1189382406 M * hparker vdlimit even 1189382512 J * click_ click@ti511110a080-2899.bb.online.no 1189382522 Q * click Ping timeout: 480 seconds 1189383356 N * Bertl_zZ Bertl 1189383364 M * Bertl ahh .. much better ... 1189383378 M * Bertl hparker: no, unless you do not have any space left 1189383413 M * Bertl hparker: i.e. the limits are set to 0,0 1189383416 M * hparker Without it set it shows 432 gig free 1189383452 M * Bertl what does vdlimit report for your context and device? 1189383470 M * hparker IDU shows 930 megs used, I set the limit to 1 gig, df showed used and available as 0 1189383512 M * Bertl vdlimit --xid /path/to/guest 1189383541 M * Bertl (upload to paste.linux-vserver.org) 1189383547 M * hparker vdlimit: vc_get_dlimit(): No such process :P 1189383561 M * Bertl well, then you didn't add a limit hash for your guest 1189383583 M * hparker limit hash? 1189383585 M * Bertl the tools (recent) will do that by default on guest restart if limits are configured 1189383622 M * Bertl how did you set the limits? 1189383687 M * hparker I've added "space_total 1048576" to /etc/vserver/guest/dlimits/dlimit and that didn't work, so then I set it with dlimit which set it, but left used and available to 0 1189383710 M * Bertl dlimit? 1189383719 M * hparker Err... vdlimit 1189383732 M * Bertl okay, how did you set it with vdlimit? (cmd line) 1189383757 M * hparker vdlimit --xid 101 --set space_total=1048576 /vservers/pcsrvc 1189383768 M * Bertl okay, and what does 1189383775 M * Bertl vdlimit --xid 101 /vservers/pcsrvc 1189383777 M * Bertl show? 1189383818 M * hparker http://rafb.net/p/gYKIvh64.html 1189383838 M * Bertl so that has at least a space limit set 1189383848 M * hparker yes 1189383866 M * hparker i have to force space_used as well? 1189383866 M * Bertl and what was the command you used above which gave your the no such process? 1189383881 M * hparker the same but didn't have the limit set :P 1189383888 M * Bertl okay :) 1189383900 M * Bertl you ahve to set all four values if you do it manually 1189383921 M * Bertl the purpose of the dlimit code is to 'track' changes to those values 1189383934 M * hparker if I use dlimits/dlimit it should do everything automatically? 1189383949 M * Bertl i.e. when you remove a file which is tagged for this xid, the *_used values change 1189383976 M * hparker makes sense 1189383976 M * Bertl yes, with moderately recent tools, a proper entry in the config causes calculation by the tools 1189384004 M * Bertl on guest startup, when the *_used values are not known 1189384011 M * hparker I'm using 30.213 1189384013 M * Bertl (they are cached over guest restart) 1189384018 M * Bertl should work for that 1189384092 M * hparker hrrmm.. stop and then start didn't pick it up 1189384116 M * Bertl what is your disk layout, and how did you configure it? 1189384126 M * hparker # cat /etc/vservers/www.pcsrvc.com/dlimits/dlimit 1189384126 M * hparker space_total 1048576 1189384152 M * Bertl http://oldwiki.linux-vserver.org/Disk+Limits 1189384158 M * hparker 500 gig raid1 with 1 partition over the whole thing 1189384178 M * hparker ext3 1189384184 M * Bertl whole thing means: the guest? the /vservers partition or the machine? 1189384202 M * hparker uhm.. guess I need tagxid 1189384217 M * hparker /vservers is all that's on that drive 1189384237 M * Bertl okay, good, you will need tagging to make the limits work 1189384254 M * Bertl otherwise the kernel cannot tell which file belongs to what context 1189384685 M * hparker On the page you linked above, why does it have dlimits/0/ but the flower page doesn't have the 0? 1189384714 M * Bertl it is an older page, but I think the directory is still true, sec 1189384787 M * Bertl http://www.nongnu.org/util-vserver/doc/conf/configuration.html 1189384796 M * Bertl yes the is a dir 1189384809 M * Bertl which contains the data 1189384843 M * hparker I'm on that page ;) My question is about the 0 in the path, is that needed now? 1189384847 M * Bertl (similar to the below 1189384867 M * Bertl no, you can name it whatever you like, but it has to be there 1189384877 M * hparker echo /vservers/test1 > /etc/vservers/test1/dlimits/0/directory 1189384886 M * Bertl for example 1189384901 M * Bertl could as well be /etc/vservers/test1/dlimits/freaking_root_dir/directory 1189384957 M * hparker Just so long as there's something between dlimits/ and the file with the limit? 1189384983 M * Bertl well, it has to be there, otherwise how would you set limits for different partitions? 1189385008 M * hparker ahhh 1189385009 M * Bertl in cases where e.g. / and /home is different for the guest 1189385025 M * hparker makes sense now ;) 1189385029 M * Bertl good :) 1189385168 M * hparker If i set one value i have to set them all? 1189385194 M * Bertl well, the values get applied all together, so the unset values will be 0 otherwise 1189385204 M * hparker oh 1189385204 M * Bertl if that is okay, you do not have to set them :) 1189385215 M * hparker 0 inodes could suck 1189385293 A * hparker applies bad math to it again 1189386278 M * hparker yay, that works ;) 1189386288 M * Bertl \o/ 1189387838 J * friendly12345 ~friendly@ppp59-167-90-160.lns1.mel6.internode.on.net 1189387981 J * coderanger_ ~coderange@kantrn.stu.rpi.edu 1189387991 M * Bertl wb coderanger_! 1189388175 Q * coderanger Ping timeout: 480 seconds 1189390042 P * friendly12345 1189391313 J * virtuoso_ ~s0t0na@pppoe-165.2.110.89-adsl.spbnit.ru 1189391717 Q * virtuoso Ping timeout: 480 seconds 1189395460 J * Aiken_ ~james@ppp121-45-250-174.lns2.bne4.internode.on.net 1189395775 Q * Aiken Ping timeout: 480 seconds 1189396835 J * coderanger__ ~coderange@c-65-96-210-168.hsd1.ma.comcast.net 1189397583 Q * zLinux Remote host closed the connection 1189398677 J * quasisane ~sanep@c-76-118-191-64.hsd1.nh.comcast.net 1189400365 J * balbir ~balbir@59.145.136.1 1189404211 M * Bertl okay, off to bed for today ... have a good one everyone! 1189404217 N * Bertl Bertl_zZ 1189404994 J * dna ~dna@237-193-dsl.kielnet.net 1189405145 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw 1189405858 J * DavidS ~david@vpn.uni-ak.ac.at 1189406218 Q * arachnist Quit: brb/bbl 1189408776 J * meandtheshell ~markus@85.127.111.224 1189409752 J * sharkjaw ~gab@158.36.44.106 1189410648 J * friendly12345 ~friendly@ppp59-167-90-160.lns1.mel6.internode.on.net 1189410874 N * phedny wilmer 1189411012 N * wilmer phedny 1189411092 J * arachnist ~arachnist@088156189068.who.vectranet.pl 1189411312 J * ema ~ema@fw.galliera.it 1189411463 Q * FireEgl cation.oftc.net venus.oftc.net 1189411463 Q * nou cation.oftc.net venus.oftc.net 1189411787 J * nou Chaton@causse.larzac.fr.eu.org 1189412453 J * morten ~morten@213-133-125-57.clients.your-server.de 1189412456 M * morten hi guys 1189412637 M * morten where can i find a good documentation about providing an own lo-interface to a guest? 1189412735 J * FireEgl FireEgl@4.0.0.0.1.0.0.0.c.d.4.8.0.c.5.0.1.0.0.2.ip6.arpa 1189412873 Q * morten Quit: [BX] Tabardation - the inability to master use of the key. See: retardation; Headcase. 1189413445 J * morten ~morten@213-133-125-57.clients.your-server.de 1189413488 M * daniel_hozac morten: just run 2.3... 1189413543 M * morten daniel_hozac: i do ;-) Linux www2 2.6.22.6-vs2.3.0.18 1189413553 M * daniel_hozac so, there you go. 1189413572 M * morten but the guests still share the loopback with the host... 1189413580 M * daniel_hozac what makes you think that? 1189413608 M * igraltist hi 1189413611 M * igraltist i get this 1189413612 M * igraltist ipc/shm.c: In function `newseg': 1189413612 M * igraltist ipc/shm.c:530: Fehler: »vxi« nicht deklariert 1189413618 M * morten debootstraped an debian.. exim can't listen on port 25 ... when i telnet to localhost 25 on the guest, i get the postfix from the host 1189413632 M * igraltist is something missing in the decleration#? 1189413635 M * daniel_hozac igraltist: so you didn't patch it right... 1189413650 M * daniel_hozac morten: and you did enable the auto_lback option, right? 1189413655 M * morten igraltist: did you really take an vanilla that you patched? 1189413679 M * igraltist no 1189413711 M * morten daniel_hozac: nope... where to do that? i was searching all docs and mans for this ... :-) 1189413724 M * morten daniel_hozac : but i enabled the support in the kernel 1189413783 M * igraltist struct vx_info *vxi = lookup_vx_info(shp->shm_perm.xid); 1189413788 M * igraltist this is defined 1189413826 M * morten igraltist: patch against a clean vanilla kernel the patch was made for and it works 1189413832 M * morten -st 1189413833 M * daniel_hozac morten: so grep AUTO_LBACK .config returns =y, yes? 1189413841 M * daniel_hozac morten: if so, you already have all that's needed... 1189413848 M * igraltist morten, i know but this not help 1189413853 M * morten for cheesy... 1189413854 M * morten root@www2:/usr/src/linux# grep AUTO_LBACK .config 1189413854 M * morten CONFIG_VSERVER_AUTO_LBACK=y 1189413856 M * igraltist how i get then rsbac inside 1189413931 M * morten daniel_hozac: hm, but as we see, the guest still shares the loopback with the host :-( 1189414061 M * daniel_hozac that's a bug :) 1189414175 M * morten hehe, you think so? :-) 1189414250 M * daniel_hozac so, what does cat /proc/virtnet//{info,status} output? (use paste.linux-vserver.org) 1189414256 M * igraltist now i try a vanilla kernel and patch it with vserver and then i can compare what is missing 1189414374 M * morten daniel_hozac : http://paste.linux-vserver.org/6542 1189414392 M * morten hmm, there is an automatic assigned ip in the info... 1189414397 M * morten lo ip 1189414533 M * morten daniel_hozac: i've updated it with an ifconfig lo output from the guest/vserver 1189414562 M * DavidS morten: use "ip a s" (from iproute2) not ifconfig 1189414574 M * DavidS the latter tends to hide to much stuff 1189414662 M * morten DavidS: as you want.. updated the "paste" 1189414978 M * morten daniel_hozac : are you still there? :-) 1189415087 Q * coderanger__ Quit: coderanger__ 1189415180 M * daniel_hozac well, i can reproduce it locally, i'm not sure of the cause yet. 1189415223 M * morten sweet, let me know if you find somethin' ;-) 1189415432 M * morten di i have to have the real machine running chbind'? 1189415453 M * daniel_hozac no. 1189415462 M * morten ok. puuh :-) 1189415549 M * igraltist put_vx_info(vxi); it was on the wrong place :) 1189415578 M * igraltist this is very helpfull now to have a vanilla patched vserver kernel to compare 1189416440 Q * hparker Quit: *burp*.. It's broke 1189416946 Q * morten Quit: BitchX-1.1-final -- just do it. 1189417493 J * kir ~kir@swsoft-mipt-nat.sw.ru 1189417661 J * morten ~morten@213-133-125-57.clients.your-server.de 1189417662 M * morten :-( 1189417667 M * morten can't get it working.. 1189417679 M * morten (lo) 1189417757 M * morten daniel_hozac : do i have to set some nflags? 1189417775 M * morten like LBACK_REMAP or somethin'? 1189417840 M * daniel_hozac you have them set already. 1189417863 M * morten okeydoke. 1189417955 M * morten is it normal that when i do a ip a s under the guest, that it says 127.0.0.1 .. or should it say the ip from the /proc/virtnet/... of the host? 1189417977 M * daniel_hozac the guest should only see/use 127.0.0.1 1189417984 M * morten k, cool ;-) 1189417988 M * morten now it only has to work :-) 1189417997 M * morten is there anything else i can look for / test? 1189418122 M * morten i guess i have it... 1189418147 M * morten postfix on master is bind to 0.0.0.0:25, not to 127.0.0.1:25 1189418214 M * daniel_hozac it shouldn't matter. 1189418233 Q * kir Quit: Leaving 1189418270 M * morten ok, but all processes listening to 127.0.0.1:specificport on the host are not reachable from the guest and can be used 1189418318 M * morten so they have different loopbacks, but the host binds port 25 to ALL interfaces, even the virtual interface from the guest 1189418338 M * morten maybe chbinding the init of the host will help!? 1189418349 M * daniel_hozac you really don't want to do that. 1189418354 M * morten :-) 1189418359 M * morten kaykay 1189418387 M * morten if i don't need to, then i don't want to .. ;-) 1189418410 M * daniel_hozac well, i think this is a bug 1189418430 M * morten who's the maintainer of the autoloopback function`? 1189418458 M * daniel_hozac Bertl_zZ maintains the kernel. 1189418473 M * morten i see, isee 1189418568 M * daniel_hozac i think this is my fault though. 1189418811 M * morten hihi, gotcha :-p 1189418895 M * morten i'll grab sum' to eat... 1189418897 N * morten morteat 1189419954 Q * arachnist Quit: brb/bbl 1189420170 Q * balbir Ping timeout: 480 seconds 1189420770 J * arachnist ~arachnist@088156189068.who.vectranet.pl 1189422299 N * morteat morten 1189422519 J * smarthouseguy ~kevin@cpe-024-074-098-089.carolina.res.rr.com 1189422529 M * smarthouseguy howdy.. anyone up 1189422530 M * smarthouseguy ? 1189422537 M * daniel_hozac most likely. 1189422651 M * smarthouseguy gr8... i have a small issue... just put my server in a datacenter, hosting a few asterisk pbxs... so they assigned me 2 stacks of ips... a /29 and a /28 network... all using the same gateway.. the /29 network iis great, but when I try to assign an ip from the /28 network, can't ping out or in... not quite sure where to go with this, and the DC is being less than helpful 1189422865 M * daniel_hozac what are the networks and are you sure they should use the same gateway? 1189422878 M * morten daniel_hozac : do you think we'll have a fix/patch for this lo problem soon? otherwise i'll try to bind the services on the host only to it's own lo and eth address... 1189423005 Q * ema Quit: leaving 1189423014 M * smarthouseguy the 2 networks are 75.127.81.144/29 255.255.255.239 and 75.127.81.224/28 255.255.255.240 all supposedly using 75.127.81.145 as gw, but 2 seperate broadcast addresses 1189423065 M * smarthouseguy and since I'm using asterisk, I'm trying not to use iptables as not to interfere with the sip traffic 1189423070 M * daniel_hozac 75.127.81.224/28 can't reach the gw. 1189423090 M * smarthouseguy didn't think so.. tried to argue with them about this... 1189423113 M * smarthouseguy then they just say oh, well we really don't support debian as an OS bla bla bla 1189423134 M * daniel_hozac morten: well, there aren't really any clean fixes (AFAICT right now), i'll have to discuss it with Bertl_zZ... 1189423204 M * smarthouseguy this was their response "It does not work like that, you have the main block /29 subnet and then you add the additional IPs block assigned. Its all on the same network, if you are not sure what you are doing ask for help." 1189423247 M * smarthouseguy and "Does not matter how many ips you need all servers come with /29 subnet and then what ever extra ip blocks you need after that, you cannot remove the main block of ips you just need to add the second stack, I can pass this onto a admin but it will take upto 24 hours to get this resolved, their will also be a charge for this issue. 1189423300 M * smarthouseguy I asked them pre-sales and pre-install if all the ips would be in sequence and on the same subnet, of course they said yes... 1189423364 M * Wonka what happens if you just fire up some tcpdump listening for one of those extra IPs and ping that IP from external? 1189423392 M * daniel_hozac (after assigning it to the host) 1189423460 M * smarthouseguy last time I tried to assign one of the new ips to the host, it went down and took them 15 hours to respond... they then proceeded to tell me they were going to charge me $45/ 15mins to get it back online, so I'm a little nervous about making to many changes on the host at this point 1189423484 M * smarthouseguy especially to the IP.. lol 1189423502 M * daniel_hozac you're not removing the current one, just adding another one. 1189423514 M * smarthouseguy i'm simply hoping I can use the new stack in VEs 1189423520 M * smarthouseguy like eth0:1 1189423544 M * daniel_hozac just do ip addr add /28 dev ethX 1189423710 M * sid3windr how they say it it can work 1189423724 M * sid3windr but every server which is in that /28 will also need an ip in the /29 1189423774 M * smarthouseguy hmmm... did ip addr add 75.127.81.225/28 dev eth0 1189423784 M * smarthouseguy but it doesn't show up... 1189423887 M * daniel_hozac hmm? 1189423889 M * smarthouseguy eth0 Link encap:Ethernet HWaddr 00:1A:A0:25:3A:F6 1189423889 M * smarthouseguy inet addr:75.127.81.147 Bcast:75.127.81.151 Mask:255.255.255.248 1189423889 M * smarthouseguy UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 1189423889 M * smarthouseguy RX packets:503402 errors:0 dropped:0 overruns:0 frame:0 1189423889 M * smarthouseguy TX packets:258131 errors:0 dropped:0 overruns:0 carrier:0 1189423891 M * smarthouseguy collisions:0 txqueuelen:1000 1189423891 M * smarthouseguy RX bytes:163617513 (156.0 MiB) TX bytes:31845121 (30.3 MiB) 1189423893 M * smarthouseguy Interrupt:177 1189423896 M * daniel_hozac don't use ifconfig. 1189423901 M * daniel_hozac it has been deprecated for years. 1189423904 M * daniel_hozac use ip addr ls 1189423955 M * smarthouseguy lol... learn more new stuff everyday 1189423966 M * smarthouseguy vs1:~# ip addr ls 1189423966 M * smarthouseguy 1: lo: mtu 16436 qdisc noqueue 1189423966 M * smarthouseguy link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 1189423966 M * smarthouseguy inet 127.0.0.1/8 scope host lo 1189423966 M * smarthouseguy 2: eth0: mtu 1500 qdisc pfifo_fast qlen 1000 1189423967 M * smarthouseguy link/ether 00:1a:a0:25:3a:f6 brd ff:ff:ff:ff:ff:ff 1189423967 M * smarthouseguy inet 75.127.81.147/29 brd 75.127.81.151 scope global eth0 1189423969 M * smarthouseguy inet 75.127.81.228/28 brd 75.127.81.239 scope global eth0 1189423969 M * smarthouseguy inet 75.127.81.149/29 brd 75.127.81.151 scope global secondary eth0 1189423971 M * smarthouseguy inet 75.127.81.148/29 brd 75.127.81.151 scope global secondary eth0 1189423971 M * smarthouseguy inet 75.127.81.236/28 brd 75.127.81.239 scope global secondary eth0 1189423973 M * smarthouseguy inet 75.127.81.150/29 brd 75.127.81.151 scope global secondary eth0 1189423973 M * smarthouseguy inet 75.127.81.225/28 scope global secondary eth0 1189423994 M * morten smarthouseguy : use http://paste.linux-vserver.org ;-) 1189424022 M * smarthouseguy can't ping the 225 1189424023 M * smarthouseguy k 1189424127 M * smarthouseguy I'm really at the point of waiting out this month, since I know I won't get a refund, and making a trip to atlanta to move this to a DC that will assign IPs the way the cust needs... and not argue about it 1189424230 M * daniel_hozac try 226. 1189424304 M * epicbjorn reminds me... i really should learn to use the iproute2 package properly 1189424331 M * smarthouseguy pasted 1189424334 M * smarthouseguy nothing 1189424351 M * smarthouseguy lol.. yeah.. same here... lots to learn still 1189424398 M * smarthouseguy only been using linux for a few years... love it so far, but have a long way to go.... started with slackware, just moved to debian for this.... 1189424700 M * smarthouseguy BTW... vserver rocks... tried several other virtualization solutions b4 this... vserver is the only one I could get asterisk to play nice with... have had no issues until this ip mess awesome project ;-) 1189424822 Q * arachnist Ping timeout: 480 seconds 1189425278 J * arachnist arachnist@088156189068.who.vectranet.pl 1189425304 J * coderanger__ ~coderange@c-65-96-210-168.hsd1.ma.comcast.net 1189425312 Q * arachnist Read error: Connection reset by peer 1189425324 J * arachnist arachnist@088156189068.who.vectranet.pl 1189425377 J * roym ~user@adsl-065-006-164-142.sip.mia.bellsouth.net 1189426650 M * smarthouseguy ok... so I just went through the iproute2 documentation, no fresh ideas here... daniel_hozac ? 1189426732 Q * coderanger__ Quit: coderanger__ 1189426926 M * Medivh smarthouseguy, i would find another ISP ;) 1189426955 M * Medivh "the 2 networks are 75.127.81.144/29 255.255.255.239 and 75.127.81.224/28 255.255.255.240" <-- the first netmask is wrong btw, it should be 255.255.255.248 for a /29 1189426956 M * Hollow hm, i have different subnets using the same gateway too ... 1189427011 M * daniel_hozac sure, it works fine if they're on the same physical network, and you have the necessary route. 1189427021 M * Medivh yeah, works if the ISP configures both gateway ips on the same VE interface on their router/switch 1189427044 M * Hollow well, i don't have any routes configured except the default gw, but yeah, they are on the same physical network of course 1189427053 M * smarthouseguy oops.. that was my mistake... it's 248 1189427072 M * daniel_hozac your host has addresses on both nets though, thus creating the route. 1189427083 M * daniel_hozac (i presume...) 1189427100 M * smarthouseguy well pass the word on to beware of NorthStorm Media Networks... 1189427122 M * smarthouseguy I specifically requested my ips be on the same subnet and in sequence... 1189427122 M * Hollow yep .. 1189427147 M * smarthouseguy ne DC suggestions in the Atlanta area?? 1189427159 M * Medivh smarthouseguy, what's your gateway ip on the /29? 1189427165 M * Hollow ps -yl rocks, btw :) 1189427214 M * smarthouseguy 75.127.81.145 1189427259 M * Medivh ok, so sounds like they use the first ip on the subnet as the gateway, as usual... which would mean that .225 is their gateway and you cannot use it... try removing it and replacing it with .226, see if that works 1189427273 M * smarthouseguy when they sent the new ips.. it had no gw addy... *** pasting now *** 1189427361 Q * rorem- Read error: Connection reset by peer 1189427439 M * smarthouseguy *** pasted more at top *** 1189427446 J * rorem- ~roremtank@bzq-219-46-202.isdn.bezeqint.net 1189427469 M * Medivh mhm, can't reach .225 from the outside, seems they don't have the /28 configured on their side at all 1189427498 M * smarthouseguy i may not know everything there is to know about networking, but as soon as I saw this stack of ips was on a different subnet i knew I was going to have issues 1189427521 M * smarthouseguy i told them that.. they said that my server is the gw to that network... 1189427586 M * smarthouseguy *** pasted more *** 1189427644 M * smarthouseguy if I can't even ping the gateway... ????? makes no sense to me 1189427668 J * Julius ~julius@p57B27BB4.dip.t-dialin.net 1189427691 M * Medivh yuk, they just put a static route on the first ip of the /29? what strange setup is that... 1189427774 M * daniel_hozac seems more likely they don't put a gateway at all. 1189427843 M * Medivh yeah, from what i get from the mails pasted, they set up the /29 as one usually does, and then added a static route for 75.127.81.224/28 to 75.127.81.146 1189427873 M * smarthouseguy lol... .146 is a baseboard management controller 1189427886 M * smarthouseguy could this be why it's not working??? 1189427902 M * smarthouseguy .147 is the 1st ip assigned to the server 1189427910 M * Medivh could definately be it ;) but it's crappy setup on your ISPs side, from my point of view at least 1189427928 M * Medivh ask them to change the static route for the /28 to route to .147 ... ;) 1189427962 M * smarthouseguy think that would get it all live until i can get this moved on 10-1 ? 1189428013 M * Medivh best would be if they just put the .225 up on their layer 3 equipment, that'd be the best setup... but it _should_ work if they put in the route, i guess... 1189428071 M * smarthouseguy no matter what I'm moving it.. they are terrible.. couldn't even follow instructions.. "please plug into gb1" took 9 hours for that issue and they made me have to change my root pass 1189428100 M * Medivh lol 1189428119 M * Medivh sounds like no matter where you go, it can only get better 1189428181 M * smarthouseguy then when they issued the new stack and I change my ips thinking it was a new stack like I had requested it took 15 hours to be live again.. I could have driven to Atlanta, fixed it, drove home, cooked dinner, washed clotes, the car and cleaned half the house in that amount of time... and I'm 3 states away.. lol 1189428199 M * Borg- yeah.. sad to here.. Im really sad for all those good ppl who lives in USA ;) 1189428211 M * Borg- s/here/hear/ 1189428232 M * smarthouseguy where r u Borg? 1189428255 M * Borg- does that matter? 1189428267 M * smarthouseguy not really... just curious 1189428283 M * smarthouseguy not the USA i assume 1189428283 M * Borg- I live in crappy country too.. 1189428289 M * Borg- but at least I do hosting somewhere else 1189428295 M * smarthouseguy aren't they all in one way or another 1189428310 M * Borg- indeed.. but I mainly focus in IT 1189428356 M * smarthouseguy i would, but atlanta is the inbetween point for my home and office... direct connection from both ISPs 1 hop to the server out of WAN ISP, so good for my voip calls... 1189428403 M * Borg- seems they know that ;) 1189428442 M * smarthouseguy just found the wrong one 1st.. there are TONS of DCs in atlanta.. so it's not like it will be hard to find another option.. and will blog about them until my fingers fall off.... 1189428496 M * smarthouseguy they are only a "reseller" for GNAX so when I go pick up my server I may have a lil meeting with the GNAX peeps 1189428537 M * Medivh i know Level(3) has a data center in atlanta, too... 1189428546 M * smarthouseguy someone tried to sell me more GNAX and I said, not on your life 1189428556 M * smarthouseguy yup.. I was hoping to get into 56 Marrietta 1189428573 J * pmenier ~pmenier@ACaen-152-1-93-62.w83-115.abo.wanadoo.fr 1189428574 M * smarthouseguy but they don't wanna talk to a 1u server like me... they only want full racks... 1189428597 M * Borg- resellers? 1189428622 M * smarthouseguy at least that's what I'm assuming.... since 56 Marrietta doesn't contact me back about my 1u 1189428622 N * pmenier pmenier_off 1189428629 Q * Aiken_ Quit: Leaving 1189428649 M * smarthouseguy yup... NorthStorm Media Networks "An Authorized GNAX reseller" 1189428713 Q * Julius Remote host closed the connection 1189429051 M * sid3windr unauthorized reseller would be funny 1189429059 J * lilalinux ~plasma@dslb-084-058-222-004.pools.arcor-ip.net 1189429130 J * Julius ~julius@p57B27BB4.dip.t-dialin.net 1189429198 M * smarthouseguy i think it would serve them right... 1189431761 J * fatgoose ~samuel@206-248-157-177.dsl.teksavvy.com 1189432484 Q * lilalinux Remote host closed the connection 1189433565 P * friendly12345 1189433664 Q * morten Quit: [BX] 2000: year of the BitchX 1189433697 J * morten ~morten@213-133-125-57.clients.your-server.de 1189435654 Q * vanboven Read error: Connection reset by peer 1189436354 J * coderanger__ ~coderange@c-65-96-210-168.hsd1.ma.comcast.net 1189437025 M * Borg- sid3windr: hows your network? :) 1189437120 M * smarthouseguy hey everyone.. thanks for all your help... noc ended up assigning me a stack of ips I can work with... everyone here rocks.. great project, keep up all the good work.... heading out now.. kind regards 1189437149 Q * smarthouseguy 1189437417 M * sid3windr Borg-: working, I think :] 1189437790 M * Borg- :) 1189437795 M * Borg- okey heading to home.. cu 1189437942 Q * rorem- Remote host closed the connection 1189438040 J * rorem- ~roremtank@bzq-219-46-202.isdn.bezeqint.net 1189438099 Q * rorem- Remote host closed the connection 1189438139 J * fatgoose_ ~samuel@76-10-154-18.dsl.teksavvy.com 1189438166 J * rorem- ~roremtank@bzq-219-46-202.isdn.bezeqint.net 1189438337 J * elesouef ~elesouef@home.lorinand.org 1189438470 Q * fatgoose Ping timeout: 480 seconds 1189439057 N * ensc Guest3184 1189439066 J * ensc ~irc-ensc@p54B4E7B2.dip.t-dialin.net 1189439176 Q * Guest3184 Ping timeout: 480 seconds 1189439312 M * tam Has anyone upgraded a centos guest from centos4 to centos5? 1189439974 M * daniel_hozac it's not different from upgrading a real machine. 1189439985 M * daniel_hozac replace centos-release, yum update. 1189440092 M * tam I'm about to try :) 1189440270 J * gerrit ~gerrit@c-67-169-199-103.hsd1.or.comcast.net 1189440770 N * pmenier_off pmenier 1189440852 J * bonbons ~bonbons@2001:5c0:85e2:0:20b:5dff:fec7:6b33 1189441087 M * tam I suspect this would be easier if my guest used internal package management 1189441107 M * tam but just installing the 5.0 centos-release packages didn't get me there 1189441130 M * daniel_hozac how did you install them? 1189441133 M * tam vrpm 1189441141 M * tam vrpm -ivh --force 1189441153 M * daniel_hozac never use --force... 1189441160 M * tam hrm. 1189441171 M * daniel_hozac anyway, with external package management, you'll need to point vyum at the centos5 repos. 1189441241 M * daniel_hozac something like ln -sf /usr/lib*/util-vserver/distributions/centos5/yum.repos.d /vservers/.pkg//yum/etc/yum.repos.d 1189441262 M * tam Ahah! 1189441267 M * tam I was just finding that when you said that 1189441268 M * tam Thank you! 1189441286 M * daniel_hozac you're welcome! 1189441525 Q * coderanger__ Quit: coderanger__ 1189441700 J * jmcaricand ~jmcarican@d83-179-196-32.cust.tele2.fr 1189442069 N * Bertl_zZ Bertl 1189442073 M * Bertl morning folks! 1189442079 M * daniel_hozac morning Bertl! 1189442355 M * Bertl hey daniel_hozac! how's going? 1189442373 M * daniel_hozac quite well, and you? 1189442433 M * Bertl still a little bit stressed, but that should be over soon 1189442576 M * Bertl did you read the (v)hashify posting (ML) regarding unification of existing links? 1189442611 M * Bertl isn't that already avoided by vhashify? 1189442624 M * daniel_hozac no, i don't think so. 1189442671 M * Bertl okay, I think it would make sense to avoid unification of files with nlink > 1 and no unification flags set, no? 1189442697 M * daniel_hozac mostly, yeah. 1189442698 M * Bertl at some point, it would also make sense to break a CoW link on guest side hard linking 1189442764 M * Bertl just an idea ... let me know if you see problems with that 1189443412 J * yarihm ~yarihm@whitehead2.nine.ch 1189443471 M * daniel_hozac btw, morten ran into an "interesting" issue... connecting to an auto_lbacked guest's 127.0.0.1 connects to the host, if the host has bound 0.0.0.0. 1189443505 Q * roym Ping timeout: 480 seconds 1189443593 M * daniel_hozac AFAICT, the bind call succeeds on the guest, but it's just not connectable. 1189443613 J * _nkukard_ ~nkukard@dsl-241-41-162.telkomadsl.co.za 1189443751 M * Bertl hmm, interesting ... 1189443794 M * daniel_hozac IMHO, we want to allow the host to bind to 0.0.0.0, and let guest's lback override this. 1189443882 M * _nkukard_ daniel_hozac, can u put me in touch with the guy that did that yum patch? 1189443890 M * daniel_hozac ensc: 1189443917 M * _nkukard_ thanks a mil 1189443938 M * daniel_hozac how's the merging going? 1189443963 M * _nkukard_ heh, rejected ... need to speak to ensc, I'll get the functionality added, just not in the way the patch does it ;) 1189443985 M * daniel_hozac due to chrootfs: and hostfs? 1189444002 M * _nkukard_ nah ... every single hunk :( 1189444010 M * _nkukard_ don't worry though, I'll work on it 1189444023 M * daniel_hozac is this on yum-devel? 1189444059 M * daniel_hozac i'd like the see the rationale this time :) 1189444120 M * _nkukard_ I'd be happier if you could more specifically identify the problems you're having with the current installroot capability 1189444134 M * daniel_hozac it's insecure. 1189444146 M * _nkukard_ it's just not something I want to accept 1189444147 M * _nkukard_ so yea 1189444148 M * daniel_hozac it'll use files from inside the chroot. 1189444154 M * _nkukard_ let me tackle it ;) 1189444182 M * _nkukard_ if u or ensc could detail the problem, I'll raise it with him he can tell me "his way" to fix it, I'll fix it and revert back to u 2 1189444215 M * daniel_hozac it's simple: yum --installroot= will use configuration from if it exists. 1189444217 M * morten Bertl: mornin ;-) have you followed the discussion about the autoloopback problem? 1189444255 M * _nkukard_ daniel_hozac, and we wan ti to use whatever config we want? 1189444258 M * _nkukard_ for instance 1189444261 M * daniel_hozac and it's not just configuration, it's cache, logfiles, lockfile, etc. 1189444262 M * Bertl morten: no, but daniel gave me a short version 1189444266 M * _nkukard_ there we go 1189444282 M * daniel_hozac we want the files to live on the host, no matter what. 1189444292 M * daniel_hozac anything else would be a security problem. 1189444341 M * morten Bertl: ok cool, so you guys don't need anything more? if so, let me know. 1189444388 M * _nkukard_ daniel_hozac, gotcha .. 1189444420 M * Bertl morten: well, the problem might easily boil down to 'don't do that' (binding anything to 0.0.0.0 on the host) 1189444453 M * daniel_hozac at the same time, i think the lback address should override it. 1189444457 M * Bertl morten: I think the current behaviour is not quite unexpected 1189444477 M * Bertl it is the same you get with any other ip right now 1189444628 M * Bertl it might become more and more interesting to put everything on the host into a separate isolation area 1189444655 M * morten Bertl: courier is not able to listen to multiple ips, proftpd not in the version i use, etc .... 1189444673 M * morten just one ip or all... :-( 1189444674 M * Bertl morten: to make it short, don't run them on the host 1189444686 M * morten hehe, k.. 1189444723 M * Bertl put a chbind/ncontext wrapper around them and everything should be fine 1189444740 M * morten k 1189444757 M * morten i'll try that later ... 1189444758 M * morten thx 1189444767 M * Bertl daniel_hozac: I don't think we can easily handle that case with some override 1189444775 M * daniel_hozac easily, no. 1189444778 M * daniel_hozac ugly hacks, yes :) 1189444786 M * Bertl daniel_hozac: it would require reordering the socket lists 1189444816 M * daniel_hozac well, not necessarily. 1189444830 M * daniel_hozac we could just not match the host's 0.0.0.0 for connects to 127.0.0.1 if LBACK_REMAP- 1189444858 M * Bertl hmm, and for udp? 1189444870 M * elesouef Hi all, I would like to know if it is possible for a vserver to start its ethernet iface with dhcp ? 1189444873 M * daniel_hozac same. not match the socket. 1189444894 M * Bertl elesouef: for the guest no, for the host, yes 1189444941 M * Bertl daniel_hozac: okay, sounds doable, if you feel like drafting up a proof of concept code ... go ahead 1189444961 M * elesouef Bertl: ok. My question was for the guest. As my host is a dhcp client itself which has to to switch from a lan to another. DHCP for the guest would have help me :) 1189445000 M * Bertl elesouef: guests are at the IP level, dhcp is below that, further, the guest is for security reasons not allowed to mess with the ip 1189445030 M * Bertl elesouef: but as I said, the host _can_ request a dhcp ip _for_ the guest (and assign it) 1189445080 M * Bertl daniel_hozac: regarding dhcp-ing guest ips, is that planned in the future? would an initialize script which does the proper config stuff suffice right now? 1189445098 M * daniel_hozac yeah, it's planned. 1189445113 M * daniel_hozac IMHO it should be completely decoupled from the guest. 1189445124 M * daniel_hozac i.e. the dhclient-script should do everything necessary 1189445138 M * daniel_hozac (or the equivalent for other DHCP-clients) 1189445154 M * Bertl i.c. but you have to specify a guest id or so? 1189445170 M * daniel_hozac right, just make the script use naddress to set the IP-addresses. 1189445325 M * elesouef Bertl: ok, thanks for your infos 1189445441 M * Bertl you're welcome! 1189445786 M * Bertl daniel_hozac: I think we still have some information leaks about connections from /proc (about host and other guests) and we should double check that, especially in regard to iptable stuff 1189445890 M * daniel_hozac hmm? 1189445907 M * daniel_hozac but yes, i'm rather certain we are missing at least interface hiding in /proc. 1189445910 M * Bertl let me forward you an email ... 1189446556 J * zLinux ~zLinux@88.213.24.240 1189446635 M * daniel_hozac Bertl: ah, i see. 1189446638 J * marcfiu ~mef@aegis.CS.Princeton.EDU 1189446641 M * marcfiu hi 1189446651 M * daniel_hozac hello 1189446667 M * marcfiu question: does the 2.2.x based scheduler work well with SMT? 1189446670 Q * yarihm Quit: This computer has gone to sleep 1189446695 M * Bertl marcfiu: I would say so, why not? 1189446699 M * marcfiu That is, will it fairly schedule things? I remember this being somewhat of an issue years ago for VMware. 1189446704 M * marcfiu Bertl: great 1189446729 M * Bertl I mean, we do not change the Linux Scheduler itself, so it will do what a normal Linux system does on SMT 1189446736 M * marcfiu right.. 1189446753 M * marcfiu and I was just curious whether it did the right thing these days. :) 1189446757 M * Bertl we will have limits for each VCPU though (and cpu sets should work there too) 1189446797 M * marcfiu ok... I'm going to stay away from SMT for now. 1189446824 M * Bertl hmm, in what way? :) 1189446858 M * marcfiu by not enabling CONFIG_SCHED_SMT 1189446870 M * Bertl hum :) 1189446887 M * Bertl that will only give you less support for HT machines :) 1189446925 M * Bertl i.e. the scheduler will treat siblings like real cpus (without that) 1189446948 M * marcfiu hmmm... what I want is to avoid using the HT siblings altogether. 1189446967 M * Bertl so you want to disable HT in the bios then 1189446987 M * marcfiu crap... I can't assume that the sites using PlanetLab will do that correctly. :( 1189447005 M * Bertl but actually, I don't see a good reason for doing so 1189447027 M * Bertl Linux uses HT properly since more than a year now 1189447027 M * marcfiu ok 1189447040 M * marcfiu I have not looked into it for a while. 1189447080 M * marcfiu Well, I suppose we don't have SMP systems with HT support. 1189447149 M * marcfiu So I am guessing that the VServer scheduler will blindly deduct tokens from processes running on siblings, even if one sibling completely dominated the other. 1189447223 M * marcfiu I think the trick that VMware did was that when both siblings were used during a quantum, then they were each charged 50%. 1189447277 M * Bertl I still don't see a need for handling that special 1189447318 M * Bertl (actually I do not see any need for handline SMT special for the hard TB scheduler) 1189447319 M * marcfiu for accuracy of hard scheduling, fairshare, and reservations. 1189447338 M * marcfiu but that's just a conjecture. 1189447342 M * marcfiu I have no hard proof. 1189447397 M * Bertl how would, on a single HT CPU system 'unfair' scheduling look like? 1189447406 M * marcfiu you are charging a process the same token in both scenarios when the HT sibling is idle and when the HT sibling is cranking away. 1189447419 M * Bertl i.e. how would a guest exploit the fact that those are HT siblings and not real CPUs? 1189447434 J * coderanger__ ~coderange@wireless-19-94.media.mit.edu 1189447440 M * Bertl no, we are charging per vpcu 1189447447 J * |pmenier| ~pmenier@ACaen-152-1-103-46.w83-115.abo.wanadoo.fr 1189447459 M * Bertl so a HT system with 2 siblings will have setups for both virtual cpus 1189447461 M * marcfiu is a HT == vcpu? 1189447493 M * Bertl HT hyper threading, vcpu is the virtual cpu (can be sibling or real) 1189447511 M * marcfiu right... you are treating a sibling and a real cpu equivalently. 1189447544 M * Bertl from the accounting and TB PoV, yes 1189447569 M * Bertl but we do not change how the kernel schedules tasks across the siblings 1189447593 M * marcfiu right... and you use that accounting to decide whether a vserver is within its hard limit / reservation. 1189447602 M * Bertl correct 1189447640 M * marcfiu but when you run on a core where both siblings are busy, then you do not get the same amount of work done as on a real CPU. 1189447659 M * marcfiu In fact, on most HT systems you cannot even count on getting 50% of the CPU. 1189447681 M * marcfiu It is entirely possible to issue instruction sequences that will rob the other sibling from getting the CPU. 1189447687 M * Bertl well, that are very old systems I would say, but yeah, those CPUs will get less done 1189447701 M * Bertl but that doesn't hurt or affect the accounting 1189447729 M * marcfiu Any way, I don't want to waste you time.... this was an issue that VMware encountered when they tried to fairly run two VMs on an HT based core. 1189447774 M * marcfiu What they observed was that depending on load they were not able to meet the CPU reservations assigned to the different processors. 1189447805 M * marcfiu They then spent a bunch of time trying to get more accurate CPU hardware accounting out of the Intel processors, which turned out to be very hard (at least back then). 1189447819 M * Bertl tests and numbers for both SMT and CMP would be interesting , but I get the feeling the way we modify the scheduler should give proper results 1189447851 Q * pmenier Ping timeout: 480 seconds 1189447851 M * marcfiu tests and numbers are the only way to prove my theory. 1189447884 M * Bertl I expect something like: 1189447901 M * Bertl guest A running, B stopped (with a 50/50 share on SMT) 1189447919 M * Bertl gives better performance for A than when guest A and B both running 1189447938 M * Bertl but I'm pretty sure, the performance of A and B will be identical 1189448142 M * Bertl but yeah, if you want 'guaranteed' cpu performance, then disabling HT (on intel) is a good idea 1189448175 M * Bertl but IMHO HT is already fading out, and gets replaced by dual and quad core cpus 1189448238 M * marcfiu HT is supposed to fade out, but I get the sense that when Intel moves towards those 80 core units that we'll see HT return. 1189448246 M * marcfiu Any way, that's a future problem. 1189448272 M * marcfiu I do want "guaranteed" cpu performance, and HT doesn't work well in those scenarios. 1189448289 M * marcfiu I'm also not sure that fair-share does the right thing when running on an HT based system. 1189448422 M * marcfiu What Vmware did was modify their accounting such that when two VMs run on a single CORE with two siblings at the same time, then they'd only account 50% of the time (tokens) to those VMs. I think there is a direct analogy to Vserver, where one should only charge 50% of the tokens when the process ran on a core where the other sibling was also busy. 1189448469 M * Bertl well, it's hard to 'define' busy 1189448488 M * Bertl what if the other sibling is 60% busy? 1189448501 Q * ktwilight Read error: Connection reset by peer 1189448576 J * ktwilight ~ktwilight@95.221-66-87.adsl-static.isp.belgacom.be 1189448598 M * marcfiu yeah.. they tried to figure that out by looking into using Intel's hardware accounting support. However, this support turned out to be fairly inaccurate at the time AND required a fair number of cycles to access in the first place. So what they did was just check whether anything ran on a sibling within a time quantum, and if so charge each 50% otherwise charge 100%. 1189448655 M * Bertl I think that would make sense on a dual guest setup, but it doesn't really make sense on a 100 guest setup :) 1189448666 M * Bertl (probably nothing VMware had to think ybout :) 1189448673 M * Bertl *about 1189448692 M * marcfiu I don't think it matters whether there is two guests or 100 guests. 1189448716 M * marcfiu well, you should know better. 1189448717 M * Bertl well, for the 100 guest case, you can simply _assume_ that the sibling is busy :) 1189448775 M * marcfiu and there you have it... case closed. :) 1189448787 M * marcfiu I'm enabling SMT. :) 1189449210 J * comfrey ~comfrey@host-64-179-9-253.buf.choiceone.net 1189449794 Q * |pmenier| Quit: KVIrc 3.2.0 'Realia' 1189449999 J * yarihm ~yarihm@84-75-130-73.dclient.hispeed.ch 1189450353 J * callahan ~chatzilla@host-181-96.dsl-sea.seanet.com 1189450365 M * Bertl wb comfrey! yarihm! callahan! 1189450797 Q * coderanger__ Remote host closed the connection 1189450829 J * coderanger__ ~coderange@1cc-dhcp-106.media.mit.edu 1189450853 M * yarihm re Bertl 1189451207 Q * callahan charon.oftc.net synthon.oftc.net 1189451207 Q * AndrewLee charon.oftc.net synthon.oftc.net 1189451207 Q * _nkukard_ charon.oftc.net synthon.oftc.net 1189451207 Q * tam charon.oftc.net synthon.oftc.net 1189451207 Q * hardwire charon.oftc.net synthon.oftc.net 1189451207 Q * Hollow charon.oftc.net synthon.oftc.net 1189451207 Q * Supaplex charon.oftc.net synthon.oftc.net 1189451207 Q * FloodServ charon.oftc.net synthon.oftc.net 1189451207 Q * comfrey charon.oftc.net synthon.oftc.net 1189451207 Q * rorem- charon.oftc.net synthon.oftc.net 1189451207 Q * coderanger_ charon.oftc.net synthon.oftc.net 1189451207 Q * Johnnie charon.oftc.net synthon.oftc.net 1189451207 Q * hallyn charon.oftc.net synthon.oftc.net 1189451207 Q * djbclark charon.oftc.net synthon.oftc.net 1189451207 Q * zLinux charon.oftc.net synthon.oftc.net 1189451207 Q * quasisane charon.oftc.net synthon.oftc.net 1189451207 Q * wenchien charon.oftc.net synthon.oftc.net 1189451207 Q * er charon.oftc.net synthon.oftc.net 1189451207 Q * neuralis charon.oftc.net synthon.oftc.net 1189451207 Q * mountie charon.oftc.net synthon.oftc.net 1189451207 Q * dilinger charon.oftc.net synthon.oftc.net 1189451207 Q * mattzerah charon.oftc.net synthon.oftc.net 1189451207 Q * puck charon.oftc.net synthon.oftc.net 1189451207 Q * coderanger__ charon.oftc.net synthon.oftc.net 1189451207 Q * fatgoose_ charon.oftc.net synthon.oftc.net 1189451207 Q * marcfiu charon.oftc.net synthon.oftc.net 1189451207 Q * mstrobert charon.oftc.net synthon.oftc.net 1189451207 Q * mugwump charon.oftc.net synthon.oftc.net 1189451207 Q * bored2sleep charon.oftc.net synthon.oftc.net 1189451207 Q * MooingLemur charon.oftc.net synthon.oftc.net 1189451207 Q * micah charon.oftc.net synthon.oftc.net 1189451254 J * comfrey ~comfrey@host-64-179-9-253.buf.choiceone.net 1189451254 J * rorem- ~roremtank@bzq-219-46-202.isdn.bezeqint.net 1189451254 J * coderanger_ ~coderange@kantrn.stu.rpi.edu 1189451254 J * Johnnie ~jdlewis@c-67-163-142-234.hsd1.ct.comcast.net 1189451254 J * hallyn ~xa@adsl-75-0-147-61.dsl.chcgil.sbcglobal.net 1189451254 J * djbclark dclark@opensysadmin.com 1189451254 J * coderanger__ ~coderange@1cc-dhcp-106.media.mit.edu 1189451254 J * marcfiu ~mef@aegis.CS.Princeton.EDU 1189451254 J * fatgoose_ ~samuel@76-10-154-18.dsl.teksavvy.com 1189451254 J * mstrobert ~mstrobert@wkstn.wycliffe.ca 1189451254 J * mugwump ~samv@watts.utsl.gen.nz 1189451254 J * micah ~micah@micah.riseup.net 1189451254 J * MooingLemur ~troy@shells195.pinchaser.com 1189451254 J * bored2sleep ~bored2sle@66.111.53.150 1189451255 J * callahan ~chatzilla@host-181-96.dsl-sea.seanet.com 1189451255 J * _nkukard_ ~nkukard@dsl-241-41-162.telkomadsl.co.za 1189451255 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw 1189451255 J * tam ~tam@gw.nettam.com 1189451255 J * Hollow ~hollow@proteus.croup.de 1189451255 J * Supaplex ~supaplex@166-70-62-194.ip.xmission.com 1189451255 J * hardwire ~bip@ak.glaven.org 1189451269 J * zLinux ~zLinux@88.213.24.240 1189451269 J * quasisane ~sanep@c-76-118-191-64.hsd1.nh.comcast.net 1189451269 J * wenchien ~wenchien@59-105-176-11.adsl.static.seed.net.tw 1189451269 J * er ~yakker@aegis.CS.Princeton.EDU 1189451269 J * mountie ~mountie@trb229.travel-net.com 1189451269 J * dilinger ~dilinger@mail.queued.net 1189451269 J * neuralis ~krstic@solarsail.hcs.harvard.edu 1189451277 J * puck ~puck@leibniz.catalyst.net.nz 1189451277 J * mattzerah ~matt@121.50.222.55 1189451544 P * callahan 1189451556 J * moun ~moun@ABayonne-151-1-22-217.w83-203.abo.wanadoo.fr 1189451576 J * callahan ~chatzilla@host-181-96.dsl-sea.seanet.com 1189451615 M * moun anybody out there 1189451652 M * moun Need some help 1189451719 M * callahan We're building our own skeleton from our existing FC6 OS. Are there any general guidelines on which init.d services aren't needed in a guest. 1189451724 M * moun why my vserver can ping the host but not the outside 1189452180 J * ema ~ema@fw.galliera.it 1189452522 M * Bertl moun: probably some misconfiguration in your setup 1189452544 M * Bertl callahan: you probably want to avoid hardware specific stuff 1189452553 M * Bertl callahan: besides that, sysv should be fine 1189452620 M * callahan What about network daemons like iptables? Are those handled within the VServer or does it depend upon my setup? 1189452690 M * Bertl iptables is host specific 1189452749 M * callahan Specific to the 'host' and thus not needed in each 'guest'. Correct? 1189452751 M * Julius is there a way to install debian without the hardware specific stuff? 1189452765 M * Bertl callahan: but, if you do a special setup, you could for example parse (and after applying a certain policy) put those in a guest specific iptables chain 1189452782 M * Bertl Julius: the default install the tools (util-vserver) do should be fine 1189452791 M * Julius it works 1189452806 M * callahan Bertl: Thanks. Understood. 1189452806 M * Julius but the system throws some errors on startup 1189452814 J * arachnist_ arachnist@088156189068.who.vectranet.pl 1189452826 M * Bertl Julius: not seen any errors with 0.30.214 and etch here 1189452837 Q * arachnist Read error: Connection reset by peer 1189452847 M * Julius i guess i did something wrong 1189452879 N * arachnist_ arachnist 1189452885 M * Julius I'm currently uploading a new system image to my server, will be using linux-vserver from now on 1189452907 M * Bertl vserver etch32 build -m debootstrap --context 10104 --hostname etch32.debian.org --interface eth1:10.1.4.32/24 -- -d etch -m http://ftp.debian.org/debian -- --arch i386 1189452917 M * Bertl (just an example) 1189452984 M * Julius what does context mean? 1189453003 M * Julius kein einziger fehler :/ 1189453007 M * Julius ups 1189453018 M * Julius wrong window, sry 1189453026 M * Bertl context is the guest context 1189453039 M * Julius thats a unique id? 1189453043 M * Bertl yep 1189453048 M * Julius okay 1189453306 P * moun Kopete 0.12.3 : http://kopete.kde.org 1189454790 Q * Borg- Quit: fuck 1189454979 Q * Julius Remote host closed the connection 1189455096 Q * DavidS Quit: Leaving. 1189456096 Q * meandtheshell Quit: Leaving. 1189456500 J * Aiken ~james@ppp121-45-250-174.lns2.bne4.internode.on.net 1189456710 M * Bertl welcome Aiken! 1189456736 M * Aiken hello 1189457527 Q * rorem- Read error: Connection reset by peer 1189457783 Q * dna Quit: Verlassend 1189457865 J * rorem- ~roremtank@bzq-219-46-202.isdn.bezeqint.net 1189457910 M * mstrobert Why are there identical configuration files (like fc6/yum.repos.d/fedora-core.repo) in both /etc/vservers/.distributions and /usr/lib/util-vserver/distributions ? Is the one in /etc the right one to edit? 1189457960 M * daniel_hozac yes. 1189457975 M * daniel_hozac util-vserver doesn't put any files in /etc/vservers/.distributions/yum.repos.d 1189458043 Q * bonbons Quit: Leaving 1189458086 M * daniel_hozac +/, of course. 1189458185 M * mstrobert daniel_hozac: why are those files also in /usr, then? Are they just for reference when making your own file in /etc? 1189458212 M * daniel_hozac util-vserver installs the files in /usr/lib*/util-vserver. 1189458327 M * bXi 2.3.0.19 O_o 1189458336 M * bXi damn i missed one 1189458730 M * Bertl :) 1189458853 M * bXi whats new in .19 1189458882 M * Bertl some fixed which were missied in .18 1189458939 M * daniel_hozac just fixes an oops on mount -o tag. 1189459054 J * Aiken_ ~james@ppp121-45-250-174.lns2.bne4.internode.on.net 1189459250 J * mugwump_ ~samv@watts.utsl.gen.nz 1189459276 Q * Aiken resistance.oftc.net larich.oftc.net 1189459276 Q * bored2sleep resistance.oftc.net larich.oftc.net 1189459276 Q * mugwump resistance.oftc.net larich.oftc.net 1189459276 Q * mstrobert resistance.oftc.net larich.oftc.net 1189459276 Q * fatgoose_ resistance.oftc.net larich.oftc.net 1189459276 Q * marcfiu resistance.oftc.net larich.oftc.net 1189459276 Q * coderanger__ resistance.oftc.net larich.oftc.net 1189459276 Q * MooingLemur resistance.oftc.net larich.oftc.net 1189459276 Q * micah resistance.oftc.net larich.oftc.net 1189459451 J * fatgoose ~samuel@76-10-154-18.dsl.teksavvy.com 1189459456 J * mstrobert ~mstrobert@wkstn.wycliffe.ca 1189459467 J * coderanger__ ~coderange@1cc-dhcp-106.media.mit.edu 1189459495 Q * comfrey Ping timeout: 480 seconds 1189459523 J * MooingLemur ~troy@shells195.pinchaser.com 1189459529 J * bored2sleep ~bored2sle@66.111.53.150 1189459571 J * ktwilight_ ~ktwilight@34.81-66-87.adsl-dyn.isp.belgacom.be 1189459724 J * micah ~micah@micah.riseup.net 1189459780 M * micah is there a vnetstat like way I can look to see what vserver is grabbing a port? 1189459797 M * Bertl yep, entering the network context should do that 1189459820 M * daniel_hozac ncontext --nid 1 --migrate -- netstat ... 1189459846 M * micah yeah, I guess I just need to write a for loop to try and locate which vserver is listening on a particular port on this IP 1189459886 M * micah maybe vsomething 1189459895 M * daniel_hozac sure 1189459975 J * coderanger ~coderange@kantrn.stu.rpi.edu 1189459975 Q * coderanger_ Read error: Connection reset by peer 1189459982 Q * ktwilight Ping timeout: 480 seconds 1189460484 Q * callahan Quit: User is away. 1189460867 M * Bertl okay, time for a nap ... back later ... 1189460873 N * Bertl Bertl_zZ 1189461706 N * mugwump_ mugwump 1189464347 J * esa ~esa@ip-87-238-2-45.adsl.cheapnet.it 1189464348 Q * eSa| Ping timeout: 480 seconds 1189464999 Q * yarihm Quit: Leaving 1189466285 Q * ensc Remote host closed the connection 1189466347 Q * coderanger__ Quit: coderanger__ 1189466595 Q * ema Quit: leaving