1316046633 J * BenG_ ~bengreen@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com
1316046637 Q * BenG Quit: I Leave
1316046675 Q * BenG_ 
1316046897 Q * clopez Ping timeout: 480 seconds
1316047040 M * Marillion daniel_hozac: your patch set work fine
1316047046 M * Marillion daniel_hozac: Linux marillion 2.6.32-rhel6-vserver #1 SMP Thu Sep 15 00:30:23 CEST 2011 i686 GNU/Linux
1316047077 M * Marillion daniel_hozac: it was a long try :)
1316051536 Q * FireEgl Quit: Leaving...
1316051970 Q * micah Remote host closed the connection
1316051988 J * FireEgl FireEgl@2001:470:e056:1:dcfe:884:fda:cfc
1316052437 J * micah ~micah@micah.riseup.net
1316054786 Q * Aiken Quit: Leaving
1316059048 Q * hparker Remote host closed the connection
1316060397 J * sannes ~ace@cm-84.209.106.118.getinternet.no
1316060721 J * hparker ~hparker@2001:470:1f0f:32c:beae:c5ff:fe01:b647
1316064887 M * Bertl off to bed now .. have a good one everyone!
1316064892 N * Bertl Bertl_zZ
1316065010 J * ghislain ~AQUEOS@adsl2.aqueos.com
1316067270 J * ncopa ~ncopa@3.203.202.84.customer.cdi.no
1316067607 Q * derjohn_mob Ping timeout: 480 seconds
1316069563 J * derjohn_mob ~aj@213.238.45.2
1316076344 J * clopez ~clopez@155.99.117.91.static.mundo-r.com
1316076930 J * kir1 ~kir@swsoft-msk-nat.sw.ru
1316076934 P * kir1 
1316079397 Q * hparker Quit: Quit
1316084127 Q * LuckyLuke Ping timeout: 480 seconds
1316085582 P * kir Leaving.
1316085842 J * BenG ~bengreen@cpc12-aztw24-2-0-cust146.aztw.cable.virginmedia.com
1316086022 M * BenG hey all
1316086036 M * BenG I'm putting together a set of scripts
1316086054 M * BenG I need somethings that gives me a list of all the directories where running vservers are living
1316086089 M * BenG I'm bodging something now, but does anyone have something elegant that does this already?
1316086282 M * BenG and can anyone give example usage of vserver-info?
1316086442 M * BenG oh, that doesn't seem that useful
1316086492 M * ccxCZ maybe you want vserver-stat?
1316086526 M * ccxCZ the name row probably corresponds to /etc/vservers
1316086581 M * ccxCZ the actual directory would be obtained by following the symlink /etc/vservers/${NAME}/vdir/
1316086856 N * Bertl_zZ Bertl
1316086860 M * Bertl morning folks!
1316086890 M * Bertl BenG: try vserver-info - -l (for tags) and in your case vserver-info <name> RUNNING
1316086903 M * Bertl return value and output might be used
1316086922 M * BenG marvellous, cheers
1316086949 M * Bertl similar with vserver-info <name> VDIR
1316086958 M * Bertl (gives you the data directory)
1316087456 Q * nkukard Quit: Leaving
1316087566 M * ccxCZ morning Bertl
1316087568 M * ccxCZ I have to admit I was bit shocked that there is no version control system in place here to support development, so I decided to give bzr-loom a try so the grsec patches can be merged more intelligently.
1316087684 M * ccxCZ http://wiki.bazaar.canonical.com/Documentation/LoomAsSmarterQuilt hopefully it will help a bit
1316088251 M * Bertl if that works for you, perfect
1316088527 M * ccxCZ It should do away with resolving the same conflicts each release
1316088768 M * ccxCZ I think it would be neat to configure buildbot so it fetches various patches (vserver, pax, grsec) and either produces combined patchset or displays what conflicts, that way everyone would have quick overview of project status and how much effort is necessary to integrate it
1316089004 J * LuckyLuke ~luca@host65-83-static.228-95-b.business.telecomitalia.it
1316089537 J * fisted_ ~fisted@xdsl-87-78-216-3.netcologne.de
1316089590 Q * fisted Ping timeout: 480 seconds
1316089717 M * Bertl go ahead .. give it a try
1316089827 M * ccxCZ Bertl: out of curiosity, what tools do you use? I see there is no official issue tracker or VCS, I don't believe you keep all the stuff in your head ;-)
1316089885 M * Bertl in the beginning, we tried out a lot, there even was a bug tracker and various revision control systems
1316089918 M * Bertl nowadays, we basically keep everything in the IRC logs and on the ML
1316089995 M * Bertl there are only a few issues with Linux-VServer, so it's rather easy to either fix them right when they come up or keep them listed on a single page (for a few days, till they are fixed)
1316090091 M * Bertl for the VCS part, it turned out that for the kernel patches, something similar to git (hard linked kernel source trees) and patches are the simplest way to work
1316090144 M * ccxCZ currently branching/cloning linux source tree, this surely takes a while
1316090195 M * Bertl well, a few seconds, very similar to what git does
1316090241 M * Bertl the most time consuming part are test builds and actual testing via kvm
1316090263 M * ccxCZ I'm asking because I'm currently investigating toolset that would integrate well with ML/IRC instead of trying to replace them
1316090334 M * Bertl well, we definitely do not want to replace either the mailing list or irc, as they are the default communication channels
1316090375 M * ccxCZ I'm writing up on it, but the document needs bit of reworking as most of it was written around 3am :]
1316090427 M * Bertl and it is rather unlikely that daniel_hozac or I will change the way we maintain and develop Linux-VServer without a good reason (which would boil down to some kind of speedup or work load reduction)
1316090487 M * Bertl but that doesn't mean that somebody (like you) couldn't or shouldn't use different tools or mechanisms
1316090531 M * ccxCZ it would be kind of silly, switching without any kind of benefit
1316090541 M * Bertl and as you probably know, it's not very hard to check in a new patch published for Linux-VServer into whatever VCS you prefer
1316090557 M * Bertl (same goes for pax/grsec I guess)
1316090590 M * ccxCZ now I'm mostly interested in what people do, what they need and what they use
1316090663 M * Bertl sounds good to me ...
1316090768 M * ccxCZ one of my main goals is real decentralization of everyting project related, kind of like fossil does it, except not tying people to one do-it-all app
1316091960 M * Bertl please elaborate ...
1316092561 M * ccxCZ I'll ping you when I clean up my document on that, okay?
1316092693 M * Bertl sure, feel free to post it on the ML
1316092987 M * ccxCZ okay
1316094803 Q * FireEgl Remote host closed the connection
1316095266 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net
1316095268 Q * LuckyLuke Quit: leaving
1316095591 J * LuckyLuke ~luca@host65-83-static.228-95-b.business.telecomitalia.it
1316095654 J * kir ~kir@swsoft-msk-nat.sw.ru
1316095654 J * kir1 ~kir@swsoft-msk-nat.sw.ru
1316095662 Q * kir 
1316095804 J * FireEgl ~FireEgl@173-16-9-169.client.mchsi.com
1316097867 J * dowdle ~dowdle@scott.coe.montana.edu
1316099501 Q * ncopa Quit: Leaving
1316099817 J * PowerKe ~tom@94-226-105-186.access.telenet.be
1316101570 Q * harobed Quit: Ex-Chat
1316102930 J * bonbons ~bonbons@2001:960:7ab:0:cc86:f0c0:d9cd:3309
1316104966 Q * kir1 Quit: Leaving.
1316105920 Q * derjohn_mob Ping timeout: 480 seconds
1316106538 Q * BenG Quit: I Leave
1316108057 J * nkukard ~nkukard@41-133-198-167.dsl.mweb.co.za
1316108561 Q * ghislain Quit: Leaving.
1316110372 J * derjohn_mob ~aj@88.128.16.156
1316110749 J * thierryp ~thierry@home.parmentelat.net
1316111346 Q * thierryp Remote host closed the connection
1316112195 M * Guy- Bertl: any news about the supplementary group membership vs. fuse issue, by any chance?
1316112253 M * Bertl not on my side ... haven't had the time to look into it
1316113237 M * Guy- OK, thanks
1316113781 M * Bertl do we have a simple test case for that btw?
1316114511 J * hparker ~hparker@2001:470:1f0f:32c:beae:c5ff:fe01:b647
1316114775 J * hijacker_ ~hijacker@cable-84-43-136-96.mnet.bg
1316115437 J * sweil ~stefan@p5086F619.dip.t-dialin.net
1316116347 Q * sweil Ping timeout: 480 seconds
1316116587 M * Guy- Bertl: depends on what you call "simple"
1316116652 M * Guy- Bertl: the test case is this: have a zfs bind mounted in a guest; on the zfs, chown user1:group file; chmod 664 file; then inside the guest, try to write to the file as user2, who is a member of group but whose primary GID isn't group
1316116674 M * Guy- Bertl: this currently results in 'permission denied' inside a guest but works on the host
1316116732 M * Bertl yeah, I remember, but I was thinking about something without zfs maybe?
1316116767 M * Guy- I don't use any other fuse fs
1316116772 M * Guy- I don't know whether it can be reproduced with those
1316116779 M * Bertl because AFAICT, it could as well be a zfs issue, no?
1316116803 M * Bertl (not saying it is)
1316116808 M * Guy- I don't know what interfaces with what at what level and how, so I can't say
1316116847 M * Guy- I mean, I don't know how the auxiliary group memberships are passed through fuse
1316116910 M * Guy- but to me, the likeliest scenario is that some guest isolation mechanism is preventing the group memberships to be seen by the fuse subsystem
1316116947 M * Guy- of course, having no idea how it actually works, I could be very wrong :)
1316116965 M * Guy- can you suggest another fuse filesystem to try it with?
1316116998 M * Bertl I'd say any supporting/checking auxiliary groups should do
1316117037 M * Bertl probably there are some quite simple filesystems, which just pass through data (as fuse examples)
1316117057 M * Bertl something like that would be perfect to toy with
1316117501 M * Guy- I'll try it with rofs
1316117726 J * ghislain ~AQUEOS@adsl2.aqueos.com
1316118342 M * Guy- Bertl: I couldn't reproduce it with rofs
1316118610 Q * sannes Remote host closed the connection
1316118624 M * Bertl but secondary groups are checked and work with that, yes?
1316118709 M * Guy- apparently yes
1316118719 Q * fisted_ Quit: leaving
1316118740 M * Guy- I'll also try with copyfs
1316118747 J * fisted ~fisted@xdsl-87-78-216-3.netcologne.de
1316118747 M * Bertl okay, thanks!
1316119069 Q * hijacker_ Quit: Leaving
1316119143 Q * bonbons Quit: Leaving
1316119420 Q * clopez Ping timeout: 480 seconds
1316119896 Q * arekm Ping timeout: 480 seconds
1316120700 J * cuba33ci_ ~cuba33ci@111-240-178-227.dynamic.hinet.net
1316121049 Q * cuba33ci Ping timeout: 480 seconds
1316121057 N * cuba33ci_ cuba33ci
1316122049 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f
1316122167 Q * derjohn_mob Ping timeout: 480 seconds
1316123041 Q * ghislain Quit: Leaving.
1316124112 M * Guy- well, copyfs doesn't seem to work at all
1316124121 M * Guy- (through no fault of vserver)
1316124346 J * thierryp ~thierry@home.parmentelat.net
1316124819 Q * thierryp Remote host closed the connection
1316125582 J * thierryp ~thierry@home.parmentelat.net
1316125727 Q * thierryp Remote host closed the connection
1316125972 M * Guy- Bertl: the version of zfs-fuse I use now logs:
1316125975 M * Guy- zfs-fuse: fuse_req_getgroups returned an error, make sure /proc is available
1316126007 M * Guy- this looks very relevant
1316126028 M * Guy- The current fuse kernel module in linux (as of 2.6.30) doesn't pass the group list to userspace, hence this function needs to parse "/proc/$TID/task/$TID/status" to get the group IDs.
1316126059 M * Guy- this is probably the cause of the problem: zfs-fuse runs on the host and doesn't have access to the /proc file it would need
1316126088 M * Bertl so, try running zfs-fuse in the spectator context then?
1316126101 M * Bertl btw, I thought we had tried running it in the guest context?
1316126107 M * Guy- that's just what I was going to try
1316126115 M * Bertl (at least I remember havin suggested that)
1316126189 M * Guy- yes, I checked my logs and apparently we found that the same problem arose then as well
1316126203 M * Guy- but maybe I was doing it wrong
1316126301 M * Bertl well, let's see how xid=1 works out
1316126518 M * Bertl but that pretty much explains why I couldn't find anything in the kernel code
1316126847 M * Guy- Bertl: OK, starting it from context 1 is a viable workaround
1316126867 M * Bertl well, it should work as well from inside the guest context
1316126875 M * Bertl i.e. the one using the zfs mount
1316126903 M * Bertl but it cannot work from the host or a different guest context
1316126922 M * Guy- if I start zfs-fuse in context 1, then it works in all guetss
1316126926 M * Guy- *guests
1316126930 M * Bertl correct
1316126962 M * Guy- since these filesystems are shared among several guests, that's the only approach I can take
1316127005 M * WMP Bertl: hello, can you tell me how to add second ip? i make interfaces/1 and files in this
1316127027 M * WMP but in ifconfig i have 2 network cards on this same ip
1316127027 M * Guy- WMP: that'll take effect the next time you start the guest
1316127035 M * WMP i now this
1316127038 M * WMP know*
1316127043 M * Bertl I asked you (yesterday?) to upload the file contents to a pastebin
1316127059 M * Bertl (the contents of all files in interfaces)
1316127097 M * Guy- easy way to obtain it: rgrep . /etc/vservers/guest/interfaces
1316127105 M * WMP Bertl: sorry, i havent this messges
1316127116 M * Bertl np, please do it now :)
1316127223 M * WMP http://pastebin.com/3cKaBsUd
1316127270 M * Bertl so that's only one interface, no?
1316127300 M * WMP yes
1316127315 M * Bertl and with this config, you see two interfaces inside the guest?
1316127356 M * WMP http://pastebin.com/FzbH4Qw3
1316127361 M * WMP yes
1316127406 M * Bertl okay, so what does 'ip a l' inside the guest show?
1316127415 M * WMP http://pastebin.com/54CJc0rQ
1316127459 M * WMP http://pastebin.com/khXMtnh6[
1316127461 M * WMP http://pastebin.com/khXMtnh6
1316127484 M * Bertl that's what I suspected
1316127498 M * Bertl well, good news is, everything is fine :)
1316127513 M * WMP but?
1316127520 M * WMP i havent this ip...
1316127524 M * Bertl the question now is, is it intentional to have that 46.105.133.108 on both interfaces?
1316127540 M * Bertl i.e. your host currently has it on eth0 and eth1
1316127563 M * Bertl if that is not what you actually want, use the following to get rid of that:
1316127574 M * WMP i have this
1316127577 M * WMP on alias to eth0
1316127605 M * Bertl ip a del 46.105.133.108/32 dev eth1
1316127621 M * WMP on quest, yes?
1316127626 M * Bertl on the host
1316127647 M * WMP emmm, but why?
1316127650 M * Bertl I presume you changed the 'dev' entry once and didn't stop the guest before that
1316127657 M * WMP aaa, ok
1316127675 M * Bertl leaves the IP assigned on the host, which might cause confusion
1316127704 M * WMP after this command i have in ifconfig on eth1 this ip: 87.98.237.234
1316127734 M * Bertl which is the second IP you assigned to the guest :)
1316127759 M * WMP ip a del 87.98.237.234/32 dev eth1
1316127761 M * WMP ?
1316127770 M * Bertl do you want to get rid of it?
1316127778 M * Bertl I thought you wanted a second IP on the guest?
1316127783 M * WMP yes
1316127798 M * Bertl http://pastebin.com/FzbH4Qw3
1316127807 M * Bertl this shows your config, and it matches what you see
1316127810 M * WMP but on host i have only eth0 and 9 alias 
1316127828 M * Bertl use 'ip a l' on the host, ifconfig doesn't see everything :)
1316127840 M * Bertl it's quite old and needs big glasses :)
1316127873 M * WMP http://pastebin.com/7n0xYk1b
1316127919 M * Bertl unless I misunderstood what you actually want, that looks about right to me
1316127948 M * WMP if i want have second ip to guest i must have eth1 in host?
1316127950 M * Bertl maybe you want the second IP to be on eth0 as well?
1316127963 M * WMP if it is possible
1316127981 M * Bertl yes, sure, stop the guest, change the dev file in interfaces/1 to contain eth0
1316127991 M * Bertl then start the guest again
1316127992 M * WMP ok
1316128021 M * Bertl don't worry when ifconfig only shows one IP
1316128034 M * Bertl the second is there and works fine, it will be shown with 'ip a l'
1316128040 M * WMP ok, i know about this ;)
1316128060 M * Bertl if you want it to show up on ifconfig, you need to make it an actual named alias
1316128091 M * Bertl (i.e. stop the guest, add an entry 'name' to the ifconfig/1 dir with something like 'bla' or so in it)
1316128144 M * WMP nice ;)
1316128145 M * WMP work
1316128159 M * Bertl cool! have fun!
1316128174 M * WMP thank
1316128180 M * Bertl you're welcome!
1316128250 M * WMP goodnight
1316128257 M * Bertl nn
1316128514 M * Bertl I'm off to bed now as well, have to get up early ...
1316128521 M * Bertl have a good one everyone! cya!
1316128526 N * Bertl Bertl_zZ