest for virtualizing Linux guests on Ubuntu on a system without VT?
[05:21] <teddy_> xen
[05:22] <jmarsden> virtualbox-ose
[05:22] <jmarsden> It all depends what the guests will be used for...
[05:24] <nick125> They'll be used as servers
[05:24] <nick125> Right now, I have a Gentoo-based Xen box that I have three VMs on, running all of my network services. I want to migrate that server to Ubuntu, but I want to reevaluate what virtualization solution I'm using.
[05:26] <jmarsden> xen or vmware server are the logical choices for that, I would think.
[05:26] <nick125> Hm. I've never used VMWare Server.
[05:26] <teddy_> 64-bit CPUs and Vmware ESXi
[05:27] <nick125> teddy_: Unfortunately, I can't spend money on new hardware...and ESXi doesn't support softraid
[06:33] <mattt> nick125: i think the latest version of ubuntu only supports kvm, so be careful which release you end up using
[06:36] <twb> Ubuntu never supported vmware directly.
[06:36] <mattt> twb: i mean as opposed to supporting both xen & kvm :)
[06:36] <twb> Ah, sorry, I wasn't paying attention.
[06:37] <twb> You mean Ubuntu libvirt doesn't support xen?
[06:37] <mattt> (i wasn't clear tho, thanks)
[06:38] <mattt> twb: i don't think there's a xen kernel in jaunty
[06:38] <twb> Does the Dom0 need a special kernel?
[06:39] <mattt> my dom0 lenny box: 2.6.26-2-xen-686
[06:39] <mattt> perhaps there is a xen kernel on jaunty now, there wasn't when it was first released tho
[06:40] <twb> mattt: did you have universe turned on?
[06:40] <mattt> twb: i ended up using lenny instead :)
[06:55] <jmarsden> http://packages.ubuntu.com/jaunty/xen-hypervisor-3.3
[06:57] <jmarsden> "... An example config file for this kernel and documentation on how to build it can be found in the xen-docs package." -- in other words, you need to build your own dom0 kernel, apparently.
[07:01] <mattt> jmarsden: PITA!
[07:05] <jmarsden> mattt: Maybe.  You're unlikely to be updating your server's dom0 kernel every week, you aren't likely to auto update that kernel either... and you can (presumably) script the kernel build process, so it's mostly a one-time cost.
[07:14] <billybigrigger> anyone alive?
[07:15] <billybigrigger> and that uses awstats? how do i get apache to either spit out 1 large access.log for awstats to read, or have awstats read access.log and keep it continuous
[07:15] <billybigrigger> every time apapche rotates the log, i loose all the stats, so i only end up having logs for a couple of days
[07:18] <jmarsden> billybigrigger: Read the comments in /etc/awstats/awstats.conf regarding the logfile and use of pipes...
[07:19] <billybigrigger> pipes, ahh, thanks jmarsden
[07:19] <jmarsden> Should be lots of ways to recombine your log files and feed them to it that way.  Maybe LogFile="cat /var/log/apache2/*.log |"  or similar
[07:38] <billybigrigger> jmarsden, the logs are ouputting to .log and .log.1 log.2
[07:38] <billybigrigger> would access.* or *.log* work better?
[07:39] <billybigrigger> althought i would like to encorporate other_vhosts_access.log
[07:39] <billybigrigger> so *.log* would be my best option yes?
[07:44] <billybigrigger> webalizer doesn't like that command
[07:44] <billybigrigger> awstats worked ok
[07:44] <billybigrigger> but webalizer.conf looks for logfile /var/......
[07:44] <billybigrigger> not logfile="/var/....."
[07:45] <billybigrigger> Webalizer V2.01-10 (Linux 2.6.28-11-server) locale: en_US.UTF-8
[07:45] <billybigrigger> Error: Can't open log file "cat /var/log/apache2/*access*.log* |"
[08:06]  * soren wonders where uvirtbot went
[08:07] <henkjan> uvirtbot is lost in cyberspace ( 19:20 -!- uvirtbot` [n=supybot@amdi.linux2go.dk] has quit [Read error: 113 (No route to host)]
[08:08]  * soren taps fingers
[08:08] <soren> uvirtbot: Welcome back.
[08:31] <Dustan> newbie lamp admin here wondering if it's a good idea to set my server to automatically update?
[08:36] <twb> Dustan: that depends on what kind of failure you want to get
[08:36] <twb> Do you want your server to fail because it has been haxxored, or do you want it to fail because an automatic update was installed and something broke because you weren't babysitting the upgrade?
[08:37] <mattt> hehe
[08:42] <mattt> Dustan: subscribe to something that informs you when updates are available, and then apply them manually :)
[08:43] <Dustan> ty
[08:45] <mattt> Dustan: i'm quite new to ubuntu and don't know how package verions work (does it go as an ex. 4.4.1 -> 4.4.2, etc.?) ... if so, those kinds of upgrades could break external PHP modules (pecl, etc.)
[09:12] <andenw_> matt, Dustan: security updates of packages should never include a new version that breaks compatibility - the security fixes are backported to apply to the older versions. (note "should" here, sometimes it is not possible to backport the fixes)
[09:13] <mattt> andenw_: thanks for confirming ... so it sounds similar to RHEL's approach
[09:14] <quizme> i tried to login to my server with ssh and failed 3 times at a password attempt, now i'm getting connection refused port 22 for that user only.  anybody know how to fix that ?
[09:15] <quizme> is that some kind of a security thing ?
[09:16] <quizme> actually it's not just for that user
[09:16] <quizme> the whole domain seems to be shut off
[09:22] <twb> mattt: the standard line is that you should not be installing packages that circumvent apt and dpkg e.g. via PEAR or cpan(1).
[09:23] <twb> quizme: there are such things, but they are not on by default (AFAIK).
[10:36] <scott_nwoktech> i'm trying to get pacemaker/openais etc for hardy server. The PPA doesn't work. Would packages built for debian work?
[10:38] <mobi-sheep> I installed Ubuntu's kernel on my grandma's pacemaker this morning.
[10:38] <mobi-sheep> It worked great. <3
[10:38] <scott_nwoktech> ha
[10:39] <mobi-sheep> Using debian packages (and/or) debian repos?  You're asking for troubles IMO.
[10:40] <scott_nwoktech> yea...i decided to build from latest source, make my own debs. that will allow apt to upgrade them eventuallly right?
[10:42] <mobi-sheep> I don't know.  I'm not the expert here. :)
[10:46] <andenw_> scott_nwoktech: apt will probably not upgrade them correctly, unless you get version numbers exactly right..
[10:47] <andenw_> scott_nwoktech: you can create your own PPA, and upload the debian source packages and tell the build servers to rebuild them for ubuntu...
[10:53] <scott_nwoktech> i haven't had any experience with using your build servers, etc. i do have an account w/ launchpad...do you mean the debs i would build the source tarballs from? just upload them and tell the servers to build for whatever ubuntu versions i want?
[11:12] <mattt> hm, trying to scp a 100GB+ file from one system to another (remote system is using XFS), but it keeps failing around 64 GB
[11:12] <mattt> as far as i can tell that's not an XFS limitation?
[11:20] <Dustan> isn't there a one line lamp installation?
[11:20] <soren> Dustan: sudo apt-get install lamp-server^
[11:20] <Dustan> lmao
[11:20] <Dustan> wow
[11:20] <Dustan> that installs all the latest stable packages?
[11:23] <soren> Well, yes.
[11:24] <soren> mattt: I'm not familiar with any limitations like that in XFS, no.
[11:24] <soren> mattt: Perhaps openssh's internal counter falls over.
[11:24] <soren> mattt: How does it fail, exactly?
[11:25] <mattt> soren: "scp" just stalls and eventually disconnects
[11:26] <mattt> soren: rather, scp just "stalls" :)
[11:27] <zende> mattt: you could try breaking up the file using split.  worth a try
[11:27] <mattt> yea, was thinking about that ...
[11:28] <soren> That really shouldn't be necessary.
[11:28] <soren> mattt: Can you try it again, and try stracing the scp process?
[11:29] <mattt> soren: i'm rsyncing it now, let me start to strace
[11:29] <soren> mattt: rsync doesn't work either?
[11:30] <mattt> soren: i'll find out shortly :) it's ~ 120 GB, so takes a while :)
[11:30] <mattt> at 58 GB now .....
[11:33] <mattt> soren: odd, it's going still
[11:33] <mattt> soren: hmmm
[11:33] <mattt> soren: i see what's happening
[11:34] <mattt> soren: at 64 GB, for whatever reason, the remote ssh server is requesting me to re-authenticate
[11:34] <mattt> soren: i caught it in time this time and it resumed
[11:35] <soren> mattt: Reauthenticate? How?
[11:35] <mattt> soren: password/key authentication
[11:35] <mattt> soren: (rsync over ssh)
[11:35] <soren> mattt: But the session is already set up?
[11:35] <soren> Oh, rsync.
[11:35] <mattt> soren: yep
[11:35] <soren> Well, in that case, the connection must have been dropped and rsync tries to connect and pick up where it left off.
[11:36] <mattt> soren: 3 times at 64 GB tho?  :)
[11:36] <soren> Yes.
[11:36] <mattt> that seems a bit strange
[11:36] <soren> Is that strange?
[11:36] <soren> Why?
[11:36] <soren> I'm not saying the connection is being dropped randomly.
[11:37] <soren> It's likely triggered by a bug that kicks in when 64 GB have been transferred.
[11:37] <mattt> yeah
[11:37] <mattt> or .....
[11:37] <zende> apt-get is blowing up when trying to download packages for me
[11:37] <mattt> my drive is read-only (hence me getting data off), and ssh can't update ~/.ssh/known_hosts
[11:38] <zende> E: Method http has died unexpectedly!
[11:38] <soren> mattt: I kidn of doubt that it would wait 64 GB before it tried to updated known_hosts?
[11:39] <mattt> i don't know if it periodically checks that and if it can't verify the authenticity of the remote host after transfering so much data you need to re-log in
[11:39] <mattt> (or it needs to re-verify the authenticity)
[11:39] <mattt> soren: no, it checks immediately
[11:39] <soren> mattt: That doesn't make sense for straight ssh.
[11:39] <soren> zende: Try in #ubuntu.
[11:40] <zende> soren: ok
[11:40] <mattt> soren: not sure, that was a weird one
[11:41] <soren> mattt: I'd try stracing ssh processes (both scp on the client, and the corresponding process on the server)
[11:41] <mattt> i have the strace output, 100+ MB tho so i need to weed through it
[11:41] <soren> mattt: From rsync or scp?
[11:42] <mattt> soren: rsync, but i -used a -f so hopefully it included the ssh subprocess
[11:44] <mattt> soren: i'll probably try disablign strict host checking later on, but need to get this file off before it's too late ... thanks for your input!
[12:36] <n3m3s1s4u> Hi all - can anyone suggest a really easy to use webmail front end (not squirrel or roundcube) that works well on sendmail and pop?
[13:01] <zende> what would cause apt-get to segfault when downloading packages when running hardy in KVM on a clean install using vmbuilder?
[13:42] <incorrect> I want to apply some performance tuning options i've been setting in /sys/ is there something like sysctl.conf where i can store them?
[13:45] <sommer> incorrect: you're probably looking for /etc/sysctl.conf
[13:49] <incorrect> i don't see how i can map stuff from /sys/ i don't see a mapping in /proc/sys
[13:58] <sommer> incorrect: the net.ipv4.ip_forward option for example maps to /proc/sys/net/ipv4/ip_forward
[14:22] <storrgie> I need to set up a machine remotely with ubuntu-desktop installed on it, and be able to vnc in.... has anyone done this?
[14:33] <billybigrigger> without ssh how would you have remote access?
[14:33] <billybigrigger> im pretty sure you need physical access for that one
[14:40] <a1fa> what was that benchmark tool for hd ?
[14:40] <fbc-mx> ls
[14:41] <fbc-mx> How can I make my virtual interfaces come up automatically on my server? I also need to logon and ifup them before I start my apache, and I'd like for them to start up automatically.
[14:45] <a1fa> edit your interfaces file
[14:46] <a1fa> auto eth0.1
[14:46] <a1fa> or whatever
[14:57] <fbc-mx> it is so wierd. even when that interface is up I can't ping it.
[14:58] <a1fa> somebody recommended a ++ something benchmark yesterday
[14:59] <_ruben> bonnie++
[15:18] <a1fa> _ruben : thanks
[16:17] <ivoks> kirkland: there's one really cool tool (iptraf) that requires bigger window than the one that's left with byobu; so, someone should look into that
[16:22] <josephpiche> not sure if this is the right place for this, but i got a question: how do i check the status of services via cli? and is there an api i could use to connect to this via python or something?
[16:23] <aruetten> josephpiche: what do you meen with service? you kann use ps to see what process are there
[16:23] <josephpiche> wow, that sounded really newbish... i mean what is the "correct" way to do it i.e. not via calling "/etc/init.d/xxx status" via python
[16:23] <ivoks> invoke-rc.d cron status
[16:24] <josephpiche> thank you very much
[16:24] <hggdh> or service cron status
[16:25] <josephpiche> hmm... is one better than the other?
[16:25] <hggdh> they pretty much do the same thing
[16:26] <hggdh> service clean up the environment before executing
[16:27] <orogor> hi here
[16:27] <josephpiche> so, would you recommend using one over the other in, say, a panel applet?
[16:27] <orogor> anyobne know if it s possible to go from a single drive install to a raid 10 install without reinstalling ?
[16:28] <Hecate> orogor, hw or sw raid? i assume the latter.
[16:28] <orogor> yes sw raid
[16:28] <orogor> else it would be at least less an issue
[16:30] <Hecate> i'm sure it can be done. the how is probably the issue. ;)
[16:30] <Hecate> gimme a minute
[16:31] <orogor> there s already lvm running
[16:31] <Hecate> you got a seperate boot-partition? otherwise a raid-level that includes a zero will be difficult/impossible?
[16:31] <Hecate> *.
[16:31] <orogor> also right now i have 2 drives diamond max 10, i can put all the data on a single drive, then ill add 2 diamond max 21
[16:32] <orogor> i could spare a 100MB partition, it s not an issue
[16:33] <Hecate> you'll need a 5th drive to get the raid10 set up.
[16:35] <hggdh> josephpiche, I would use 'service'
[16:37] <Hecate> this is how i'd do it: 1) get 4 empty drives 2) boot the box from a live-cd/-usb 3) create 2 partitions on the 4 empty drives (1 GiB for boot and the rest for the lvm's data-stripe)
[16:37] <Hecate> ^^ orogor
[16:37] <Hecate> _|__
[16:38] <josephpiche> hggdh: thanks
[16:38] <Hecate> orogor, 4) create a raid1 (copy) on all 4 small partitions, format it with ext2, install grub (QUESTION: Can grub boot from raid10?)
[16:39] <Sam-I-Am> Hecate: just not raid 5 iirc
[16:39] <josephpiche> hggdh: i need an applet for doing simple status, start, stop, restart actions on a couple services and couldn't find any, so I figured I should just write on
[16:39] <orogor> humm
[16:41] <Hecate> 5) create a raid10 on the remaining partitions, create a new lvm inside that raid, copy the data into it 6) get the system ready to boot: mount the system inside the lvm, chroot into it, adjust the mdadm.conf, modprobe.conf 7) update the initramfs 8) give it a shot
[16:42] <Hecate> thanks Sam-I-Am
[16:42] <orogor> hufff
[16:42] <hggdh> josephpiche, I guess you can easily wrap a shell script (or perl, or python) around either service or invoke-rc.d
[16:43] <hggdh> both of them will accept any parameters to be sent to the real init script being called
[16:43] <Hecate> orogor, this is a very complicated thing to do, i'm afraid. and if i were you i would not tackle it, unless you either really need it or know what you're doing (i.e. you're really experienced)
[16:44] <Hecate> i just hope a) my explanations are correct and b) you did understand them ;)
[16:44] <orogor> experienced, but not with raid
[16:44] <Hecate> orogor, maybe you could give the entire thing a dry run using a virtual machine?
[16:45] <Hecate> or maybe you just want to start with sth a little less difficult. for example turning the entire thing into a raid1?
[16:46] <josephpiche> hggdh: interesting. thanks for the advice
[16:46] <Hecate> orogor, but before you start doing anything, verify that grub can boot from raid10. it does for raid1, since i use that myself.
[16:46]  * Hecate is afk now. Got to fix himself sth to eat.
[16:46] <orogor> Hecate, going to buy the disks before the store close, brb
[16:47] <orogor> Hecate, that must somehow works, peoples do it , i worry more for theswitch itself
[17:14] <a1fa> hostname,9632M,40107,95,43868,21,22237,7,30215,58,64024,10,457.4,0,16,+++++,+++,+++++,+++,+++++,+++,+++++,+++,+++++,+++,+++++,+++
[17:14] <a1fa> is this any good
[17:14] <a1fa> who can interpret bonnie data
[17:15] <orogor> nobody :)
[17:15] <orogor> i never understood it
[17:15] <Sam-I-Am> its fields
[17:15] <Sam-I-Am> they're well documented
[17:16] <a1fa> url?
[17:16] <Sam-I-Am> man page
[17:17] <a1fa> ;P
[17:17] <a1fa> no
[17:18] <Sam-I-Am> no?
[17:18] <mattt> bonnie can spit out html, no?
[17:18] <Sam-I-Am> it should have output the field names above the data
[17:18] <a1fa> nah
[17:18] <a1fa> i want to compare it to other results
[17:19] <a1fa> sorry
[17:19] <a1fa> to ther setups by other people
[17:19] <Sam-I-Am> http://www.coker.com.au/bonnie++/readme.html
[17:19] <Sam-I-Am> that was hard using google :)
[17:20] <a1fa> hm
[17:21] <a1fa> 39mb/s sucks
[17:22] <a1fa>  Timing buffered disk reads:   86 MB in  3.03 seconds =  28.39 MB/sec
[17:22] <a1fa> lol
[17:22] <a1fa> ghey
[17:23] <Sam-I-Am> sounds a tad slow
[17:23] <Sam-I-Am> then again, there are some craptastic raid controllers out there
[17:23] <Sam-I-Am> like dell perc3
[17:32] <mathiaz> kirkland: are you writing up the minutes of today's ubuntu-server meeting?
[17:32] <kirkland> mathiaz: heh :-)
[17:33] <kirkland> mathiaz: i'm pretty slammed right now, do you mind?
[17:33] <mathiaz> kirkland: I'll try to get to it today then
[17:37] <kirkland> mathiaz: cool, thanks.
[19:18] <resno> Hello, I am trying set up a virtual server. And I am not sure what to do.
[19:24] <resno> I need help configuring my server to be a vps.
[19:44] <Hecate> resno, what exactely do you wanna do?=
[19:45] <resno> i have django running and i would like to setup php to run.
[19:45] <resno> the best way i can think to do it is with two seperate vpses
[19:46] <billybigrigger> hey all
[19:47] <billybigrigger> anyone here run a virtual server inside of vbox?
[19:47] <billybigrigger> im trying to run a copied disk and created a new machine...
[19:47] <billybigrigger> everything works good
[19:47] <billybigrigger> except my networking
[19:47] <billybigrigger> i have guest additions installed, nat set for network type in vbox, and i still get nothing
[19:47]  * Nafallo thought vbox was some graphical thing :-)
[19:48] <billybigrigger> yeah i run my www/mysql/mail services through it though
[19:48]  * resno agrees with Nafallo
[19:48] <billybigrigger> better than having 2 computers running
[19:48] <billybigrigger> and nicer than combining server apps/services with desktop
[19:49] <billybigrigger> so i run the server inside a vm
[19:49] <billybigrigger> whats wrong with that?
[19:53] <Hecate> what's django?
[20:04] <KillMeNow> anyone here know of a better way to update your Ubuntu 8.04lts to 9 w/o having to download the install ISO?
[20:06] <katakaio> Have you attempted a dist-upgrade?
[20:06] <sbeattie> KillMeNow: do-release-upgrade doesn't work for you?
[20:06] <KillMeNow> yea, i've done the dist-upgrade whenver i see a new kernel update
[20:07] <KillMeNow> haven't tried the do-release-upgrade yet
[20:07] <KillMeNow> lemme try that
[20:07] <KillMeNow> course, lemme back up everything first
[20:07] <RoAkSoAx> KillMeNow, https://help.ubuntu.com/community/UpgradeNotes
[20:20] <alex-weej> halp
[20:20] <ivoks> hulp
[20:20] <alex-weej> we have an LDAP server which we're using for authentication etc. on a few hosts
[20:20] <alex-weej> we have a server which is in a DMZ which we want to restrict "access" to
[20:20] <resno> how do i setup a virtual server?
[20:20] <alex-weej> anyone know the best way to do that?
[20:21] <alex-weej> posix group membership?
[20:22] <henkjan_> aan de moezel
[20:22] <henkjan_> err, E_WINDOW
[20:23] <alex-weej> i can limit ssh access to some groups, but that still lets you log in at the seat
[20:23] <alex-weej> i need a more general way of limiting login to some group
[20:48] <Sam-I-Am> alex-weej: netgroups
[20:59] <\sh> anyone using pxe booting linux kernels and using root nfsmounts ?
[21:00]  * \sh sees very strange dhcp request behaviour of ipconfig (klibc)...waiting too long and not sending directly dhcp requests to the server..an ipconfig -t 360 helps here..but this is so insane
[21:00] <\sh> (pxe dhcp requests are working like a charm)
[21:01] <\sh> switch runs on portfast mode for all native trunk ports...and we know it works
[21:02] <\sh> the only candidate who can fck up this setup looks like ipconfig during initramfs
[21:10] <AtomicSpark> So I'm a staff member on Halflife2.net and our website goes down when there is a spike in traffic. The summary is here http://pastie.org/private/enx6ksykaueil5qa6d7ug Is there anyway we can tweak a default LAMP install? Would using lighttpd make this better?
[21:11] <ivoks> or you could configure the system?
[21:12] <ivoks> default config is just that - default config
[21:12] <ivoks> it doesn't cover all the scenarios
[21:12] <ivoks> if you have big iowait, don't allow so many connections
[21:12] <RoyK> AtomicSpark: I guess that depends on where the bottleneck is located
[21:12] <ivoks> or buy faster disks
[21:13] <ivoks> if it's an php app, you should also consider eaccelerator
[21:13] <ivoks> and biger buffers for mysql
[21:13] <AtomicSpark> I personally don't have access to the server. I'm just looking for tips to give him.
[21:13] <\sh> AtomicSpark: remove all apache mods from the installation which you don't need...adjust the fork config in apaches configuration...you need stress your apache config to get good values...and increase your memory...2GB for mysql with a php app in front of it, is not enough...
[21:14] <ivoks> high iowait is... bad luck
[21:14] <\sh> AtomicSpark: well, another option may be to pay a sysadmin a good salary  ;)
[21:14] <ivoks> basicaly, you reached the limits
[21:14] <ivoks> you could try fix it with bigger buffers for mysql and apache
[21:14] <AtomicSpark> It is a high traffic website. I'm surprized it's on such a weak system.
[21:15] <ivoks> but with 2GB you don't have lots of space...
[21:15] <ivoks> AtomicSpark: what disks are there?
[21:15] <ivoks> probably sata
[21:16] <AtomicSpark> I'm not sure. I'd assume SATA, but the website has been around for awhile. Might be SCSI.
[21:16] <ivoks> or IDE :)
[21:16]  * AtomicSpark shivers
[21:16] <ivoks> if you get iowait with scsi, then you have serious problems
[21:17] <ivoks> you'll have to cut on number of allowed connections
[21:17] <ivoks> and fix the application
[21:17] <ivoks> it's hard to talk about it without real numbers
[21:20] <AtomicSpark> True. Well thanks for the advice. I'll forward it on to him.
[21:21] <a1fa> who has ubuntu/debian boot cd that can do dell firmware updates?
[21:22]  * \sh doesn't use dell hw...hp has some debian packages for x86 / x86_64
[21:24]  * \sh is gone home
[21:25] <AtomicSpark> a1fa: If dell comes with a firmware maintanince cd like HP does, just use it.
[21:25] <KillMeNow> i guess Ubuntu is a microsoft product now:  http://linuxologist.com/linuxhumor/ubuntu-is-a-microsoft-product-now/
[21:29] <a1fa> KillMeNow
[21:29] <KillMeNow> yes?
[21:30] <a1fa> thats funny
[21:31] <KillMeNow> isn't it though?
[21:31] <KillMeNow> Dell still hasn't corrected the mistake
[22:47] <dinger1986> does anyone know what command you use with cp to not overwrite
[22:50] <KillMeNow> cp -n
[22:50] <dinger1986> i just tried that it says invalid option
[22:51] <Tom_Ass> what port should I open in ufw to make my nfs shares accessible?
[22:53] <jpds> Tom_Ass: 2049
[22:54] <KillMeNow> i suppose you can try to use the --update option
[22:54] <KillMeNow> it will only overwrite if the file is newer, but there is always rsync --ignore-existing
[22:57] <dinger1986> ah rsync sounds ok
[22:58] <dinger1986> sweet works a wee treat
[23:04] <KillMeNow> glad to help dinger
[23:10] <Tom_Ass> thanks, jpds, but it doesn't seem sufficient
[23:12] <KillMeNow> are you running IPTables?
[23:14] <Tom_Ass> KillMeNow, yes
[23:15] <KillMeNow> try flushing your iptables and then see if NFS can mount
[23:16] <KillMeNow> if you can mount it after IPTables are flushed and empty then it's the IPTables that's killing the connection