1278633872 Q * dowdle Remote host closed the connection
1278634804 J * derjohn_foo ~aj@d003199.adsl.hansenet.de
1278635121 Q * PowerKe Ping timeout: 480 seconds
1278635236 Q * derjohn_mob Ping timeout: 480 seconds
1278636352 Q * MeCooL 
1278637555 J * imcsk8 ~ichavero@189.231.5.18
1278637737 J * PowerKe ~tom@d5153A6A9.access.telenet.be
1278638018 Q * imcsk8 Quit: This computer has gone to sleep
1278638041 J * imcsk8 ~ichavero@189.231.5.18
1278638392 Q * imcsk8 Quit: This computer has gone to sleep
1278638934 J * imcsk8 ~ichavero@189.231.5.18
1278640086 Q * imcsk8 Quit: This computer has gone to sleep
1278642946 Q * FireEgl Quit: Leaving...
1278643036 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1278643163 J * FireEgl FireEgl@Sebastian.Tcldrop.US
1278644026 Q * cehteh Ping timeout: 480 seconds
1278644088 J * imcsk8 ~ichavero@189.231.5.18
1278644295 Q * imcsk8 
1278644434 J * SauLus_ ~SauLus@d025200.adsl.hansenet.de
1278644434 Q * SauLus Read error: Connection reset by peer
1278644449 N * SauLus_ SauLus
1278645926 J * cehteh ~ct@pipapo.org
1278654994 J * ncopa ~ncopa@180.40.189.109.customer.cdi.no
1278655178 J * petzsch ~markus@dslb-092-078-148-121.pools.arcor-ip.net
1278656445 Q * derjohn_foo Ping timeout: 480 seconds
1278656739 Q * petzsch Quit: Leaving.
1278658131 J * derjohn_foo ~aj@213.238.45.2
1278658733 J * petzsch ~markus@dfn131.rz.tu-ilmenau.de
1278658867 Q * petzsch 
1278659237 J * BenG ~bengreen@cpc6-aztw22-2-0-cust100.aztw.cable.virginmedia.com
1278659391 J * petzsch ~markus@dfn131.rz.tu-ilmenau.de
1278661011 J * barismetin ~barismeti@zanzibar.inria.fr
1278661320 J * ktwilight_ ~keliew@91.176.225.94
1278661716 Q * ktwilight Ping timeout: 480 seconds
1278662081 J * ktwilight ~keliew@91.176.211.29
1278662154 Q * ktwilight_ Read error: Connection reset by peer
1278663263 Q * Piet Remote host closed the connection
1278663320 J * Piet ~Piet__@7GDAADNK5.tor-irc.dnsbl.oftc.net
1278663681 N * click_ click
1278663812 Q * petzsch Quit: Leaving.
1278663984 J * tanjix tanjix@dslb-084-059-035-185.pools.arcor-ip.net
1278663998 M * tanjix hello @ll
1278664021 M * tanjix is there a way to receive an output of all used (open) ports from one context?
1278664042 M * tanjix without entering the guest
1278664833 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d
1278664993 J * petzsch ~markus@dfn131.rz.tu-ilmenau.de
1278665591 Q * petzsch Quit: Leaving.
1278665769 J * petzsch ~markus@dfn131.rz.tu-ilmenau.de
1278668355 Q * BenG Quit: I Leave
1278669285 Q * petzsch Quit: Leaving.
1278669905 J * thierryp ~thierry@zankai.inria.fr
1278670282 J * petzsch ~markus@dslb-092-078-148-121.pools.arcor-ip.net
1278670876 Q * swenTjuln Ping timeout: 480 seconds
1278672508 J * imcsk8 ~ichavero@evdomip-67-179.iusacell.net
1278672996 N * Bertl_zZ Bertl
1278673012 M * Bertl morning folks!
1278673032 M * Bertl tanjix: you mean a list?
1278673096 M * ard maybe he means a netstat -nl or ss -nl
1278673157 M * Bertl probably, well, depends on the guest setup, such a list can be gathered by entering the spectator context(processes) and the network context or network namespace
1278673197 M * Bertl (or by entering the guest process context, instead of the spectator context)
1278673245 M * ard I think tanjix doesn't know that there are more ways to get into the right context without doing a complete vserver .... enter
1278673356 M * ard this original util-vserver has a *very* *old* debian/rules, I know, because it looks like my original debian/rules ;-)
1278673392 Q * petzsch Quit: Leaving.
1278674302 J * ktwilight_ ~keliew@91.176.233.224
1278674532 Q * ktwilight Ping timeout: 480 seconds
1278674714 J * ktwilight__ ~keliew@91.176.139.70
1278675097 M * ard compiling... woei
1278675155 Q * ktwilight_ Ping timeout: 480 seconds
1278675840 Q * hijacker Quit: Leaving
1278675844 J * hijacker ~hijacker@213.91.163.5
1278675969 M * ard Almost...
1278675973 M * ard ard@lennydev64:~/work/vserver/svn/trunk$ dpkg -c ../util-vserver_0.30.216-pre2896_amd64.deb |wc -l
1278675973 M * ard 6
1278676372 M * ard ard@lennydev64:~/work/vserver/svn/trunk$ dpkg -c ../util-vserver_0.30.216-pre2896_amd64.deb |wc -l
1278676372 M * ard 449
1278676375 M * ard better :-)
1278677956 M * ard so
1278677981 M * ard And now I am going to shoot myself... in the lxc experiments I also experimented with the cgroup mount
1278677989 M * ard #vserver    /dev/cgroup    cgroup   ns,cpuset,cpu,cpuacct,devices,freezer,net_cls,memory    0 0
1278677990 M * ard vserver    /dev/vcgroup    cgroup   rw,relatime,blkio,net_cls,freezer,devices,memory,cpuacct,cpu,cpuset    0 0
1278677996 M * ard grrrr
1278678034 A * ard changed /dev/cgroup into /dev/vcgroup to try to mount it with an lxc compliant /dev/cgroup
1278678086 M * Bertl but then it won't be Linux-VServer compliant :)
1278678110 M * ard yes :-). So I just stopped trying to combine them... /me ditched lxc
1278678155 Q * manana Remote host closed the connection
1278678191 M * ard Hmmmm, we made a bug in svn yesterday
1278678215 M * ard when spaces/net is empty, it doesn't work anymore
1278678299 M * Bertl probably easy to fix
1278678304 M * ard yes :-)
1278678312 M * ard vserver.functions
1278678393 M * ard Bertl : the function getFileValue, will that b0rk/bail out when the file is empty?
1278678463 M * ard if you don't know from head, I am already searching
1278678723 M * ard yes, it will return a != 0
1278678744 M * ard or better: read will return non-zero if end-of-file is reached without input
1278678837 M * ard alright: instead of an empty spaces/net, I now have a "" in spaces/net , which reads as an empty string
1278678943 M * ard ownoes !
1278678953 M * ard  /usr/sbin/vspace --enter '""' --net
1278678955 M * ard never mind
1278678967 M * ard I will put a test-if-empty-file around :-)
1278678988 M * Bertl good idea, or even better, test for existance
1278679034 M * ard well, it should exist, but can contain either a value, or nothing. In the later case the getFileValue fails
1278679048 M * Bertl why can it contain nothing?
1278679063 M * ard since r2894 yes :-)
1278679090 M * ard spaces/net contains the xid of the "parent" vserver with which it shares the network namespace :-0
1278679125 M * ard so when it is empty it now fails, but if nobody uses it yet, we might just as well make it illegal
1278679144 M * ard and the "parent" vserver should have 0 to make the initial network namespace
1278679147 M * Bertl what is the purpose of having an empty file there?
1278679175 M * ard the previous version created a network namespace for just that vserver
1278679201 M * Bertl which makes sense in a typical setup
1278679220 M * ard But I wanted to have a dummy network namespace vserver, and have different vservers belong to the same network namespace
1278679225 J * ghislain ~AQUEOS@adsl2.aqueos.com
1278679248 M * Bertl so why not do it like with network contexts then?
1278679249 M * ard So I can have multiple vservers per vlan, and for each vlan a seperate ipstack
1278679268 M * ard well, that's the part daniel_hozac was fixing :-)
1278679281 M * Bertl same number means same context ... for network namespaces, same name means same space, no?
1278679306 M * ard You mean like cgroup/name ?
1278679318 M * Bertl for example
1278679323 M * ard you want to share the ipstack, but not the remainder...
1278679357 M * ard f.i. you can have multiple vservers in one vlan that talk to a vserver in another vlan, that might or might not be on another host
1278679370 M * Bertl same ist true for the network contexts, no?
1278679387 M * ard you mean without network namespace?
1278679389 P * thierryp ciao folks
1278679433 M * Bertl look, network contexts (identified by nids) and process contexts (identified by xids) are orthogonal
1278679446 M * ard ah :-)
1278679466 M * Bertl i.e. you can put two processes in the same xid but different nids and vice versa
1278679467 M * ard and then we add another dimension with multiple ip-stacks
1278679487 M * Bertl yep, same should be doable with any namespace
1278679529 M * ard I combine the security of network contexts with network namespaces ;-)
1278679549 M * ard I want *both* to work :-)
1278679552 M * Bertl no problem there either
1278679571 M * Bertl a process can belong to a network namespace and a network context
1278679587 M * ard well, that's the part daniel was fixing yesterday in util-vserver :-)
1278679597 M * ard and it's almost working :-)
1278679621 M * ard I mean, as in out of the box normal util-vserver functionality...
1278679859 M * ard daniel_hozac :
1278679863 M * ard         shared=""
1278679863 M * ard         test -s "$d"/net && getFileValue shared "$d"/net
1278679957 A * ard also added:
1278679958 M * ard         if test -z "$shared" || [ "$shared" -eq 0 ]; then
1278680070 M * ard yes!
1278680075 M * ard it works :-)
1278680130 M * ard Bertl : the vserver <something> enter now also works with the correct network namespace
1278680156 M * Bertl congrats!
1278680156 M * ard only the interface/ip setup still happens in the wrong network namespace, but I can live with that.
1278680172 M * ard Well, it was daniel_hozac that actually made it happen...
1278680196 M * ard and I have packaged it in your favorite distro format... .deb :-)
1278680348 M * ard only the starting and the stopping of the vserver must still be done with  vspace -e "dummyvserver" --net --
1278680605 J * mnemoc ~amery@shell.opensde.net
1278680920 M * ard daniel_hozac : http://paste.linux-vserver.org/16013 against r2896
1278680929 Q * arekm Ping timeout: 480 seconds
1278681225 J * BenG ~bengreen@cpc6-aztw22-2-0-cust100.aztw.cable.virginmedia.com
1278681794 Q * BenG Quit: I Leave
1278682115 J * arekm arekm@carme.pld-linux.org
1278683348 J * petzsch ~markus@p4FF45962.dip.t-dialin.net
1278684751 J * dna ~dna@p54BCA55B.dip0.t-ipconnect.de
1278684779 Q * petzsch Quit: Leaving.
1278685337 J * petzsch ~markus@p4FF47183.dip.t-dialin.net
1278685833 Q * Piet Write error: connection closed
1278686082 J * Piet ~Piet__@7GDAADNXD.tor-irc.dnsbl.oftc.net
1278687961 J * dowdle ~dowdle@scott.coe.montana.edu
1278689077 M * ard daniel_hozac : there must be an option to disable/enable network contexts combined with network namespaces
1278689124 M * ard and the [ "$shared" -eq 0 ] can better be left out :-(
1278689195 M * ard for now I have moved the disabling of the chbind to the part where new network namespaces are created
1278689244 A * ard wonders one thing:
1278689290 M * ard is the network context per network namespace? Not that it really matters as far as I can see. I thought it was a global thing.
1278689619 M * Bertl it will be global, if the nid is unique, but it will operate ontop of the network namespace
1278689659 M * Bertl nap attack ... bbl
1278689663 N * Bertl Bertl_zZ
1278689666 M * ard O/~
1278690304 Q * imcsk8 Quit: This computer has gone to sleep
1278692779 Q * derjohn_foo Remote host closed the connection
1278693701 J * derjohn_mob ~aj@213.238.45.2
1278694211 Q * derjohn_mob Ping timeout: 480 seconds
1278694615 Q * ncopa Quit: Ex-Chat
1278694718 Q * petzsch Quit: Leaving.
1278694720 J * petzsch ~markus@p4FF47183.dip.t-dialin.net
1278695201 Q * petzsch Ping timeout: 480 seconds
1278695987 J * imcsk8 ~ichavero@evdomip-67-179.iusacell.net
1278696203 J * petzsch ~markus@p4FF47183.dip.t-dialin.net
1278696375 Q * petzsch 
1278697742 Q * ghislain Quit: Leaving.
1278697904 Q * barismetin Remote host closed the connection
1278698115 N * Bertl_zZ Bertl
1278698119 M * Bertl back now ...
1278698205 M * ard aargh...
1278698210 M * ard BUG: unable to handle kernel NULL pointer dereference at 0000000000000028
1278698236 M * Bertl upload the complete trace
1278698245 M * ard never mind (for you) :-)
1278698255 M * ard it's a bug in bridging in 2.6.33
1278698277 A * ard was waiting for the system to reboot to exercise more tests with the vserver setups
1278698291 M * ard and the bridging stuff are leftovers from.... lxc :-)
1278698321 M * Bertl so I presume it is fixed in more recent kernels upstream?
1278698342 M * ard I guess so :-)
1278698355 M * ard I can crash 2.6.33 with lxc setup
1278698360 M * ard 2.6.34 seems to work
1278698387 M * ard but even then... there have been some changes in the structures of these "virtual" layered devices
1278698395 M * Bertl a recent 2.6.33?
1278698429 J * petzsch ~markus@p4FF47183.dip.t-dialin.net
1278698464 M * ard Haven't tried one yet... I am now at 2.6.33.3 that works pretty well if you don't use bridges on vlans on ethernet devices
1278698484 M * ard that was a step down from bridges on vlans on bonds on ethernet
1278698530 M * ard Once I get this vserver setup working I will iron out any kernel bugs related to networking
1278699622 Q * petzsch Quit: Leaving.
1278699868 M * fback 1
1278700318 M * ard 2
1278700696 J * petzsch ~markus@p4FF47183.dip.t-dialin.net
1278700777 J * manana ~mayday090@84.17.25.144
1278702009 M * ard It's that time again
1278702013 M * ard (to go home)
1278702015 M * ard O/~
1278702028 J * thierryp ~thierry@82.226.190.44
1278702082 Q * thierryp Remote host closed the connection
1278702738 Q * http203 Quit: Leaving
1278703913 J * ghislain ~AQUEOS@adsl2.aqueos.com
1278703937 Q * petzsch Quit: Leaving.
1278703946 Q * Piet Ping timeout: 480 seconds
1278704069 J * Kermit cf3840c2@ircip1.mibbit.com
1278704115 M * Kermit I have a question I would like to know if anyone would like to help me
1278704207 M * Kermit I would like to know if there is a way, like with memory, to limit the number of CPU's a virtual machine can actually see?  I have scheduling down pat, but I want the server to look like it only has 2 processors, when in fact the host has 8.  Can this be achieved?
1278704263 M * Kermit Anyone?
1278704373 M * Bertl sure, cpu sets should allow that
1278704470 M * daniel_hozac you need a patch though to hide the other CPUs.
1278704913 M * Kermit Okay, awesome
1278704920 M * Kermit So can I find this on the webpage?
1278705054 M * Kermit What I would ultimately like is to show two CPU's in cpuinfo, and also to show 100% utilization on the virtual box, while showing 25% on the host, make sense?
1278705128 M * Bertl well, if two cpus are assigned to a guest, and those two cpus are used to 100% ...
1278705157 M * Kermit Is that all done through cgroups?
1278705265 M * Kermit I am having a hard time finding how to do all this in the documentation.  I saw cpu_set in cgroup, but I was not able to mount cgroup, which I think has something to do with that patch Daniel was talking about?
1278705276 M * daniel_hozac nope
1278705284 M * daniel_hozac the patch is only to hide CPUs not in the cpuset.
1278705343 M * Kermit Okay, so the patch will hide the cpu's, but with scheduling, it will still only look like it is using 12.5% of the total CPU?  assuming you use one processor?
1278705376 M * daniel_hozac i'm not sure what you mean.
1278705389 M * Bertl the question is, how do you 'look' at the cpu usage :)
1278705462 M * Kermit Well when I run top, and peg the CPU's, I see 12.5% cpu utilization, which is 1/8th of the processing power of 8 processors.  So I want to see 1 CPU in /proc, and when pegged I would like to see 100% utilization, is this achievable?
1278705490 M * Kermit Using top at the moment
1278705608 M * daniel_hozac http://www.bullopensource.org/cpuset/source.html
1278705610 M * daniel_hozac has the patch
1278705663 J * petzsch ~markus@p4FF47183.dip.t-dialin.net
1278705689 M * daniel_hozac albeit for very old kernels.
1278705744 M * Kermit That is okay, I have many Linux "experts" at my fingertips
1278705766 M * Kermit By the way love this software, Daniel, yer one of the lead developers are you not?
1278705798 M * Bertl good, then I'd suggest to 'let' them port the patch to your kernel, and I guess once the other cpus are hidden, top will report back properly
1278705843 M * daniel_hozac yes, as is Bertl.
1278705860 M * Kermit If setup with scheduling, or once hidden the vserver will only have access to those cpu's?
1278705900 M * daniel_hozac the guest will only have access to those CPUs if you setup a cpuset.
1278705906 M * daniel_hozac with the patch, they will also be hidden.
1278705915 M * daniel_hozac (the ones it doesn't have access to)
1278705940 M * Kermit Okay, so I am getting a little confused
1278705959 M * Kermit Can you point me to directions on setting up cpuset?
1278705993 M * daniel_hozac Documentation/cpusets.txt in your favorite kernel tree.
1278706016 M * daniel_hozac well, i suppose it is Documentation/cgroups/cpusets.txt these days.
1278706067 M * daniel_hozac then check the great flower page for the util-vserver part of it.
1278706079 M * Kermit cgroups, okay, that was where I found cpusets before.
1278706101 M * Kermit I had issues though, I can't mount cgroup, it says cgroup is not a valid filesystem, am I missing something?
1278706114 M * daniel_hozac likely the kernel options for it.
1278706140 M * Kermit Okay, so I can look into that, thank you for being so helpful, I am no guru, but I do understand what you are saying now, thank you so much
1278706179 M * Guy- is there a known reason why I shouldn't use 2.6.34.1-vs2.3.0.36.30.4.pre8?
1278706197 M * Guy- the "pre" is making me somewhat uneasy :)
1278706639 J * BenG ~bengreen@cpc6-aztw22-2-0-cust100.aztw.cable.virginmedia.com
1278706703 J * BenG_ ~bengreen@cpc6-aztw22-2-0-cust100.aztw.cable.virginmedia.com
1278706708 Q * BenG_ Remote host closed the connection
1278706825 M * Kermit So do I set the cpu sets up on the virtual machine, in the kernel or the host?
1278707045 Q * BenG Quit: I Leave
1278707358 Q * Kermit Quit: http://www.mibbit.com ajax IRC Client
1278709887 J * harobed ~harobed@arl57-1-82-231-110-14.fbx.proxad.net
1278710524 Q * ghislain Quit: Leaving.
1278710620 Q * dna Quit: Verlassend
1278712503 Q * petzsch Quit: Leaving.
1278715005 Q * harobed Ping timeout: 480 seconds
1278716049 Q * bonbons Quit: Leaving
1278716398 Q * dowdle Remote host closed the connection
1278717921 Q * FireEgl Remote host closed the connection
1278718601 J * FireEgl FireEgl@173-16-9-10.client.mchsi.com
1278719881 Q * FireEgl Ping timeout: 480 seconds