1308187610 M * Bertl off to bed now ... have a good one everyone!
1308187615 N * Bertl Bertl_zZ
1308188798 J * ichavero__ ~ichavero@148.229.1.11
1308189003 Q * ichavero_ Ping timeout: 480 seconds
1308189961 Q * ichavero__ Ping timeout: 480 seconds
1308190189 J * nou Chaton@causse.larzac.fr.eu.org
1308191645 Q * ryker Quit: Leaving.
1308192378 J * ryker ~Adium@c-76-16-115-27.hsd1.in.comcast.net
1308193526 J * fichte` ~fichte@bashpipe.de
1308193740 Q * FIChTe Ping timeout: 480 seconds
1308193955 Q * ryker Quit: Leaving.
1308197936 J * sannes ~ace@cm-84.209.81.224.getinternet.no
1308199478 Q * Hunger Read error: Connection reset by peer
1308199482 J * Hunger- ~Hunger@Hunger.hu
1308204416 Q * quasisane Ping timeout: 480 seconds
1308204922 J * jaqm ~jaqm@223.121.217.87.dynamic.jazztel.es
1308204941 J * ghislain ~AQUEOS@adsl2.aqueos.com
1308205300 Q * derjohn_mob Ping timeout: 480 seconds
1308206518 J * derjohn_mob ~aj@213.238.45.2
1308206970 J * petzsch ~markus@dslb-092-078-236-175.pools.arcor-ip.net
1308210605 Q * petzsch Quit: Leaving.
1308216517 J * petzsch ~markus@dslb-092-078-236-175.pools.arcor-ip.net
1308218495 N * Hunger- Hunger
1308219301 J * thierryp ~thierry@zanzibar.inria.fr
1308219311 J * kir ~kir@swsoft-msk-nat.sw.ru
1308219992 Q * petzsch Quit: Leaving.
1308221562 N * Bertl_zZ Bertl
1308221566 M * Bertl morning folks!
1308221647 M * WMP hello ;)
1308224692 M * Guy- WMP: to rename a guest, stop it, grep for its name under /etc/vserver/guestname, modify the affected files, rename the rootdir, rename the /etc/vservers/guestname directory, and look at the symlinks it contains, changing them as necessary (it will be obvious); you can then start it again
1308224711 M * Guy- (under the new name, of course)
1308224734 M * WMP so i make path co vserver-utils ;)
1308224739 M * WMP patch*
1308225426 M * Bertl daniel_hozac: the ipv6 changes in your patch, are they handwaving or are they tested/known to work?
1308225439 M * Bertl (3.0-rc3)
1308225460 M * daniel_hozac i haven't tested anything yet :)
1308225472 M * daniel_hozac they are what was needed to make it build.
1308225519 P * kir Leaving.
1308225542 M * Bertl okay
1308226201 M * Bertl what do we need the VXF_STATE_SETUP check in cap_capable?
1308226237 M * daniel_hozac i'm not sure. that was just merging of the reject
1308226259 M * daniel_hozac i'm not entirely sure when cap_capable gets invoked...
1308226270 M * daniel_hozac but we need to be able to do privileged calls during SETUP for sure.
1308226285 M * Bertl okay
1308226408 M * Bertl I've merged the following: http://vserver.13thfloor.at/ExperimentalT/delta-bertl-dhozac.diff
1308226421 M * Bertl but I have to check the short->int part in inode
1308226445 M * Bertl will upload an updated patch shortly so that we have a common base
1308226590 M * daniel_hozac changing the existing flags to short seemed worrisome.
1308226683 M * Bertl but still valid, as S_IXUNLINK is the largest with 8192
1308226724 M * Bertl I guess int gives better performance, but short saves a whole int for each inode in memory
1308226733 M * daniel_hozac yeah, i suppose that is true...
1308226741 M * daniel_hozac as long as we keep an eye on it.
1308226765 M * Bertl yup
1308226785 M * LuckyLuke ok given that I cannot use the latest linux-2.6.38.8-vs2.3.0.37-rc17 (due to a regression on hppa) and that the currently running 2.6.38.4-vs2.3.0.37-rc15 has some routing issues, do you have any other version to recommend that's worth trying? (I suppose the -dontuse version is not)
1308226820 M * Bertl correct, what's the problem with the hppa regression?
1308226877 M * LuckyLuke arch/parisc/kernel/built-in.o: In function `sys_call_table':
1308226877 M * LuckyLuke (.rodata+0x2a20): undefined reference to `compat_sys_clock_adjtime'
1308226877 M * LuckyLuke arch/parisc/kernel/built-in.o: In function `sys_call_table64':
1308226878 M * LuckyLuke (.rodata+0x3a20): undefined reference to `sys_clock_adjtime'
1308226887 M * LuckyLuke with 2.6.38.8
1308226903 M * LuckyLuke not present in 2.6.38.4 nor in 2.6.39, afaik
1308226926 M * LuckyLuke (that's not a linux-vserver issue, of course)
1308226947 M * Bertl and did you check what changed in arch/parisc/kernel between 2.6.38.8 and 2.6.39?
1308226981 M * Bertl as it is likely just a missing _ or minor change in name or maybe even a missing include?
1308226994 M * LuckyLuke not yet, I'm compiling vanilla 2.6.38.4 2.6.38.8 and 2.6.39 kernels to make sure of what I just said
1308227302 M * Bertl http://paste.linux-vserver.org/19994
1308227308 M * Bertl this probably fixes it
1308227581 Q * nkukard_ Quit: Leaving
1308227755 M * LuckyLuke make started, let's see. meanwhile, many thanks.
1308232271 J * ghislain1 ~AQUEOS@adsl2.aqueos.com
1308232547 Q * ghislain Ping timeout: 480 seconds
1308232582 J * ryker ~Adium@c-76-16-115-27.hsd1.in.comcast.net
1308233115 Q * ghislain1 Read error: Operation timed out
1308233135 J * ghislain ~AQUEOS@adsl2.aqueos.com
1308233717 J * imcsk8 ~ichavero@148.229.1.11
1308233989 Q * LuckyLuke Quit: trying new kernel
1308234398 Q * ryker Quit: Leaving.
1308234409 J * ryker ~Adium@c-76-16-115-27.hsd1.in.comcast.net
1308234418 Q * ryker Remote host closed the connection
1308234437 J * ryker ~Adium@c-76-16-115-27.hsd1.in.comcast.net
1308234473 Q * ryker 
1308234710 J * ryker ~Adium@c-76-16-115-27.hsd1.in.comcast.net
1308235665 J * dowdle ~dowdle@scott.coe.montana.edu
1308236813 J * petzsch ~markus@dslb-092-078-236-175.pools.arcor-ip.net
1308238385 J * LuckyLuke ~luca@host65-83-static.228-95-b.business.telecomitalia.it
1308238461 J * ghislain1 ~AQUEOS@adsl2.aqueos.com
1308238811 Q * ghislain Ping timeout: 480 seconds
1308238929 M * LuckyLuke Bertl: kernel compiled, many thanks
1308238937 M * Bertl you're welcome!
1308238941 M * LuckyLuke and, btw, it solves the routing issue of linux-vserver that I had hit
1308239662 Q * thierryp Remote host closed the connection
1308240892 M * WMP Bertl: is possible to add 2 IP to one VM?
1308240905 M * daniel_hozac or 273
1308240911 M * WMP ok ;)
1308240928 M * WMP --interface eth0:$IP/32 --interface eth1:$IP2/32 
1308240950 M * daniel_hozac if you have an eth1
1308240965 M * WMP hmmm
1308240977 M * WMP in host i have 10 alias to eth0
1308241018 M * Bertl won't help if you specify eth1 then :)
1308241201 Q * ncopa Quit: Leaving
1308241326 M * WMP ok ;)
1308241676 Q * derjohn_mob Ping timeout: 480 seconds
1308241914 J * quasisane ~sanep@c-76-24-80-97.hsd1.nh.comcast.net
1308242840 M * Bertl daniel_hozac: how do I configure a list of packages to be installed for a new guest?
1308242852 M * Bertl in this specific case a debian squeeze system
1308242930 M * daniel_hozac for debootstrap, i honestly don't know how to extend it.
1308242955 M * Bertl hmm, okay, so enter guest and apt-get whatever is missing then?
1308242965 M * daniel_hozac yeah, basically.
1308243119 M * WMP isn;t possible to set in host 250hz timing kernel and in guest 1000hz timing kernel?
1308243124 M * daniel_hozac no
1308243156 M * Bertl as there is only one kernel, you need to solve the problem to setting two different timer values on a single kernel :)
1308243158 M * WMP :(
1308243178 M * daniel_hozac why do you think you need to change the timer value at all?
1308243205 M * WMP daniel_hozac: in this same mahcine i want to have WWW/mysql and game servers
1308243262 M * Bertl and what would you do if you had to install those without Linux-VServer ?
1308243273 M * Bertl i.e. mysql and a game server on the same machine?
1308243326 M * daniel_hozac Bertl: maybe try -- --arch amd64 --include packageA,packageB,packageC?
1308243389 M * WMP Bertl: on 1000hz i have problem with mysql, but game servers work good
1308243406 M * daniel_hozac what problem do you have?
1308243423 M * WMP slowly
1308243483 M * WMP hmmm
1308243496 M * WMP now i cloning guest and all work
1308243505 M * WMP without edit context
1308245374 M * WMP hmmm, game server on Vserver have big lagg :(
1308245566 M * Bertl raise the priority and/or dedicate a cpu
1308245596 M * Bertl or simply try with vmware just to see how much faster Linux-VServer is :)
1308245757 M * WMP Bertl: i havent prioryt to cpu
1308249654 J * Piet_ ~Piet__@28IAAB1PS.tor-irc.dnsbl.oftc.net
1308250062 Q * Piet Ping timeout: 480 seconds
1308250880 J * derjohn_mob aj@tmo-076-77.customers.d1-online.com
1308251337 Q * ghislain1 Ping timeout: 480 seconds
1308254229 M * WMP vserver have default limit to CPU?
1308254239 M * Bertl nope
1308254264 M * Bertl by default, a guest is unlimited, except for dangerous stuff
1308254275 M * Bertl i.e. hardware or kernel access
1308254287 M * WMP so why this work on host fast, and in guest slow?
1308254306 M * Bertl what exactly
1308254328 M * Bertl but probably the answer is, because your guest is configured differently than the host
1308254344 M * WMP Bertl: what i shoud check?
1308254364 M * Bertl a simple test is to chroot into the guest and try the same stuff from that chroot
1308254380 M * Bertl i.e. it is then running _on_the_host_ per definition
1308254394 M * WMP ok
1308254433 M * Bertl if that is fast, and inside the guest it is still slow, then you have some misconfiguration in the guest config
1308254436 M * WMP Bertl: i shoud mount eq, procs ?
1308254460 M * Bertl if that is as slow as inside the guest, then your guest itself (i.e. the installation) is misconfigured
1308254475 M * Bertl if your application requires proc, then yes
1308254515 J * bonbons ~bonbons@2001:960:7ab:0:24bc:54b2:2ac3:7c03
1308254647 Q * derjohn_mob Ping timeout: 480 seconds
1308254725 M * WMP Bertl: and i have lags in chroot...
1308254837 M * Bertl so then the problem is not in the guest config but inside the guest installation
1308254857 M * Bertl i.e. you want to look for differences inside the guest filesystem comared to the host filesystem
1308254888 M * Bertl most common problem is nameservice related, btw, so that is a good start to check
1308255248 M * WMP i have this same distro
1308255267 M * Bertl good, then comparing the settings should be easier
1308255289 M * WMP but, my program is in java
1308255310 M * WMP and shoudnt use system configs
1308255345 M * Bertl so you 'think' it doesn't use a resolver or the glibc, I doubt that :)
1308255369 M * WMP yes
1308255377 M * WMP resolv.conf i have good
1308255385 M * WMP glibc in this same verison
1308255447 M * Bertl well, something is obviously different, otherwise the chroot (which enables you to run your java app on the host) would be as fast as the installation on the host
1308255519 M * WMP OMG!...
1308255526 M * WMP epic fail...
1308255548 M * WMP my user have turn on video on hd
1308255559 M * WMP and server work good, but his computer...
1308255629 M * Bertl so your test was flawed :)
1308255692 M * WMP yes...
1308255698 M * WMP stupid minecraft server
1308255747 J * hijacker ~hijacker@87-126-142-51.btc-net.bg
1308256146 M * LuckyLuke didn't tickless kernels remove all the problems of HZ?
1308256189 M * Bertl it removed a lot of problems, and created brand new ones :)
1308256217 M * LuckyLuke kewl
1308256516 Q * sannes Remote host closed the connection
1308256547 Q * dowdle Remote host closed the connection
1308256746 J * cary ~cary@027c89ff.bb.sky.com
1308256776 J * dowdle ~dowdle@scott.coe.montana.edu
1308256799 M * cary Can anyone answer a question about openvpn clients on linux vserver?
1308256836 M * Bertl not without somebody asking the question ...
1308256882 M * pmjdebruijn cary: in other words, just ask, and wait around to see if anybody knows
1308256893 M * cary Cool, I'm trying to create a connection to an openvpn server from a vserver running debian lenny and I'm having difficulty with the following
1308256917 M * cary I'm trying to connect to an openvpn server from a vserver
1308256929 M * cary running debian lenny.  getting the following error
1308256959 M * cary TCP/UDP: Socket bind failed on local address [undef]:1194: Address already in use
1308256978 M * pmjdebruijn which version of vserver
1308256995 M * pmjdebruijn so the vserver is your client?
1308257006 M * daniel_hozac is your host running and openvpn server as well?
1308257024 M * cary yeah, vserver is the client.  host is not running openvpn
1308257031 M * cary it's a remote server
1308257045 M * pmjdebruijn and the vserver can access the tap device?
1308257053 M * pmjdebruijn tap/tun or what does openvpn use again:?
1308257058 M * cary Hold on a sec, I'll get that version number
1308257067 M * daniel_hozac address already in use means you have something else listening on the port
1308257075 M * pmjdebruijn but the client doesn't listen?
1308257090 M * cary netstat doesn't show anything on 1194
1308257111 M * daniel_hozac does it show anything on the host?
1308257128 M * daniel_hozac (note it's most likely UDP, not TCP)
1308257176 M * cary vserver verstion vserver 0.30.216
1308257185 Q * hijacker Quit: Leaving
1308257189 M * Bertl not released yet :)
1308257212 M * Bertl so they still use openvpn in the future?
1308257223 M * pmjdebruijn cary: address already in use would suggest your running server not client
1308257278 M * cary Bertl: huh?  
1308257324 M * pmjdebruijn cary: that vserver version doesn't exist yet
1308257324 M * Bertl 0.30.216 was not released yet, so I presume you are from the future
1308257335 M * cary pmjdebruijn: perhaps, I'm doing something else wrong
1308257336 M * pmjdebruijn maybe that's util-vserver
1308257342 M * pmjdebruijn cary: what does uname -a say
1308257363 M * pmjdebruijn cary: no clue, just make sure your config declares it's a client
1308257378 M * cary Bertl:  Ahh, I see, this is the correct one then 2.6.26-2-vserver-amd
1308257396 M * pmjdebruijn uh crap
1308257399 M * pmjdebruijn the debian kernel :(
1308257430 M * pmjdebruijn isn't that ancient^2 :)
1308257439 M * pmjdebruijn cary: most folks on Debian build their own kernels
1308257447 M * cary pmjdebruijn, ahh
1308257455 M * pmjdebruijn not that that is your issue per se
1308257467 M * pmjdebruijn just wanted to briefly rant about that :)
1308257471 M * Bertl http://linux-vserver.org/Installation_on_Debian#Issues_with_Lenny.27s_2.6.26_Kernel_and_Util-vserver
1308257481 M * cary I'll look into that.
1308257493 M * pmjdebruijn cary: we have a fairly big install running on 2.6.27.53-vs 
1308257531 M * Bertl hopefully to bed updated soon too
1308257537 M * Bertl s/bed/be/
1308257579 M * cary pmjdebruijn: Let me check my openvpn client.  Bertl: I'm still a novice vserver user.  The details really escape me at the moment.
1308257606 M * pmjdebruijn cary: there is no openvpn "client" or "server"
1308257617 M * pmjdebruijn cary: it's the same program, it operates either way based on config
1308257624 M * Bertl simplified, throw the version away debian provides, compile your own recent kernel and try again
1308257630 M * pmjdebruijn cary: #openvpn will probably be more helpful on the specific of this
1308257649 M * pmjdebruijn Bertl: I'm not entirely convinced the issue is related to vserver at all
1308257676 M * cary pmjdebruijn: OKay, I'll check them out.  So there's no peculiarities related to openvpn on vserver, other than what's inthe FAQ, 
1308257685 M * pmjdebruijn I'm not sure
1308257697 M * pmjdebruijn we don't use it in vserver
1308257707 Q * jaqm Ping timeout: 480 seconds
1308257707 M * Bertl but that kernel is so broken that problems can be expected
1308257733 M * daniel_hozac as long as you've assigned a tun to the guest you're fine
1308257747 M * pmjdebruijn cary: I highly recommend the long-term maintained kernels 2.6.27.57/58 or 2.6.32.41 
1308257758 M * pmjdebruijn they offer a very stable platform
1308257776 M * daniel_hozac is 2.6.27 even maintained still?
1308257781 M * pmjdebruijn not anymore
1308257785 M * daniel_hozac so...
1308257788 M * pmjdebruijn but it was only recently dropped
1308257791 M * daniel_hozac hardly something to recommend.
1308257796 M * pmjdebruijn the point is it's very much patched
1308257800 M * pmjdebruijn and thus very stable/reliable
1308257814 M * cary pmjdebruijn:  Will definitely consider that.  Is it tough to upgrade vserver kernels?
1308257815 M * daniel_hozac won't be so reliable in a month.
1308257822 M * pmjdebruijn daniel_hozac: huh?
1308257829 M * pmjdebruijn daniel_hozac: kernels don't destabilize due to time
1308257839 M * pmjdebruijn only security issues remain, which is highly dependant of the usecase
1308257841 M * daniel_hozac when the next DoS/exploit comes around...
1308257853 M * daniel_hozac besides, those kernels are far less tested than others.
1308257863 M * daniel_hozac it's not very uncommon for things to break in the -stable series.
1308257880 M * pmjdebruijn daniel_hozac: it's much more common for things to break in not -stable series
1308257882 M * daniel_hozac because people don't actually care that much, they just submit it to -stable because, well, it's a good fix.
1308257894 M * daniel_hozac not really.
1308257906 M * pmjdebruijn daniel_hozac: we've had several use cases that run only stable with -stable kernels
1308257910 M * pmjdebruijn because something is broken
1308257915 M * pmjdebruijn be it cifs/nfs /etc
1308257932 M * pmjdebruijn the 2.6.27.x tree is the only one where everything we need works
1308257938 M * pmjdebruijn expecting the same for 2.6.32.x
1308257958 M * pmjdebruijn obviously that's very dependant on every specific use case
1308257962 M * daniel_hozac sounds like you're either doing something seriously wrong, or very horrible at reporting bugs.
1308257996 M * pmjdebruijn I am horrible at reporting NFS related bugs
1308257999 M * pmjdebruijn :)
1308258005 M * pmjdebruijn or CIFS for that matter yes
1308258036 M * pmjdebruijn daniel_hozac: also we have a reasonably sized infrastructure
1308258045 M * pmjdebruijn daniel_hozac: lots of things break, when you scale up
1308258051 M * pmjdebruijn which worked fine in smaller setups
1308258082 M * daniel_hozac unless you're scaling up to tens of thousands machines, scale is usually not an issue.
1308258148 M * pmjdebruijn heh
1308258186 M * pmjdebruijn daniel_hozac: with 10 machines you're 10 times as likely to run into $RANDOM_BUG as with 1
1308258189 M * pmjdebruijn scales linearly
1308258204 M * pmjdebruijn assuming your likely to hit it at al
1308258237 M * pmjdebruijn daniel_hozac: we've even hit a specific gcc+kernel+64bit issue in the bnx2 driver in the past
1308258315 M * pmjdebruijn the broadcom dude was amazing help though
1308258498 M * Marillion pmjdebruijn: which mainboard (bnx2)?
1308258638 M * pmjdebruijn HP ProLiant DL360 G5
1308258638 Q * Romster Read error: Connection reset by peer
1308258665 M * Marillion ah ok, thanks
1308258671 M * pmjdebruijn but that's hardly relevant
1308258679 M * pmjdebruijn since it was a driver issue
1308258683 M * pmjdebruijn not a hardware one
1308258740 M * Marillion sure, but broadcom have many chips of that
1308258771 M * pmjdebruijn yeah but it doesn't matter
1308258789 M * pmjdebruijn any bnx2 would have been affected in this particular case
1308258809 M * Marillion not for you, for me wll :)
1308258815 M * pmjdebruijn wll?
1308258821 M * Marillion wll/well
1308258833 M * pmjdebruijn it was using an old gcc
1308258837 M * pmjdebruijn so no need to worry about that now
1308258848 M * pmjdebruijn I think it happened using etch's gcc
1308258856 M * pmjdebruijn and only when building 64bit kernels
1308258857 M * Marillion gcc compile not firmware! ;)
1308259125 Q * cary Ping timeout: 480 seconds
1308260868 Q * bonbons Quit: Leaving
1308261538 Q * petzsch Quit: Leaving.
1308261625 J * derjohn_mob ~aj@d142199.adsl.hansenet.de
1308262637 M * Marillion patch-2.6.38.8-vs2.3.0.37-rc17.diff make fun, not problem at the time found
1308262744 M * Marillion Kernel looks good
1308263254 A * Marillion ...my statement should not act derogatory
1308266408 Q * ensc|w Quit: Lost terminal
1308266848 N * ensc Guest4890
1308266858 J * ensc ~irc-ensc@p5DF2F3F9.dip.t-dialin.net
1308267266 Q * Guest4890 Ping timeout: 480 seconds