1109002849 T * * http://linux-vserver.org/ | latest stable 1.2.10, devel 1.9.4.7, ng9.0 -- He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime 1109002849 T * Bertl - 1109002953 M * Bertl the community of course! ;) 1109002966 M * erwan_taf :b 1109003223 Q * erwan_taf Remote host closed the connection 1109003646 J * jsambrook ~jsambrook@aelfric.plus.com 1109004230 M * Bertl roswel_home: alias + hide_netif seems sufficient for the portmap checks (well, at least find_local_ip) 1109004263 M * roswel_home ok 1109004268 M * roswel_home i checks it just now 1109004298 M * roswel_home but i don't understand where i should define the mysterious alias 1109004318 M * Bertl you have an existing new-style config? 1109004330 M * roswel_home yes 1109004348 M * roswel_home i use the latest version of alpha util-vserver 1109004356 M * Bertl okay, sec 1109004357 M * roswel_home what does mean hansi=eth0: 1109004363 M * roswel_home ? 1109004366 A * Bertl is looking at the flower page ... 1109004377 M * roswel_home hehe, ok no problem :) 1109004384 M * roswel_home it's not really clear to me in the doc 1109004425 M * Bertl # name 1109004425 M * Bertl When this file exists, the interface will be named with the text in this file. Without such an entry, the IP will not be shown by 'ifconfig' but by 'ip addr ls' only. Such a labeled interface is known as an "alias" also (e.g. 'eth0:foo'). 1109004436 M * Bertl so, what you want is to add 1109004449 M * roswel_home hummm 1109004459 M * Bertl # /etc/vservers//interfaces//name 1109004466 M * roswel_home but i've this file 1109004471 M * roswel_home aahh in iface 1109004472 M * roswel_home ... 1109004474 M * roswel_home ok sorry 1109004475 M * roswel_home ;) 1109004478 M * Bertl np 1109004479 M * roswel_home i read to fast... 1109004891 M * BWare morning 1109004898 M * Bertl morning BWare! 1109004981 M * BWare Is the old style configuration (.conf/.sh) gonna be obsolete any time soon ? 1109005074 M * Bertl heh, well, I guess according to enrico, better sooner than later 1109005086 M * BWare too bad 1109005107 M * Doener why, i never liked to old style ;) 1109005111 J * Pazzo ~thomas@host130-250.pool8172.interbusiness.it 1109005117 M * Doener s/,/?/ 1109005124 M * Doener s/to/the/ 1109005142 M * BWare I did 1109005176 M * Doener any particular reason? 1109005186 M * BWare Although I dislike sendmail configuration, I like the postfix config files 1109005220 M * BWare well, probably because of the fact that I only need to edit one file 1109005222 M * Pazzo hi doener :) 1109005229 M * Doener hi Pazzo 1109005242 M * BWare instead of finding my way around the configuration tree 1109005278 M * BWare but I guess I'll have to migrate to the new skool sooner or later 1109005320 M * ntrs_ Hi everyone. 1109005335 M * BWare hi ntrs_ 1109005336 M * ntrs_ I hope there are some companies here that use Vserver commercially. 1109005353 M * Bertl evening Pazzo! ntrs_! 1109005367 M * roswel_home hummm, Bertl , i created /usr/local/etc/vservers/hestia/interfaces/0/name with a name inside, and nothing happened... could i forget someting? 1109005368 M * Doener ntrs_: at least lycos (rs) is here 1109005395 M * ntrs_ We all need to come up with some sort of a community fund to support the vserver development team. Initially, we need to take care of Bertl and eventually include other developers including Doener and ensc (Enrico) 1109005417 M * ntrs_ Any pledges or at least ideas on how to implement this the best? 1109005462 M * Pazzo hi bertl! 1109005462 M * ntrs_ I know of at least two companies that have already pledged certain monthly amounts to support vserver development. 1109005467 M * BWare Well, I was thinking about donating .50 euro of every sold vserver ;) 1109005518 M * Bertl roswel_home: did you restart hestia? 1109005523 M * ntrs_ Ok. I was thinking more of a fixed amount that all of you can set aside each month regardless of your sales. It has to be a monthly recurring amount. If you care about this product we need to support it and we need to support it right now 1109005526 M * roswel_home of course i did... 1109005537 M * Bertl and was an alias created? 1109005544 M * roswel_home no 1109005557 M * roswel_home i should see it on my root server is it? 1109005564 M * Bertl please restart with --debug again and upload the output somewhere ... 1109005582 M * ntrs_ If we do not support it financially it may end up on the dark side and I am sure non of you want that to happen. 1109005643 M * Bertl hmm, community fund sounds like a great idea! 1109005681 M * ntrs_ Absolutely, and it is really necessary if we are planning to be able to reject certain hostile takeovers of the main developers and such. 1109005723 M * roswel_home i agree too ntrs_ :) 1109005737 M * Bertl well, I guess there are not that many companies present now and here ... 1109005757 M * ntrs_ You can pledge your contribution to Bertl directly in private or also publicly here in the channel. As long as Bertl and the other developers can make ends meet, this project will be alive and kicking. 1109005790 M * Pazzo Bertl: "diffing" around I discovered the new note about the "legacy API" in Kconfig - I think this will help some people to not run into trouble :o) 1109005799 M * Bertl ntrs_: so maybe we should take the idea to the mailinglist (probably other companies will be subscribed there too) 1109005818 M * ntrs_ Probably not. Bertl, maybe you can put some info to the topic above and also send that to the mailing list. 1109005821 M * ntrs_ Right. 1109005826 M * ntrs_ I was just about to say that. 1109005829 M * ntrs_ :) 1109005873 M * Bertl okay, hmm .. well, it's your idea, so would you like to present it on the ml, or should I write some key-note? 1109005934 M * ntrs_ I can post something, but you probably know the situation better than anyone else. You can say that the initiative came from chanel 1109005953 M * Bertl Pazzo: yes, that's the intention, after somebody disabled the legacy option and ran into big troubles ... 1109005992 M * Bertl okay, I'll post something and you follow up with some details and/or ideas ... 1109006020 M * Pazzo Bertl: 2.6.11-rc2_vs1.9.4-rc4 is still running fine - I'll put another not-so-important host online today, probably with 1.9.4.7. any concerns about doing so? if not I'll do so and let you know if it's doing a good job :o) 1109006061 M * Bertl no, 1.9.4.7 should be pretty fine ... 1109006133 M * ntrs_ Bertl, ok lets do that then 1109006148 M * Pazzo is 2.6.11-rc4 pretty fine too? I don't like vanilla 2.6.10... 1109006164 M * Bertl yeah, it looks like, sec 1109006521 M * Bertl # uptime 1109006521 M * Bertl 18:24:57 up 2 days, 18:17, 1 user, load average: 26.91, 26.32, 26.28 1109006529 M * Bertl Linux quad 2.6.11-rc4-vs1.9.4.7 #2 SMP Fri Feb 18 22:40:59 CET 2005 i686 unknown unknown GNU/Linux 1109006555 M * Bertl it's under heavy stress testing since it was booted ... 1109006604 M * Bertl processes 2068443867 1109006612 M * Bertl I guess that gives some idea ;) 1109006635 M * Bertl (from /proc/stat) 1109006774 M * Pazzo thnx bertl! nice load :o) 1109006817 M * Pazzo processes 2068443867? fork bomber's paradise... 1109006936 M * Bertl yep, killer* 1109006949 M * Bertl that's why it took a little to logon ;) 1109007405 M * Bertl ntrs_: okay, I hope the posting reflects what you meant ... 1109007501 M * ntrs_ Yes, exactly, that was it. Very well said. 1109007510 M * roswel_home Bertl, ok i got the alias bug 1109007530 M * roswel_home it's because a change the configuratoin while the vserver was running, 1109007550 M * roswel_home then as i was stopping my vserver ++ set -- 192.168.0.16 broadcast + label eth0:hestia dev eth0 1109007550 M * roswel_home ++ /sbin/ip addr del 192.168.0.16 broadcast + label eth0:hestia dev eth0 1109007550 M * roswel_home RTNETLINK answers: Cannot assign requested address 1109007569 M * roswel_home cause the alias was'nt created, as i just modified the conf 1109007593 M * roswel_home the when i try to start the vserver again, the ip address is still active, so i got 1109007608 M * roswel_home ++ set -- 192.168.0.16 broadcast + label eth0:hestia dev eth0 1109007609 M * roswel_home ++ /sbin/ip addr add 192.168.0.16 broadcast + label eth0:hestia dev eth0 1109007609 M * roswel_home RTNETLINK answers: File exists 1109007640 M * Bertl i.c. just remove the entries with ip, and restart it 1109007642 M * roswel_home now it works fine :) 1109007660 M * roswel_home yep i did it, no problem 1109007698 M * roswel_home this is not a critical bug, but hsould be reported to ensc 1109007700 M * Bertl and does portmap work now? 1109007707 M * roswel_home yep it work great :) 1109007710 M * roswel_home Thank you 1109007730 M * Bertl well, changing the config while the vserver is running is a big nono ... 1109007764 M * roswel_home humm you're right, 1109007792 M * roswel_home anyway if you have many paramter to change could be better than to stop your vserver during 10 minutes. 1109007823 M * Bertl you can basically change most of the parameters while the vserver is running 1109007831 M * Bertl but you have to do it differently 1109007845 M * roswel_home differently? 1109007855 M * Bertl ntrs_: thanks for the great idea! 1109007873 M * ntrs_ You are welcome. 1109007882 M * Bertl roswel_home: well, that's something for the future (tools) 1109007911 M * Bertl basically flags and ccaps as well as all limitation can be changed while the vserver is running 1109007918 M * ntrs_ I hope more companies will realize the importance of this proposed fund, unless they want to end up having to use virtuozzo for the rest of their lives. 1109007928 M * Bertl roswel_home: of course not by editing the files, but by using special tools for that 1109007968 M * roswel_home ok 1109007999 M * Bertl I could imagine a solution where you would 'adapt' the config with vserver config ... calls 1109008000 M * roswel_home And do you know why this alias is not done by default? 1109008029 M * Bertl and the script could then update the config, and 'adjust' the setting (which can be adjusted) on the fly 1109008051 M * roswel_home humm could be nice! :) 1109008091 M * Bertl ad alias: simple, the alias is not really used/required by most tools/programs (at least recent ones) and it makes some things simpler not to have it 1109008446 M * Doener hm, the only problematic changes are ip addresses and context number i guess... the former will probably be basically gone with ngnet, the latter is a bad thing to change while running ;) 1109008474 M * Doener the problem will be gone of course, not ip addresses ;) 1109008474 M * Bertl precisely, but changing the xid would be possible ... 1109008518 M * Doener Bertl: when you start using tagctx it gets worse ;) 1109008854 Q * jsambrook Quit: Download Gaim: http://gaim.sourceforge.net/ 1109008910 M * Bertl Doener: yep definitely ... 1109008941 M * daniel_hozac so what's the correct way to figure out the IP address? 1109008946 M * daniel_hozac netlink foo? 1109008966 M * Bertl ip (from iproute2) shows everything (so yes) 1109008983 M * daniel_hozac i mean programatically, like in say portmap. 1109008987 M * Bertl but the 'alias' option is still there, so it's fine to use that 1109009006 M * Bertl yeah, I'd say netlink it is ... 1109009029 M * daniel_hozac well, i like to fix the issues rather than work around them ;) 1109009057 M * daniel_hozac does glibc have some wrappers for the netlink stuff? 1109009067 M * daniel_hozac getifaddrs or something like it? 1109009124 M * Bertl I hope so, (man netlink) 1109009128 M * Doener getifaddrs is patched in some distros to use netlink 1109009133 Q * rs Quit: leaving 1109009158 M * Doener f.e. debian's glibc has the patch, and IIRC redhat/fedora use that patch, too 1109009171 M * daniel_hozac right, thanks. 1109009238 M * roswel_home Thanx for help to everybody :) Have a good night. Time to sleep here... 1109009244 N * roswel_home roswel_Zzz 1109009269 M * Bertl night roswel_Zzz! 1109009579 J * matta ~matta@69.93.28.254 1109009744 M * matta Bertl ? 1109009784 M * matta read the Virtuozzo thing... 1109009799 M * Bertl hey matta! 1109009835 M * matta hi! 1109009839 M * Bertl how are you, fine? 1109009846 M * matta yeah, still going well 1109009852 M * matta work a lot :) 1109009877 M * matta I had some input on the situation, since I still keep an interest in vserver and still run it on a few servers here 1109009895 M * matta this has got to be a tough situation for you, i'm sure swsoft is will to pay a pretty penny for your services :) 1109009944 M * Bertl well, I guess so ... 1109009970 M * matta the downside is knowing that VServer project will die without you, and swsoft will have eliminated the only other technogly that prevents them from being a complete monopoly in there area of expertise 1109009980 M * Bertl Doener: do you remember which option (kernel wise) enables the nat table? 1109009998 M * matta Bertl: ip_conntrack + full nat 1109010000 M * Bertl matta: what about VPS? 1109010007 M * matta Bertl: yeah. 1109010009 M * Bertl ah the contrack, tx! 1109010020 M * Doener 2fast4me ;) 1109010035 M * Bertl yeah, matta is smart! ;) 1109010056 M * matta there is uml and xen... i don't personally like it anymore, xen is quite interesting... but virtuozzo/vserver and uml/xen are completely different 1109010101 M * Bertl yes, uml/xen has other qualities ... 1109010181 M * Bertl matta: do you utilize virtuozzo? maybe a recent version? 1109010209 M * matta i'd be willing to send a paypal subscription to you, can't afford a whole lot... but i'd do just if it would help ensure vserver still exists as a project, mainly because I spent so much time on it previously 1109010219 M * matta Bertl: yep... I have lots of virtuozzo servers 1109010228 M * matta 2.6.0/2.6.1 1109010252 M * Bertl would you mind a few questions regarding this product? 1109010650 J * prae ~prae@sherpadown.net 1109010668 M * Bertl welcome prae! 1109010692 M * prae Hi Bertl ! 1109011039 M * lars||| just a quick note: my company would also donate money to the vserver project as soon as we have some vservers in production :) 1109011064 M * Bertl lars|||: thanks for the feedback! 1109011607 J * jfried ~jfried@p3EE0B233.dip.t-dialin.net 1109011608 M * jfried hi 1109011723 M * Bertl welcome jfried! 1109011810 M * jfried is it possible to have network inside a vserver?. i have a rootserver from server4you with one nic interface and a public ip. in my vservers i have private ip's (192.168.0.*). the vservers are on virtual devices (eth0:apache, for ex.). but when i ssh into the vserver i cannot ping the world 1109011894 M * Bertl which vserver version is installed there? 1109011925 M * jfried kernel: 2.6.10-vs1.9.4, util-vserver 0.30.203 1109011986 M * Bertl okay, you should add a capability (ccapability) to allow ping to work ... 1109012028 M * Bertl now looking how it is called within the tools ;) 1109012053 M * Bertl DECL("icmp", VC_VXC_RAW_ICMP), 1109012063 M * Bertl // some deprecated values... 1109012064 M * Bertl DECL("ping", VC_VXC_RAW_ICMP), 1109012076 M * Bertl so icmp is the better way to go ... 1109012102 M * Bertl * /etc/vservers//ccapabilities 1109012119 M * Bertl (I guess you are using the new-style config, right?) 1109012136 M * jfried bertl, no the old one 1109012164 M * Bertl okay, then you have to set it manually after vserver start 1109012203 M * jfried with? 1109012204 M * Bertl vattribute --set --xid --ccap icmp 1109012208 M * jfried a thank you 1109012218 M * Bertl you're welcome! 1109012431 M * jfried hm must i set a default gateway inside the vserver? route add default gw does not work 1109012459 M * Bertl no, the routing (for now) is still done (and configured) on the host 1109012461 M * Doener routing and stuff is handled within the host, not the vserver 1109012505 M * Doener you need to do some SNAT to make your private ip's get access to the outside 1109012539 M * jfried Doener, well but the private ips are all on virtual devices 1109012549 M * Doener there's no such thing ;) 1109012570 M * jfried Doener, that means? no internet for the vservers? 1109012570 M * Doener eth0:apache is just an aliased ip addresse that is assigned to eth0 1109012590 M * jfried Doener, yes thats right (for me it is a "virtual" device) 1109012605 M * Doener with the right SNAT (and DNAT?) rules you should get access to the outside... 1109012629 M * jfried hm 1109012632 M * Bertl jfried: simple example here (because I'm just answering an email) 1109012642 M * Doener something like: iptables -t nat -A POSTROUTING -o eth0 -s 192.168.0.x -j SNAT 1109012655 M * Bertl eth0 Link encap:Ethernet HWaddr 52:54:00:12:34:56 1109012659 M * Bertl inet addr:10.0.0.2 Bcast:10.255.255.255 Mask:255.0.0.0 1109012663 M * Bertl eth0:XXXX Link encap:Ethernet HWaddr 52:54:00:12:34:56 1109012667 M * Bertl inet addr:192.168.0.2 Bcast:10.255.255.255 Mask:255.0.0.0 1109012670 M * Doener --to-source 1109012678 M * Doener i always forget the important parts ;) 1109012702 M * Bertl I have the following entries in my iptables: 1109012708 M * Bertl iptables -t nat -A PREROUTING -p tcp -d 10.0.0.2 --dport 80 -j DNAT --to-destination 192.168.0.2 1109012725 M * Bertl iptables -t nat -A POSTROUTING -s 192.168.0.2 -j SNAT --to-source 10.0.0.2 1109012762 M * Bertl this allows a host on 10.x.x.x to reach the vserver XXXX on 192.168.0.2 via the 'official' 10.0.0.2 address for http 1109012797 M * Bertl and it also allows the vserver (from inside) to reach the internet via 10.0.0.1 (default gateway) 1109012812 M * jfried bertl, ok i'll try it thank you 1109012823 M * Bertl jfried: if you replace the 10.x.x.x by your 'public' ip that will work fine 1109012838 M * jfried bertl, i'll try it :) 1109012860 M * Bertl k, let us know if you encounter any issues ... or have some further questions ... 1109012881 M * jfried sure :) 1109012988 M * jfried bertl, yeah right :) it works :-) 1109013089 M * Bertl excellent! 1109013128 M * jfried :) 1109013182 Q * prae Quit: Pwet 1109013216 J * tad ~truex@static-151-204-232-50.bos.east.verizon.net 1109013233 M * Bertl welcome tad! 1109013247 M * daniel_hozac does http://daniel.hozac.com/vserver/portmap-4.0-getifaddrs.patch look sane? it seems to work alright... 1109013343 M * Bertl looks fine to me ;) 1109013884 M * tad Greetings all. I am working my way through installing an FC3 vserver using the yum install method. I had problems early on because the util-vserver packages don't have any GPG keys with them. Copy the keys from a known good source to the fc3/pubkeys directry in /usr/lib/util-vserver/distributions directory got me over the initial hurdle and allowed me to do the install 1109013892 M * jfried bertl, i do not need the vattribute, perhaps because of this S_CAPS="CAP_NET_RAW" in my config file? 1109013917 M * tad Now I'd like to get networking running, but get a bit confused around netdev and ip stuff in the alpha utils branch. 1109013926 M * Bertl jfried: well, you should get rid of that CAP_NET_RAW thingy 1109013952 M * jfried bertl, ok 1109013961 M * Bertl tad: ask away ... 1109013988 M * Bertl jfried: unless you want vserver root to be able to sniff host packages ... 1109014010 M * tad On my server, I have a number of ip's aliased to eth1, I'd like to use the one bound to eth1:1 (192.168.1.50 for example) for the vserver called fc3 1109014042 M * tad I can't quite tell which files under /etc/vserver/fc3 to edit - I assume ip and adding nodev would do it, but no dice 1109014050 M * Bertl the vserver is already created or to-be-created? 1109014075 M * tad Already created - vanilla -d fc3 install. No networking stuff speced on the build line 1109014083 Q * JViz Ping timeout: 480 seconds 1109014118 M * Bertl okay, so if you had used --interface 192.168.1.50/24 or something like that, the tools would have created the entries for you .. but you can simply do that yourself 1109014154 M * Bertl the key issue here usually is the directory hierarchy 1109014175 M * tad OK - do I need a nodev file? (It looked to me like that would use the already existing alias on the host) 1109014194 M * Bertl /etc/vservers//interfaces/ 1109014202 M * Bertl yes, this is the dir you want to create 1109014220 M * Bertl where vserver-name is fc3 and iface is for example 00 1109014237 M * tad I think I'm mostly there. I have /etc/vservers/fc3/interfaces/0/ with an ip file (with the address in it - no netmask) 1109014272 M * Bertl then you want to populate this with bbcast, nodev, prefix, mask ip but not dev 1109014305 M * Bertl prefix and mask are basically the same 1109014315 M * Bertl no dev and no name entries 1109014316 M * tad If I add a /24 to the ip does it figure that stuff out automatically? 1109014336 M * Bertl nope, ythe 24 has to go into prefix then ... 1109014613 M * Bertl ah, just because we had that a few moments ago ... 1109014630 M * Bertl of course it is advantageous to stop the vserver before you change the config ;) 1109014672 M * tad OK - May be getting closer. In the vserver, I now see eth1 with MAC but no IP, and eth1:1 with the correct IP (and the same MAC). This is roughtly what I see on the host 1109014688 M * tad eth1 has an ip address on the host of course 1109014710 M * tad routing tables all point to eth1, and don't appear to get me off the host 1109014712 M * Bertl the ifconfig output inside the vserver is fine ... 1109014747 M * Bertl you do not do routing to the vserver, as networking happens on the host 1109014778 M * Bertl i.e. if your setup has everything in place for routing (without any vservers) it will work fine with the vserver 1109014808 M * tad Let me look into that - the machine I'm using had been a router long ago... Probably some bit rot. 1109014814 M * Bertl you might need to SNAT/DNAT ips if you do not want to use the vserver's ip in outbound connections 1109014869 Q * nano_ Quit: Cerrando la cliente 1109014876 M * Bertl simple tests involve ping -I and tcpdump on the host 1109014962 M * tad I suspect I have forwarding turned off, but have lost the place to look 1109014984 M * Bertl forwarding is not really involved 1109014995 M * Bertl as the host does not do forwarding to itself 1109015016 M * tad From within the vserver I can ping the host, but no one on the network 1109015016 M * Bertl maybe you could explain your network setup a little 1109015032 M * Bertl on the eth1 network? 1109015078 M * tad yes 1109015111 M * Bertl what if you try to ping a host on the eth1 network with ping -I on the host? 1109015119 M * tad The network is pretty strigtforward, the machine I'm on is 192.168.1.22 (one enet card at eth0) 1109015140 M * tad I have a single alias (192.168.1.50) configured for eth1:1. 1109015158 M * Bertl well, and what is connected to eth1? 1109015160 M * tad The rest of the world is reached through a dlink router at 192.168.1.1 1109015197 M * tad Sorry - typo - the only card is on eth1 with an alias eth1:1 - there is no eth0 1109015221 M * Bertl okay, well, then let's try with: 1109015235 M * Bertl ping -I 192.168.1.50 192.168.1.1 (on the host) 1109015249 M * Bertl if that works as expected, let's try: 1109015271 M * Bertl ping -I 192.168.1.50 google.com 1109015391 M * tad I think you are onto it. The first ping (on the host) didn't work 1109015417 M * Bertl and the second did work? 1109015483 M * mugwump http://sourceforge.net/projects/linuxjail/ # jailing LSM module 1109015535 M * Bertl thanks! it's Sergei's project, right? 1109015537 M * tad Nope - looks like 192.168.1.50 is blocked somewhere 1109015561 M * Bertl probably the router does not accept that ip ... 1109015581 M * Bertl you can work around it by 'reusing' the 192.168.1.22 ip 1109015596 M * tad Don't know why - there are lots of other machines around that are off right know, but use a variety of address on this network. 1109015604 M * mugwump http://kerneltrap.org/node/3823 # original posting, yes "Serge Hallyn" according to that post 1109015633 M * Bertl yeah, we had a longer email exchange regarding that lsm ... 1109015702 M * mugwump was that cc'ed to the list ? 1109015717 A * mugwump checks 1109015724 M * Bertl hmm, guess part of it should be on lkml ... 1109015729 J * prae ~prae@sherpadown.net 1109015734 M * Bertl wb prae! 1109016275 M * tad OK - I had a thinko... My firewall was blocking most icmp packets. Fixed that and now the vserver seems to be doing just fine. 1109016277 M * tad Thanks! 1109016300 M * Bertl you're welcome! 1109016334 M * Bertl don't forget to add yourself to the list of happy linux-vserver users (or companies) once you're happy ;) 1109016361 M * Bertl (unless you don't want to do that, of course) 1109016489 M * tad Will do - With any luck I'll even contribute something back! 1109016504 M * Bertl that would be great! 1109016590 P * tad Leaving 1109019380 M * lars||| I get: 1109019380 M * lars||| * The configured vshelper '/sbin/vshelper' does not match the 'vshelper' 1109019380 M * lars||| script of the util-vserver package 1109019423 M * lars||| I've compiled util-vserver with no problems 1109019461 M * Bertl well, probably the script is right ;) 1109019482 M * Bertl look for vshelper scripts and check if they differ ... 1109019497 M * Bertl (locate, find, diff, ls -la should be of help there) 1109019506 J * sylvio sylvio@lea.boerde.de 1109019557 M * lars||| ehmm ok! 1109019577 M * Bertl welcome sylvio! 1109019747 M * mugwump reading this thread on lkml I can't help but think that the reason that Serge wrote that LSM module is because linux-vserver wasn't included in the main kernel earlier 1109019789 M * mugwump earlier than the current target release of 2.7.823 1109019821 M * Bertl hmm, yes, not so wrong ... 1109019859 M * Bertl one of his latest replies to my mails asking why he doesn't improve linux-vserver, was along the lines: 1109019896 M * Bertl we are going to improve that module to full linux-vserver functionality later ... 1109019992 M * mugwump ... and he seems to be taking the approach of not bothering to explore the history of changes of the code base of linux-vserver. after all, what do we know about coding kernel stuff anyway 1109019998 M * mugwump 1109020072 M * Bertl hehe ... well, I take this as a good example what the LSM framework is _not_ able to do, and what can be done there ... 1109020611 M * eyck is there anything that LSM framework CAN do? 1109020955 J * JViz Anomaly@gso167-184-225.triad.rr.com 1109021143 M * Bertl eyck: hum, of course ... 1109021148 M * Bertl welcome JViz! 1109021269 J * duckx ~Duck@dyn-83-157-138-21.ppp.tiscali.fr 1109021475 M * Bertl wb duckx! 1109021491 M * matti Uf. 1109021504 M * Bertl hey matti! :] 1109021521 M * matti Hello Bertl ;] :] 1109021528 M * matti What's up? 1109021529 M * matti :> 1109021566 M * Bertl well, nothing extraordinary ... 1109021588 M * matti No news is good news ;] 1109021612 M * Bertl is it? 1109021641 M * matti In some point of view. 1109021642 M * matti ;] 1109021652 M * Bertl of course! ;) 1109021655 M * matti :> 1109021795 M * duckx hy :) 1109021816 M * duckx Have heard that roswel spent some time here 8) 1109021841 M * Bertl yeah, was that a bad thing? 1109021849 M * duckx Nop :) 1109021861 M * duckx Pretty pleasant ... 1109021872 M * duckx I got no time those days ... busy @work 1109021894 M * duckx And party of course 8) 1109021934 M * Bertl of course ... 1109021938 M * duckx did he talk about my multicast issue ? 1109021973 M * Bertl hmm, maybe ... 1109021985 M * duckx I tried to setup slp for fun ... 1109021993 M * duckx But I had no success 1109022016 M * duckx Seems the multicast don't go through the interface ... 1109022047 M * duckx As far as I remember I played with the capabilities ... with no success :( 1109022053 M * Bertl did you assign any multicast addresses? 1109022061 M * duckx Nop 1109022076 M * duckx By default it should go to the default route ... 1109022083 M * Bertl then it probably wasn't able to bind to that ip/port 1109022104 M * duckx Hmm let me see :) 1109022187 M * duckx Listening on loopback... 1109022187 M * duckx Couldn't bind to multicast for interface 192.168.0.10 (Cannot assign requested address) 1109022194 M * duckx Is the error I get ... 1109022238 M * Bertl sounds like that's the issue ... try adding the multicast ips to the vserver 1109022271 M * duckx route add -net 224.0.0.0 ... 1109022285 M * duckx can't rememember the end :) 1109022313 M * Bertl hmm, well, I thought more about ../interfaces/MC/ip here ... 1109022344 M * Bertl (or in legacy style: IPV4ROOT=" ... 224.0.0.0/8" or so ... 1109022406 M * duckx ok 1109022410 M * duckx Let me test this :) 1109022475 M * duckx What should I put /interfaces/MC/ip ? 1109022488 M * duckx Dump question, but I don't know really well multicast !( 1109022491 M * Bertl 224.0.0.0 1109022494 M * duckx ok :) 1109022509 M * Bertl if that is what you normally use ... no idea 1109022626 M * Bertl if your app uses 225.0.0.37, then of course, you have to specify that 1109022777 M * duckx Damn ... 1109022784 M * duckx The reboot bug is back ! 1109022804 Q * lars||| Quit: using sirc version 2.211+KSIRC/1.3.11 1109022895 M * duckx vservers are down !( 1109022982 J * hillct ~hillct@65.76.245.91 1109023008 M * Bertl welcome hillct! 1109023024 M * hillct hello Bertl 1109023050 M * Bertl duckx: rebootbug as in your host interfaces goes down? 1109023069 M * duckx vserver xxxx stop stops the server ... 1109023140 M * duckx But also the host server ... 1109023228 M * Bertl well, as I tried to explain last time, that's no bug, that is a configuration issue 1109023258 M * Bertl basically it's the trickier version of putting 'ifconfig eth0 down' in the shutdown script ... 1109023296 M * duckx Yop 1109023305 M * duckx It's may be a capabilities issue :( 1109023466 M * duckx vs-reboot (: failed to exec (/sbin/vshelper) 1109023512 M * Doener night folks! 1109023516 N * Doener Doener_zZz 1109023519 M * duckx Night 1109023531 M * Bertl night Doener_zZz! 1109023695 Q * prae Quit: Pwet 1109023760 M * duckx Server back m:) 1109023791 Q * duckx Quit: Client exiting 1109024264 Q * jfried Quit: Leaving 1109024422 J * prae ~prae@sherpadown.net 1109025866 Q * prae Quit: Pwet 1109026960 P * Pazzo 1109028652 M * Bertl night everyone, off to bed now ... 1109028665 N * Bertl Bertl_zZ 1109028952 P * hillct