1352678845 Q * Ghislain Quit: Leaving.
1352689565 M * Bertl off to bed now ... have a good one everyone!
1352689569 N * Bertl Bertl_zZ
1352703045 J * Ghislain ~aqueos@adsl1.aqueos.com
1352704478 Q * nkukard Remote host closed the connection
1352704567 J * nkukard ~nkukard@41-133-139-203.dsl.mweb.co.za
1352707424 J * kir ~kir@swsoft-msk-nat.sw.ru
1352707521 Q * fisted Remote host closed the connection
1352709494 P * kir PING 1352709494
1352713230 Q * FireEgl Quit: Leaving...
1352713846 Q * geos_one Quit: ChatZilla 0.9.89 [Firefox 16.0.2/20121029000647]
1352714890 J * BenG ~bengreen@cpc29-aztw23-2-0-cust144.18-1.cable.virginmedia.com
1352714979 J * fisted ~fisted@gw-2s00.inf.fh-bonn-rhein-sieg.de
1352715700 Q * fisted Ping timeout: 480 seconds
1352716010 J * jokar ~mohsen@188.75.84.162
1352717901 J * FireEgl ~FireEgl@173-25-83-57.client.mchsi.com
1352719564 Q * Guy- Remote host closed the connection
1352720132 J * clopez ~clopez@242.17.60.213.dynamic.mundo-r.com
1352720451 Q * ircuser-1 Ping timeout: 480 seconds
1352720548 Q * BenG Quit: I Leave
1352720596 J * BenG ~bengreen@cpc29-aztw23-2-0-cust144.18-1.cable.virginmedia.com
1352720731 M * nkukard mb
1352723012 Q * Aiken Remote host closed the connection
1352723903 J * ircuser-1 ~ircuser-1@35.222-62-69.ftth.swbr.surewest.net
1352723909 J * Guy- ~korn@elan.rulez.org
1352725320 Q * ncopa Quit: Leaving
1352725412 J * ncopa ~ncopa@3.203.202.84.customer.cdi.no
1352727133 J * nezafat ~nezafat@04ZAABGY9.tor-irc.dnsbl.oftc.net
1352727824 Q * BenG Quit: I Leave
1352728021 Q * nezafat autokilled: . If you feel an error has been made, please contact support@oftc.net. (2012-11-12 13:47:01)
1352728066 J * nezafat ~nezafat@659AABZXD.tor-irc.dnsbl.oftc.net
1352728066 Q * nezafat 
1352728669 Q * clopez Ping timeout: 480 seconds
1352729198 J * clopez ~clopez@242.17.60.213.dynamic.mundo-r.com
1352729224 N * Bertl_zZ Bertl
1352729230 M * Bertl morning folks!
1352731310 J * shaggy63 ~packethel@c-98-212-73-99.hsd1.in.comcast.net
1352731398 M * shaggy63 vserver is pretty damn amazing with hashify.  I was able to get 725 virtuals using 451 mb each down to 19 GB total with apache and mysql with 0.00 load
1352731459 M * shaggy63 Are there any downsides to using hashify?
1352731486 J * geos_one ~chatzilla@vpn2-29.tu-graz.ac.at
1352731523 J * fisted ~fisted@p5DE8DAFC.dip.t-dialin.net
1352731575 M * Bertl if done properly, not really
1352731612 M * shaggy63 It's a beautiful thing
1352731622 M * Bertl hashify basically uses 'protected' hard links
1352731660 M * Bertl so you get all the advantages of hard links without the security issues 
1352731746 M * shaggy63 I'm assuming the COW has to be a patch to the kernel though.
1352731759 M * Bertl you probably could get similar savings disk wise with overlay filesystems, but you will lose all the memory and performance gains which result from the single inode setup
1352731784 M * Bertl of course, kernel needs to be patched to do the CoW link breaking
1352731806 M * daniel_hozac just like it needs to be patched to do Linux-VServer ;)
1352731836 M * shaggy63 of course.
1352731848 M * shaggy63 Its a great feature, wonder why it isn't mainline.
1352731877 M * shaggy63 How's the stability compared to openvz?
1352731939 M * Ghislain i wonder how daniel or hebert could say something about comparing stability of another project ^^
1352731966 M * Bertl IMHO it is sufficient to compare the patches to get an idea :)
1352731984 M * Ghislain in  vserver you are not the testbed of parrallels company but otherwise i cannot tell i have enough problem with their plesk panel to risk myself with their product ^^
1352731988 M * shaggy63 I understand, I'm just wondering if anyone has experience with both.  Openvz seems to crash often from what I read on forums
1352732038 M * Ghislain i quite never had a crash on vserver if i remove from my count the buggy mainline 3.2.x kernels that were a nightmare until .18
1352732076 M * Bertl I have a Linux-VServer guest with an uptime of 1104 days, and it is running on a host with moderate load
1352732077 M * Ghislain and i used vserver on some very highly loaded systems
1352732091 M * shaggy63 Nice
1352732118 M * shaggy63 I like stability not new fancy shit that breaks because its not properly tested.
1352732168 M * Ghislain the difference is that in vserver not all is done for you it demand a minimum involvement. So you need to build your kernel and utils or trust and use the one provided by contributor. In my case this is very solid piece of software and works like a charm since 2005 that i use it
1352732217 M * Bertl we are not a company advertising for a commercial product either :)
1352732237 M * shaggy63 I passed up vserver a while back and went openvz because everything I read stated vserver was difficult.  I set it up over the weekend, it's actually quite easy to admin.
1352732289 J * fisted_ ~fisted@p5DE8DAFC.dip.t-dialin.net
1352732295 M * shaggy63 I like it a lot more than openvz
1352732300 Q * fisted Remote host closed the connection
1352732305 N * fisted_ fisted
1352732338 M * Bertl glad to hear!
1352732514 M * shaggy63 When I tested openvz I started about 400 guests and it brought the server to it's knees.  With hashify on vserver I started almost double that with 0 load.  Its impressive.  I think the reason everyone is going openvz is because of suspend/resume but come on it doesn't take long to stop, rsync and start a vserver guest.
1352732544 Q * geos_one Quit: ChatZilla 0.9.89 [Firefox 16.0.2/20121029000647]
1352732614 J * thierryp ~thierry@zankai.inria.fr
1352732782 M * Bertl do/did you actually use the suspend/resume?
1352732837 M * Bertl also, did you try to suspend, let's say 100 guests at once?
1352732837 M * shaggy63 only in testing
1352732845 M * shaggy63 no
1352732878 M * shaggy63 I figured it would be nice to have if a guest was using a lot of resources I could move it
1352732939 M * Bertl btw, mainline already supports the freezer subsystem
1352732970 M * shaggy63 ? explain please.
1352732981 M * Bertl i.e. suspend/resume for an arbitrary group of processes
1352733023 M * Bertl http://www.kernel.org/doc/Documentation/cgroups/freezer-subsystem.txt
1352733040 M * Bertl so that is something you can use in Linux-VServer as well
1352733131 M * shaggy63 Sweet, just when I thought I was done testing I get something else to test !
1352733252 M * shaggy63 Why does 2.3.4.3.1 not have cow?  Its not going away is it?
1352733690 M * Bertl no, it just isn't ported to and tested on 3.6 yet
1352733774 M * shaggy63 ok, damn they are up to 3.6 already. what the heck is going on in kernel dev land.
1352734698 M * shaggy63 Well, thanks for the chat and the awesome software, mind if I idle?
1352734857 M * Bertl not at all! you're welcome!
1352735077 M * Ghislain but is there a sytem to copy the memory and context of process that you freeze with the cgroup ?
1352735134 M * Ghislain if there is i really have to try that ^^, not sure i will use it but for fun of it ^^
1352736042 M * Bertl in theory, that can be done from userspace, but the only system I know working and tested is the suspend to disk from mainline kernels
1352736189 J * fisted_ ~fisted@p5DE8FFD1.dip.t-dialin.net
1352736509 Q * fisted Ping timeout: 480 seconds
1352736511 N * fisted_ fisted
1352737835 M * Ghislain ok ^^
1352737847 J * fisted_ ~fisted@p5DE8FFD1.dip.t-dialin.net
1352738095 Q * fisted Ping timeout: 480 seconds
1352738095 N * fisted_ fisted
1352738717 Q * fisted Ping timeout: 480 seconds
1352739196 Q * thierryp Remote host closed the connection
1352740619 J * bonbons ~bonbons@2001:960:7ab:0:a810:1531:16cb:c729
1352740972 Q * Chlorek Quit: bumbum
1352742142 Q * ravid Remote host closed the connection
1352742514 J * arekm ~arekm@ixion.pld-linux.org
1352742536 J * Chlorek chlorek@chlorek.com
1352742685 Q * ncopa Ping timeout: 480 seconds
1352742972 J * fisted ~fisted@xdsl-78-35-92-4.netcologne.de
1352743929 J * ncopa ~ncopa@3.203.202.84.customer.cdi.no
1352744084 J * geos_one ~chatzilla@80.123.185.198
1352745291 Q * ensc|w Remote host closed the connection
1352745296 J * ravid ~quassel@196-210-239-253.dynamic.isadsl.co.za
1352745300 J * ensc|w ~ensc@www.sigma-chemnitz.de
1352746270 Q * ncopa Quit: Leaving
1352750351 J * thierryp ~thierry@home.parmentelat.net
1352751141 Q * arekm Quit: leaving
1352751490 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f
1352753362 M * Bertl off for a nap ... bbl
1352753366 N * Bertl Bertl_zZ
1352755070 J * dkg ~dkg@lair.fifthhorseman.net
1352755139 M * dkg i quite regularly get messages about some processes failing to terminate when stopping a vserver guest.  is this a known or common situation?
1352755166 M * dkg i'm using the debian vserver packages, with the squeeze vserver kernel, fwiw.
1352756102 M * shaggy63 go to linux-vserver.org
1352756122 M * shaggy63 err, go to www.linux-vserver.org/Upstart_issues
1352756127 M * shaggy63 search for timeout
1352756135 M * shaggy63 about halfway down the page
1352756205 M * Chlorek it's shit
1352756238 M * Chlorek don't tell people what they should do if you are the new one
1352756295 M * shaggy63 I'm just pointing him to information that might be useful
1352756325 M * Chlorek squeeze doesn't use it
1352756387 M * shaggy63 interesting, then enlighten me because I have the same issue.
1352756400 M * Chlorek and I don not.
1352756410 M * Chlorek event if I use
1352756445 M * Chlorek *do
1352756553 M * Chlorek the point is upstart doesn't work in vserver, and it's the end of story
1352756645 M * shaggy63 ok, then what is causing the timeout when shutting down vserver guests
1352756721 Q * bonbons Quit: Leaving
1352756815 J * fisted_ ~fisted@xdsl-84-44-220-122.netcologne.de
1352757143 M * dkg i don't use upstart
1352757220 Q * fisted Ping timeout: 480 seconds
1352757220 N * fisted_ fisted
1352758300 Q * thierryp Remote host closed the connection
1352758796 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:a4be:d1b4:1fc0:9422
1352760573 N * Bertl_zZ Bertl
1352760577 M * Bertl back now ...
1352760695 M * Bertl dkg: in almost all cases, this is a misconfigured guest
1352760738 M * Bertl i.e. some service or the guest's init (doesn't matter what kind of init) doesn't shut down properly when signalled to do so
1352760741 M * dkg Bertl: misconfiguration within the guest itself?  or within the host's /etc/vservers ?
1352760747 M * dkg ah, ok.
1352760781 M * Bertl util-vserver basically sends a signal (if there is an init) or runs the sysv sequence to shut down the services
1352760786 M * dkg these are using the "plain" init style, fwiw.
1352760809 M * Bertl so you have a guest init, which is simply signalled
1352760838 M * Bertl after that, util-vserver waits for a certain amount of time, and then forces whatever is still running to shut down via direct signalling
1352760862 M * Bertl and reports the processes still found running to you
1352760880 M * Bertl (usually that list gives a good hint what was running)
1352760901 M * dkg yep, it's quite explicit :)
1352760940 M * Bertl if it basically lists all processes usually running in a guest, then the guest's init is misconfigured to handle the shutdown
1352760981 M * dkg Bertl: are you ok if i cut and paste this conversation (the last ~5 minutes) into some public documenation with a group i'm working with?
1352760987 M * Bertl if it only lists a specific service, but other services did shut down properly, then either that service really needs longer than the configured (guest config) timeout or the service doesn't shut down properly
1352761007 M * dkg thanks, that makes sense to me.
1352761012 M * Bertl no proble
1352761014 M * Bertl +m
1352761201 Q * fisted Remote host closed the connection
1352761252 J * fisted ~fisted@xdsl-87-78-80-138.netcologne.de
1352762509 Q * thierryp Remote host closed the connection
1352764599 Q * Ghislain Quit: Leaving.