1326931233 Q * dowdle 1326931931 N * Bertl_oO Bertl 1326931934 M * Bertl back now ... 1326934159 J * psycodad ~cschnee@snappy.schneebi.com 1326939159 J * fLoo ~fLoo@31-19-187-252-dynip.superkabel.de 1326940932 M * Bertl off to bed now ... have a good one everyone! 1326940938 N * Bertl Bertl_zZ 1326946275 Q * FireEgl Quit: Leaving... 1326947089 Q * micah Ping timeout: 480 seconds 1326951264 J * sannes ~ace@cm-84.209.106.118.getinternet.no 1326954363 J * micah ~micah@micah.riseup.net 1326957126 J * ghislain ~AQUEOS@adsl2.aqueos.com 1326958364 J * thierryp ~thierry@home.parmentelat.net 1326958376 Q * thierryp Remote host closed the connection 1326958408 J * thierryp ~thierry@home.parmentelat.net 1326960018 J * Marbug ~Marbug@83.101.67.3 1326960554 M * DelTree is it possible to have the host listen to 0.0.0.0:foo and a guest listening on foo too ? never managed to make it... 1326961035 J * petzsch ~markus@dslb-088-075-170-200.pools.arcor-ip.net 1326961951 Q * petzsch Quit: Leaving. 1326963150 M * daniel_hozac no. 1326963163 M * DelTree damn... 1326963166 M * daniel_hozac since host's 0.0.0.0:foo includes all the guest's addresses.... 1326963194 M * DelTree then I'm f**ked with my nfs problem... 1326963278 M * DelTree great piece of crap going on... really... ^_^ 1326963358 M * DelTree nothing will work if portmap doesn't listen 127.0.0.1, but it won't go far if it only listens 127.0.0.1, but it will only listen 127.0.0.1 or not listen 127.0.0.1 unless it listens 0.0.0.0... f**k... 1326963377 Q * ensc|w Remote host closed the connection 1326963386 J * ensc|w ~ensc@www.sigma-chemnitz.de 1326963470 M * daniel_hozac so switch to a better portmap daemon, or put it in a network context. 1326963473 M * DelTree (is it understandable that way ?) 1326963486 M * DelTree better portmap daemon ? where ? 1326963518 M * DelTree anywhere at aptitude reach ? 1326963534 M * daniel_hozac i would have no idea 1326963793 M * DelTree ok... then my present conclusion is that it's a complete no-go... ^_^ 1326964084 M * daniel_hozac sure, you can never accomplish anything without putting some work... 1326964114 M * DelTree :) 1326964175 M * DelTree here "some work" would mean hacking some piece of software hard... not a thing that I can do right now... not a thing I can do at all in the case of nfs-kernel-server... 1326964225 M * daniel_hozac no 1326964246 M * daniel_hozac it means running chbind --nid 42 --ip / service portmap restart 1326964258 M * DelTree ? 1326964292 M * DelTree what's that dark magic spell ? 1326964372 M * daniel_hozac the fundamentals of Linux-VServer. 1326964650 M * DelTree so what should I do with that ? 1326964670 M * DelTree force that portmap on the host to release the guest addresses ? 1326966205 M * hijacker DelTree, this should start portmap inside context 42 on the IP address specified and run inside it 1326966302 M * DelTree :) 1326966321 M * DelTree and the host will be able to contact it *on 127.0.0.1* ? 1326966524 M * daniel_hozac if that's the address you specify. 1326966554 M * DelTree but there the guests won't be able to contact it... 1326966572 M * DelTree not anyone else at that... 1326966577 M * DelTree nor* 1326966594 M * daniel_hozac you can specify as many IP addresses as you wish... 1326966654 M * DelTree ok... I'll have to try it.. 1326968884 Q * fisted_ Read error: Connection reset by peer 1326969109 J * fisted ~fisted@xdsl-87-78-220-147.netcologne.de 1326971103 J * sladen ~paul@starsky.19inch.net 1326971222 J * BenG ~bengreen@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com 1326971516 N * Bertl_zZ Bertl 1326971520 M * Bertl morning folks! 1326972794 Q * LuckyLuke Ping timeout: 480 seconds 1326973952 J * LuckyLuke ~luca@host65-83-static.228-95-b.business.telecomitalia.it 1326974868 Q * LuckyLuke Ping timeout: 480 seconds 1326975092 J * LuckyLuke ~luca@host65-83-static.228-95-b.business.telecomitalia.it 1326977448 Q * thierryp Remote host closed the connection 1326978699 Q * arekm Quit: leaving 1326978920 Q * Aiken Remote host closed the connection 1326981650 J * LuckyLuk1 ~luca@host65-83-static.228-95-b.business.telecomitalia.it 1326981650 Q * LuckyLuke Read error: Connection reset by peer 1326982703 Q * LuckyLuk1 Ping timeout: 480 seconds 1326982908 J * LuckyLuke ~luca@host65-83-static.228-95-b.business.telecomitalia.it 1326983091 M * Bertl off for a nap ... bbl 1326983103 N * Bertl Bertl_zZ 1326983184 Q * ccxCZ Remote host closed the connection 1326986494 N * ensc Guest24518 1326986504 J * ensc ~irc-ensc@p54ADF28B.dip.t-dialin.net 1326986913 Q * Guest24518 Ping timeout: 480 seconds 1326987435 Q * BenG Quit: I Leave 1326987470 J * BenG ~bengreen@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com 1326987486 Q * BenG 1326988108 J * dowdle ~dowdle@scott.coe.montana.edu 1326989215 N * Bertl_zZ Bertl 1326989222 M * Bertl back .. 1326991340 J * ccxCZ ~ccxCZ@new.webprojekty.cz 1326993110 Q * eyck_ Remote host closed the connection 1326993235 J * bonbons ~bonbons@2001:960:7ab:0:400d:b435:103d:ca99 1326993595 J * eyck ~eyck@77.79.198.60 1326994063 Q * guerby Read error: No route to host 1326994515 M * ghislain hello there, i have a little question, for vroot system why do we put UFS as the filesystem instead of the real one ? 1326994582 M * Bertl to trick certain versions of the quotatool into using the kernel interface instead of direct low-level device i/o 1326994705 J * guerby ~guerby@nc10d.tetaneutral.net 1326996160 M * ghislain ok so i guess we should NOT change that then :p 1326996184 M * ghislain never let a chance for quotatool to try to be smart 1326996189 M * ghislain too dangerous 1326996257 M * Bertl yup 1326996728 Q * fLoo Ping timeout: 480 seconds 1326996850 Q * bonbons Quit: Leaving 1326996868 J * bonbons ~bonbons@2001:960:7ab:0:400d:b435:103d:ca99 1326997213 Q * daniel_hozac Ping timeout: 480 seconds 1326997743 J * daniel_hozac ~daniel@c-bb3271d5.08-230-73746f22.cust.bredbandsbolaget.se 1326998392 M * Bertl off for now .. bbl 1326998396 N * Bertl Bertl_oO 1327000671 Q * ncopa Quit: Leaving 1327001283 J * BenG ~bengreen@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com 1327001698 Q * BenG Quit: I Leave 1327004243 Q * geb Ping timeout: 480 seconds 1327004635 J * destructive ~Adium@ip-83-99-90-132.dyn.luxdsl.pt.lu 1327004645 M * destructive hi 1327004685 J * geb ~geb@mars.gebura.eu.org 1327004738 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f 1327004860 M * destructive where can i find some docs about how vserver guests are connected to the network? 1327005084 J * petzsch ~markus@p57B66AD4.dip.t-dialin.net 1327005607 M * Guy- destructive: have you tried the wiki at linux-vserver.org? 1327005777 M * destructive yes, haven't found what i was looking for 1327005888 M * destructive for example I would like to know which interface a guest uses by default for networking 1327006337 M * fback destructive: by default all networking is done on the host 1327006422 M * fback destructive: you have to specify which interface(s) belong to which guest 1327006441 M * fback destructive: and optionally you can get auto loopback 1327006579 N * Bertl_oO Bertl 1327006583 M * Bertl back now ... 1327006636 M * Bertl destructive: by default, Linux-VServer uses IP isolation, i.e. you assigned a subset of the host IPs to each guest 1327006670 M * Bertl the interfaces carrying at least one of those IPs will be visible inside the guest 1327006707 M * Bertl the guest itself can bind to one or all of the assigned IPs to run services on, but not to IPs outside the assigned set 1327006770 M * destructive i made me this little script to setup the networking: http://pastebin.com/dsjuTb5A 1327006806 M * destructive is now all traffic of my vserver going through nat? 1327006997 M * destructive so in general all network connections of vservers are using one of the interfaces specified in /etc/vservers/$VSERVER/interfaces/ ? 1327007429 M * Bertl no, the 'dev' entry specified in the guest config is used by util-vserver to add the IP when the guest is started and to remove it when the guest is shut down 1327007543 M * Bertl and network conenctions do not 'use' interfaces in the strict way of speaking, routing decides where a packet is sent to 1327007648 M * Bertl forwarding, as you 'enable' in your script is not involved in traffic originating from a guest, as it is basically originating from the host, so only output, not forwarding 1327008198 M * destructive thanks for the explanation 1327008307 M * Bertl but your postrouting rule will work (at least for outgoing traffic), as it doesn't involve the interface .. 1327008354 M * Bertl the host will use ethX (whichever interface used to send traffic to) 1327008422 M * destructive ok, i don't know if you have already recognized but i tried to follow this setup: http://linux-vserver.org/Networking_vserver_guests 1327008532 M * Bertl I'm not sure where you're heading to, i.e. what's the plan in the end? 1327008688 M * destructive i have a server with one external ip address, but at the moment i'm testing vserver in a virtualbox 1327008722 J * hijacker_ ~hijacker@cable-84-43-136-96.mnet.bg 1327008728 Q * hijacker_ Remote host closed the connection 1327008740 M * Bertl okay, so playing around for now ... good 1327008752 M * destructive yes 1327009082 M * destructive Where does a guest send packets to? to the broadcast? 1327009087 Q * dowdle 1327009123 M * Bertl it depends on the routing setup on the host 1327009150 M * Bertl the kernel only ensures that a guest IP is used as source IP 1327009179 M * Bertl this allows to have guest specific routing (based on the source IP) on the host 1327009200 M * Bertl i.e. that way you can control/manipulate the routing on a per guest basis 1327009458 M * destructive but that iptables rule uses the broadcast address as source address? 1327009491 M * daniel_hozac "that" rule? 1327009508 M * destructive iptables -t nat -A POSTROUTING -s $DUMMY_BROADCAST/$DUMMY_PREFIX ! -d  $DUMMY_BROADCAST/$DUMMY_PREFIX -j SNAT --to-source $HOSTIP 1327010033 Q * sannes Remote host closed the connection 1327010176 Q * bonbons Quit: Leaving 1327010785 Q * petzsch Quit: Leaving. 1327011152 M * destructive Why are there in my "ip route" output of the guest other routes than that which uses the dummy interface? 1327011152 M * destructive 10.0.2.0/24 dev if2 proto kernel scope link src 10.0.2.15 1327011152 M * destructive 169.254.0.0/16 dev if2 scope link metric 1000 1327011152 M * destructive 172.16.0.0/12 dev dummy0 proto kernel scope link src 172.16.0.100 1327011152 M * destructive default via 10.0.2.2 dev if2 1327011170 M * daniel_hozac because that is your routing table. 1327011199 M * daniel_hozac if it would only use the dummy interface, it would never go anywhere. 1327011207 M * daniel_hozac since dummy interfaces are the /dev/null of networking 1327011271 M * m_ueberall Hi. I just started to build kernel-vserver-3.2.1 (using patch-3.2.1-vs2.3.2.5) on Mageia Linux. It's still running (so far so good), but I see the line "boolean symbol IPV6 tested for 'm'? test forced to 'n'". Does anyone know from which CONFIG option this stems from? 1327011321 M * Bertl CONFIG_IPV6=m 1327011337 M * m_ueberall No, that I corrected to CONFIG_IPV6=y in .config 1327011340 M * daniel_hozac infiniband. 1327011469 M * Bertl but you can safely ignore it 1327011513 M * m_ueberall :) 1327013127 J * fisted_ ~fisted@xdsl-87-78-215-73.netcologne.de 1327013156 Q * destructive Quit: Leaving. 1327013295 Q * fisted Ping timeout: 480 seconds 1327014498 M * Bertl off for now .. bbl 1327014502 N * Bertl Bertl_oO 1327015452 Q * ghislain Quit: Leaving. 1327015547 Q * Marbug Ping timeout: 480 seconds