1300924835 Q * dowdle Remote host closed the connection
1300924856 Q * Piet Ping timeout: 480 seconds
1300924951 Q * BenG Quit: I Leave
1300925117 J * derjohn_mob aj@88.128.59.34
1300925573 J * Piet ~Piet__@04ZAAALYV.tor-irc.dnsbl.oftc.net
1300937260 Q * bsingh Read error: Connection reset by peer
1300937999 J * bsingh ~balbir@122.166.161.182
1300943432 Q * tolkor Read error: Connection reset by peer
1300943453 J * tolkor ~rj@tdream.lly.earlham.edu
1300943936 J * yarihm ~yarihm@guest-docking-nat-1-004.ethz.ch
1300944273 N * ensc Guest1978
1300944283 J * ensc ~irc-ensc@p5DF2C382.dip.t-dialin.net
1300944320 Q * Radiance Ping timeout: 480 seconds
1300944440 Q * Guest1978 Ping timeout: 480 seconds
1300946485 J * petzsch ~markus@p57B638CB.dip.t-dialin.net
1300946785 J * ghislain ~AQUEOS@adsl2.aqueos.com
1300947915 M * Bertl off to bed now ... have a good one everyone!
1300947920 N * Bertl Bertl_zZ
1300948152 J * harobed ~harobed@arl57-1-82-231-110-14.fbx.proxad.net
1300949921 J * ncopa ~ncopa@3.203.202.84.customer.cdi.no
1300952210 Q * derjohn_mob Ping timeout: 480 seconds
1300953435 Q * harobed Ping timeout: 480 seconds
1300954907 J * puck ~puck@leibniz.catalyst.net.nz
1300955937 Q * Romster Quit: Geeks shall inherit properties and methods of object earth.
1300956249 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net
1300956978 Q * petzsch Quit: Leaving.
1300958106 J * edit ~blagoj@193.104.33.10
1300958214 M * fichte` is it possible to set the time within a vserver?
1300958253 J * kir ~kir@swsoft-msk-nat.sw.ru
1300958357 M * Mr_Smoke Probably, I wouldn't want that though
1300958444 M * edit When I try to start the VPS limited with CPU limit I get this error.
1300958446 M * edit /usr/lib64/util-vserver/vserver.functions: line 1521: /dev/cgroup/testvps/cpu.shares: Permission denied
1300958457 M * edit The VPS is started but no CPU limit available.
1300958542 M * Mr_Smoke edit: is /dev/cgroup/cpu.shares available at all ?
1300958573 N * ensc Guest2000
1300958583 J * ensc ~irc-ensc@p5DF2C382.dip.t-dialin.net
1300958619 M * edit No, but is not that file need to be created automatically?
1300958643 M * Mr_Smoke Well, yes, by your kernel, if your CPU CGroup is enabled and mounted
1300958654 M * Mr_Smoke So either you haven't enabled it, or haven't mounted it, or both
1300958729 M * edit It is enabled and it is mounted, all other files are there, as cpuset.cpus cpuset.mem_exclusive, cpuset.mem_hardwall etc....
1300958740 Q * Guest2000 Ping timeout: 480 seconds
1300959991 M * daniel_hozac that's not the same subsystem
1300959992 M * daniel_hozac cpuset is one
1300959994 M * daniel_hozac cpu is another.
1300960724 M * edit aha, ok, thanks.
1300961099 N * ensc Guest2003
1300961099 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1300961109 J * ensc ~irc-ensc@p5DF2C382.dip.t-dialin.net
1300961270 Q * Guest2003 Ping timeout: 480 seconds
1300961356 M * uranus does anyone has experience with the oom killer in 2.6.38? In older versions we experienced a very high load and oftern has to reset the whole system while one guest is in oom
1300961378 M * Mr_Smoke How older?
1300961388 M * uranus 2.6.35 and 2.6.36
1300961407 M * Mr_Smoke Hm
1300961445 M * Mr_Smoke Had a panic/oops once, with one of these,
1300961451 M * Mr_Smoke can't say about oom
1300961577 M * uranus with kernel 2.6.36.4  i experienced some very confusing loads: suddenly sys usage goes up to 70% and wi/o goes to 30% server is not usable at this point and ssh not showing any response
1300961691 M * uranus vs patchset is vs2.3.0.36.39
1300961704 M * Mr_Smoke Hm
1300961714 M * Mr_Smoke I'm running 2.3.0.37-rc* these dyas
1300961717 M * Mr_Smoke days*
1300961761 M * uranus on one of these hosts i also run 2.6.38 now with rc8 patchset, hopfully this error does not occur any more :)
1300961816 M * Mr_Smoke How long do your sys spike last ?
1300962054 Q * bsingh Read error: Operation timed out
1300962451 M * edit The value here "/dev/cgroup/cpu.shares" will be used as a default value if I did not change it manually, right?
1300962526 M * edit Also, I guess there is no need for "rlimits" now, right?
1300962619 M * Mr_Smoke Right :)
1300962938 M * edit Let's say that I give a VPS to use 14% CPU. This line "echo "0-1" > /dev/cgroup/testvps/cpuset.cpus " will tell the VPS to use 14% of both CPU's at a same time and will use 28% or will use dynamically, for example from the core that is low loaded?
1300962978 M * edit I have 8 CPU's on this machine.
1300963093 M * edit Actually I have 2 on this one where I'm experimenting.
1300963495 Q * mikezzz Ping timeout: 480 seconds
1300963803 J * bsingh ~balbir@122.172.34.205
1300964344 M * daniel_hozac rlimits can still be useful to enforce e.g. the number of tasks.
1300964724 N * ensc Guest2008
1300964729 M * uranus Mr_Smoke, till i reset the server
1300964734 J * ensc ~irc-ensc@p5DF2C382.dip.t-dialin.net
1300964873 M * Mr_Smoke Ouch.
1300964895 Q * Guest2008 Ping timeout: 480 seconds
1300966091 Q * bsingh Read error: Operation timed out
1300966309 J * bsingh ~balbir@122.172.34.205
1300967227 Q * s0undt3ch Remote host closed the connection
1300967327 J * s0undt3ch quasselcor@80.69.34.153
1300967742 N * ensc Guest2016
1300967752 J * ensc ~irc-ensc@p5DF2F280.dip.t-dialin.net
1300967888 J * mib_6wnpt4 c3340e49@ircip1.mibbit.com
1300967923 J * zucchini c3340e49@ircip3.mibbit.com
1300967943 Q * zucchini 
1300968150 Q * Guest2016 Ping timeout: 480 seconds
1300969328 J * ghislain1 ~AQUEOS@adsl2.aqueos.com
1300969379 J * petzsch ~markus@p57B638CB.dip.t-dialin.net
1300969618 Q * ghislain Ping timeout: 480 seconds
1300970029 Q * FireEgl Ping timeout: 480 seconds
1300970503 Q * petzsch Quit: Leaving.
1300971328 J * petzsch ~markus@p57B638CB.dip.t-dialin.net
1300974700 Q * bsingh Read error: Connection reset by peer
1300975753 J * petzsch1 ~markus@p57B6517D.dip.t-dialin.net
1300975881 M * uranus how do i tell a guest that he has to use a specific Parition as root? if using normal guest fstab and add a / line it says vsysctl: open("."): Permission denied
1300975946 M * daniel_hozac did you set a barrier on it?
1300975981 M * uranus just a  moment
1300976034 M * uranus ----buicX on /var/lib/vservers/guest
1300976070 Q * petzsch Ping timeout: 480 seconds
1300976181 M * daniel_hozac what does dmesg on the host say?
1300976216 M * uranus vxW: [vsysctl,539:#6005|6005|6005] did hit the barrier.
1300976229 M * uranus do i have to remove the barrier on the vdir?
1300976240 M * daniel_hozac what's the "vdir"?
1300976246 M * uranus /var/lib/vservers/guest
1300976253 M * uranus softlink to that dir
1300976253 M * daniel_hozac i thought it didn't have one.
1300976264 M * daniel_hozac where do you have barriers, exactly?
1300976272 M * uranus /var/lib/vservers 
1300976472 M * uranus i'm using the deb packages of your util-vserver pre's 
1300976482 M * uranus currently 0.30.216-pre2952
1300976606 Q * harobed Quit: Ex-Chat
1300977070 Q * petzsch1 Quit: Leaving.
1300978266 J * dowdle ~dowdle@scott.coe.montana.edu
1300978325 J * bsingh ~balbir@122.172.5.88
1300979207 N * Bertl_zZ Bertl
1300979211 M * Bertl morning folks!
1300979384 M * edit Yo!
1300979541 N * ensc Guest2040
1300979551 J * ensc ~irc-ensc@p5DF2F280.dip.t-dialin.net
1300979713 Q * Guest2040 Ping timeout: 480 seconds
1300980170 J * yarihm_ ~yarihm@guest-docking-nat-2-019.ethz.ch
1300980419 Q * yarihm Ping timeout: 480 seconds
1300980611 M * edit If I do not have this file which CPU/core will be used by the VPS's by default?
1300980612 M * edit /etc/vservers/<guestname>/cgroup/cpuset.cpus 
1300980718 M * daniel_hozac all of them
1300980889 M * edit So, if one VPS can use 10% and I have 2 CPU's he will use 1.) 5% from the first one or 5% from the second one or 2.) 10% from the first one to complete some operation and once the operation is completed and the second CORE is free than he will use the second one for the next operation (assuming that the first CORE is already taken by another process from another VPS etc...), something like that, right?
1300980906 M * daniel_hozac how did you say it can use 10%?
1300980945 M * edit in /dev/cgroup/<guest name>/cpu.shares
1300980979 M * edit or /etc/vservers/testvps/cgroup/cpu.shares
1300980990 M * edit I'm reading this tutorial.
1300980992 M * edit http://linux-vserver.org/util-vserver:Cgroups#Real_world_Examples_of_Scheduling
1300981110 M * daniel_hozac okay?
1300981143 M * edit Okay what?
1300981145 M * edit :)
1300981346 M * daniel_hozac so what is your actual configuration?
1300981454 M * edit So, I have testvps
1300981476 M * edit in /etc/vservers/testvps/cgroup
1300981477 M * edit I have
1300981493 M * edit cpuset.cpus
1300981502 M * edit when I put 0 it use only the first CPU
1300981511 M * edit when I put 1 it use the seccond CPU
1300981522 M * edit Now I want to know if I remove that file.
1300981544 M * edit It will use all the CPU's (I have 2 total)
1300981558 M * daniel_hozac yes
1300981610 M * ghislain1 not in real time of course as the cgroup are built at start time but yes
1300981644 M * ghislain1 to change in real time just directly use the cgroup's files
1300981718 M * edit But how,  I will use 10% of the CPU's as total 10%, or 10% from the first core and 10% from the second CORE, which in that case will be 20%.
1300981980 J * edit2 ~blagoj@89.205.105.228
1300982005 P * edit 
1300982080 M * daniel_hozac it completely depends on what kernel you're on, and how you told it to do 10%.
1300982090 M * daniel_hozac and the load on the system.
1300982390 M * edit2 I told you, I define in the "/etc/vservers/testvps/cgroup/cpu.shares" to use 10% from the total CPU.
1300982402 M * edit2 Load is low because I have 2-3 test vPS's on the machine.
1300982420 M * edit2 kernel is  2.6.36.2-vs2.3.0.36.38.2
1300982433 M * daniel_hozac what does "10%" mean? that file doesn't accept that as input.
1300983422 M * uranus what must i do when i receive the following message after start: vsysctl: open("."): Permission denied - dmesg: vxW: [vsysctl,539:#6005|6005|6005] did hit the barrier. 
1300983474 M * daniel_hozac figure out where your barriers are set.
1300983513 M * uranus ok i found a barrier on /var/lib/vservers (vdir is below that dir)
1300983574 M * uranus must i remove these?
1300983577 M * daniel_hozac no
1300983584 M * daniel_hozac the directory above is where you want it.
1300983595 M * daniel_hozac typically, anyway.
1300983604 M * daniel_hozac you can try removing it to see if that helps.
1300983658 M * uranus okay i found the problem
1300983677 M * uranus there was a barrier on the partition i tried to mount during vserver start
1300983688 M * daniel_hozac right....
1300983707 M * uranus thanke daniel_hozac 
1300986256 Q * yarihm_ Quit: This computer has gone to sleep
1300986506 J * bonbons ~bonbons@2001:960:7ab:0:45de:46e0:1c34:62c
1300988350 J * BenG ~bengreen@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com
1300988359 M * edit2 Daniel, I do not have "10%" as a value in /etc/vservers/testvps/cgroup/cpu.shares I have 800 as a value in that file, which is 10% of the CPU, so the total CPU is 8000.
1300989035 J * petzsch ~markus@p57B6517D.dip.t-dialin.net
1300989760 M * ghislain1 this 10% is only if 100% of the cpu is taken if i am not misslead
1300989837 M * ghislain1 i think it trigger only if there is cpu starvation but i am not sure. Saying it otherwise it is a balance when the cpu is starved but if not a single group can rise way over this
1300989866 M * ghislain1 if the other are not using too much cpu, so this is not HARD limit as i understand it
1300990065 M * Bertl yep, hard limit is a separate project mainline is working on
1300990079 M * Bertl I think I even have a patchset which adds this somewhere
1300990605 Q * petzsch Quit: Leaving.
1300990987 Q * mib_6wnpt4 Remote host closed the connection
1300991319 Q * BenG Quit: I Leave
1300993023 Q * ghislain1 Quit: Leaving.
1300993890 Q * bonbons Ping timeout: 480 seconds
1300994127 J * yarihm ~yarihm@guest-docking-nat-1-053.ethz.ch
1300994327 Q * yarihm 
1300994502 J * bonbons ~bonbons@2001:960:7ab:0:45de:46e0:1c34:62c
1300995066 J * FireEgl ~FireEgl@173-25-19-139.client.mchsi.com
1300995612 Q * Piet Remote host closed the connection
1300995955 M * edit2 Ok, more/less, let's say, I have 2 CPU's with 2 GHz on the machine so total 4 GHz, can I limit 1 VPS to use, 1 GHz and other VPS's to use the rest without interfere with eachother?
1300996154 M * daniel_hozac depends on your kernel.
1300996171 M * daniel_hozac i don't think 2.6.36 had the hard limits.
1300996562 J * manana ~mayday090@84.17.25.149
1300996647 M * edit2 Which kernel has it?
1300996936 J * Piet ~Piet__@04ZAAAMSB.tor-irc.dnsbl.oftc.net
1300997162 J * derjohn_mob aj@88.128.7.135
1300998177 J * imcsk8 ~ichavero@148.229.1.11
1300998820 M * Bertl we removed the hard limits from earlier kernels because they were broken
1300998837 Q * Piet Ping timeout: 480 seconds
1300998875 M * daniel_hozac right
1300999194 M * edit2 Ok, than please tell me which kernel suppors it?
1300999294 M * daniel_hozac old ones...
1300999325 M * edit2 aaahhh, so the new kernels do not have hard limit.
1300999328 M * edit2 Ok, good to know.
1300999330 M * edit2 Thanks!
1300999416 J * Piet ~Piet__@04ZAAAMTK.tor-irc.dnsbl.oftc.net
1300999703 Q * manana Remote host closed the connection
1300999964 M * Bertl unless you apply the patches, of course
1301000132 Q * FloodServ Service unloaded
1301000305 M * edit2 but the patches can broke the kernel, right?
1301000323 M * edit2 Anyhow just for testing can you please tell me where can I download the patches?
1301000335 M * Bertl the older ones did, there are newer ones posted to lkml
1301000433 J * FloodServ services@services.oftc.net
1301000607 J * yarihm ~yarihm@178-83-213-36.dynamic.hispeed.ch
1301000726 M * Bertl https://lkml.org/lkml/2011/2/15/963 for example
1301000807 J * ghislain ~AQUEOS@adsl2.aqueos.com
1301000939 N * ensc Guest7
1301000949 J * ensc ~irc-ensc@p5DF2F280.dip.t-dialin.net
1301001110 Q * Guest7 Ping timeout: 480 seconds
1301001169 M * Bertl edit2: if you test them, let us know how they do
1301001315 M * edit2 Sure, no rpboem. But I'm unable to see where can I download them? Or maybem I'm blind?
1301001316 M * edit2 :)
1301001355 M * Bertl it's the start of the thread, the 1/7, 2/7 ... 7/7 contain the patches
1301001429 M * Bertl e.g. https://lkml.org/lkml/2011/2/15/965
1301001436 M * Bertl scroll down to 'Get diff 1'
1301001449 M * Bertl https://lkml.org/lkml/diff/2011/2/15/965/1
1301001463 M * Bertl which is the first of seven patches
1301001514 M * edit2 Ok Ok, I found them. 
1301001534 M * edit2 I need to apply all 7 patches or random, one, than pay with it, than another one etc....?
1301001551 M * Bertl all seven
1301001570 M * edit2 okay
1301001613 M * edit2 I will test them tomorrow, it is late now, here where I live , and will let you know the results.
1301001613 Q * bonbons Quit: Leaving
1301001627 M * Bertl np, no need to hurry
1301001663 Q * ghislain Quit: Leaving.
1301001747 M * edit2 One more question, which kernel is recommended, I have at this moment 2.6.36 but it is not a problem to upgrade it?
1301001762 M * Bertl no, I'd go for 2.6.38 atm
1301001768 M * edit2 ok, will do.
1301001782 M * edit2 Going to sleep now and ready for testing tomorrow.
1301001790 M * edit2 Bye all.
1301001798 M * Bertl good night, cya
1301002437 Q * yarihm Ping timeout: 480 seconds
1301002508 J * yarihm ~yarihm@vpn-global-243-dhcp.ethz.ch
1301002797 Q * uranus Quit: Verlassend
1301004617 J * ser ~ser@cpc3-nott15-2-0-cust186.12-2.cable.virginmedia.com
1301004794 Q * ser 
1301006089 Q * yarihm Quit: This computer has gone to sleep
1301006112 J * yarihm ~yarihm@178-83-213-36.dynamic.hispeed.ch
1301007401 Q * yarihm Quit: This computer has gone to sleep
1301009789 J * veki ~chatzilla@109-93-126-186.dynamic.isp.telekom.rs
1301009888 P * veki 
1301010932 Q * dowdle Remote host closed the connection