[00:01] yes [00:01] what version/patches? [00:02] h/o [00:04] ok im trying now with new quota files and quotacheck and quotaon using -F vfsv0 [00:04] je (~je@hd5e25b7f.gavlegardarna.gavle.to) joined #vserver. [00:04] je_ (~je@hd5e25b7f.gavlegardarna.gavle.to) left irc: Read error: Connection reset by peer [00:04] linux-2.4.24+patch-2.4.25-pre7+patch-2.4.25-pre7-vs1.24+patch-2.4.25-pre7-vs1.24-q0.12+delta-2.4.25-pre7-vs1.24-q0.12-t1 [00:05] hmm okay, sounds reasonable ... [00:05] thanks [00:05] Action: mids reads backlog [00:06] hmm nope still doesnt seem to be working. [00:07] i can try with those set of patches though. [00:07] talon: could you try with the patches mids uses, just to verify that this works for you (or not?) [00:07] yeah, but keep the old ekrnel [00:07] I mean the sources, we can do a diff between them later ... [00:08] yeah i will keep my old 2.4.24 kernel. [00:08] I'm officially updating the quota hashes stuff for 2.4.25-pre8 now ... will take a little ... [00:09] i have the 2.4.25 i built last night too but it basicly does the same thing which is why im back to the old kernel. but im building a new one now (keeping the old sources for both) [00:10] good [00:10] oops, I got the hang-fix patch too [00:10] but I did do that manually [00:10] so forgot to mention i [00:10] wasn't sure if this was included in the t1, but it seems t1 is only the debuging stuff ... [00:10] +t [00:11] manually = using vi instead of patch? [00:11] yes [00:11] Bertl: debugging + getting it to compile [00:13] Bertl: i can send you my kernel config file if that helps. [00:14] well, not atm, maybe later ... [00:14] I also use UID24/GID24 [00:15] but not bind mount (dont even know what it is/does) [00:16] mids: want to know? [00:16] remounts part of a filesystem somewhere else. bind mount extensions makes it able to mount read only. [00:16] thx [00:17] basically mount --bind /tmp /mnt/tmp [00:17] would allow you to access /tmp via /mnt/tmp [00:17] it's some vfs magic ... [00:17] the funny part is while mount --bind -o ro /tmp /mnt/tmp [00:17] works, it doesn't do what you would expect ;) [00:18] what does it do then? [00:18] it mount it rw, as the original ... [00:18] this is because there isn't a superblock info which contains the ro/rw stuff ... [00:19] so I coded a kernel aptch, about a year ago, which adds this (ro) and other mount flags to the --bind mounts ... [00:19] (called bind mount extensions, or short bme) [00:20] ah [00:20] patch-2.4.25-pre7-vs1.24 i cant find in the experimental section or the releases. did you mean the pre6-1.24 instead? [00:21] Action: mids wouldnt be surprosed if pre7 was pre6 with a sed -e 's/pre6/pre7/' over it [00:22] I got tired of having to fix the Makefile patch failure [00:22] i just want to be sure im applying the right patch files. [00:22] yeah, use pre6 [00:24] Filther (Kati@dial055008.vnet.hu) joined #vserver. [00:24] hi [00:25] hi Filther! [00:25] talon: ext2 or ext3? [00:25] has anyone experience on running udp based apps within a vserver? [00:25] hmm, like ftp? [00:25] ? [00:26] like udp servers. [00:26] like game servers [00:26] udp like udp/ip, right? as used in ftp? [00:26] ftp uses udp??? [00:26] ok so once i have the release of 1.24 for 2.4.25 installed i install patch-2.4.25-pre7-vs1.24-q0.12.diff and http://vserver.13thfloor.at/Experimental/delta-2.4.25-pre7-vs1.24-q0.12-t1.diff and the quota hang fix right? [00:27] udp like udp, as opposed to tcp/ip :) [00:27] okay, any issues with that? [00:27] I have something like this, running in a vserver: [00:27] udp 0 0 v1.vpserver:27888 *:* 1762/avp2linux [00:28] but [00:28] clients withing my LAN can't detect the server [00:28] other, tcp-based servers work fine [00:28] that is some inetd based entry right? [00:28] this is from "netstat -ap" [00:29] hmm looks like teh hang fix patch is already included in the delta. [00:29] I've given this vserver full privileges considering the NET_* ones [00:30] which isn't such a good idea ... [00:30] yes, it is not [00:30] ...and it didn't change anything [00:30] okay, are you able to connect to that port? [00:30] ;> [00:30] with what? [00:30] hmm, good point, ... nmap for example? [00:31] oh would anyone be interested in a patch i made that prevents users with CAP_NET_RAW form setting promisc mode? its documented in capabilitys.h that only CAP_NET_ADMIN should be able to do so. but af_packet.c lets you anyway without a two line fix. [00:31] miller7 (none@213.239.180.106) left irc: Ping timeout: 492 seconds [00:31] talon: sure, we might include that immediately ;) [00:31] could you tell me how to? [00:31] you know nmap? [00:32] not really. [00:32] use google, and isntall it on the client ... [00:32] Bertl: im just not sure ive covered all place in teh kernel where you can set promisc mode. im new to the linux kernel. [00:32] alright [00:32] talon: np, I'll review it anyway ... [00:33] Bertl: ok i will put it up so you can look at it. [00:34] or post it on the ml [00:36] if I installed nmap, what should I do next? [00:36] http://www.amoebasoft.com/~talon/promiscfix.patch [00:36] Filther: make yourself familiar with the command line options, and try to conenct to the port [00:36] if you want i can post it to the ml i havent gotten around to subscribing yet. [00:37] wouldn't hurt ... I guess ... [00:37] talon: what do you use to work on the kernel sources? [00:38] Bertl: just vim and the userland tools that come with slackware. [00:38] so basicly an editor diff and patch. [00:38] just a suggestion, cscope is a nice tool ;) [00:38] oh ive used cscope as well. [00:38] although not in this case. [00:39] had to pull it out last night to figure out what was going on with the dquot struct though. [00:39] very handy. [00:39] yeah, took me some time to figure out the -p switch ... [00:40] if you look at the /~talon part of that page you will find a utility to make certain Sun machines power back up on a specified date/time after power down. [00:41] im more of a solaris/BSD guy than a linux guru. ive only had to work on linux projects in the past year or so. [00:41] hmm, so you probably have some suns standing around right? [00:42] yeah i have 5 ultra 10s under my desk now. [00:42] two of them not being used atm. [00:42] that would allow to test vserver on sparc, right= [00:42] s/=/? [00:42] yeah i suppose i could throw linux on one and play with it. [00:42] would be great to get some feedback on that platform ... [00:42] i might even be able to arrange for you to have access to one to play with. [00:43] cool [00:44] waiting for the kernel to finsih compiling. im using a vmware box so its a bit slow. [00:45] okay, just verified that the 'new' quota hash patch works ... now for context quota ... [00:45] root -- 25 0 0 4 0 0 [00:45] root -- 1055 0 0 5 0 0 [00:45] bash -c 'touch /mnt/part{1,2}/{a,b,c,d}{1,2,3,4}' [00:45] root -- 25 0 0 20 0 0 [00:45] so your inode count actually increases then? [00:45] root -- 1055 0 0 21 0 0 [00:46] thats better than what i have. [00:46] that is without the context quota stuff yet ... [00:46] oh. [00:46] only quota hash abstractions on 2.4.25-pre8 [00:46] nmap says the udp port is open [00:47] qha is the base for all accounting done with the context quota patches ... [00:47] Filther: that means that something arrives there and comes back ... [00:47] now I would suggest to try 'normal' udp probing and make a log with tcpdump on the vserver hsot [00:48] s/hsot/host/ [00:48] oh... [00:49] mcp (~hightower@wolk-project.de) left irc: Ping timeout: 492 seconds [00:49] tcpdump -s 10000 -vvnei [00:49] ill trz [00:49] i'll try [00:49] :) [00:49] tcpdump -s 10000 -vvnei port 27888 [00:51] Nick change: _Doener_ -> Doener [00:51] hmm.. [00:51] SELECT [00:51] CONCAT(u.vorname, ' ', u.nachname) AS kundenname, [00:51] IF(u.emailForLimit='', u.email, u.emailForLimit) AS email, [00:51] IF(r.emailFrom='', r.email, r.emailFrom) AS sender, [00:51] IF(r.firmenname='', CONCAT(r.vorname, ' ', r.nachname), r.firmenname) AS sendername [00:51] FROM [00:51] damn... [00:51] arch/mips/Kconfig:88:warning: defaults for choice values not supported [00:51] drivers/ieee1394/Kconfig:60:warning: enable is only allowed with boolean and tristate symbols [00:51] that's what i wanted... [00:51] Action: Doener still has that copy'n'paste curse on him... [00:53] i get this error generally when doing configuration for mips... [00:53] okay, so that means what? we have to implement the defaults for choices for mips? [00:53] but it's only a warning right? [00:53] does this appear on i386 too? [00:53] no [00:53] i don't know for now... but i found some mips/linux how-to saying that the mips branch is quite often broken [00:54] Filther (Kati@dial055008.vnet.hu) left irc: Remote host closed the connection [00:54] seems so ... [00:55] sgi mips or some other barnd mips box? just curious. [00:55] i couldnt imagine running linux on my indy. [00:55] talon: we are doing x86 cross compiling ... [00:55] just cross compiling... [00:55] Filther (Filther@dial055008.vnet.hu) joined #vserver. [00:55] re [00:56] my computer crashed [00:56] ;> [00:56] i've tried some rXXXX cpu's as well as mips and sb1 or what it is called [00:56] most didn't even start to compile [00:56] i never had much success with cross compiling. [00:56] Doener: try MIPS32 and MIPS64 .. wait a moment I have a piece of config here ... [00:57] ok finally got the kernel built. [00:57] could you please type the tcpdump command again? [00:57] CONFIG_MIPS=y [00:57] CONFIG_MIPS32=y [00:57] CONFIG_CPU_MIPS32=y [00:57] CONFIG_MIPS=y [00:57] CONFIG_MIPS64=y [00:57] CONFIG_CPU_MIPS64=y [00:57] CONFIG_64BIT=y [00:57] 3 for mips32 , 4 for mips64 [00:58] 22:51 < Bertl> tcpdump -s 10000 -vvnei port 27888 [00:58] mips32 failed for me (at least compile started ;) [00:59] but this configuration errors made me suspicious [00:59] guess we have to contact mips folks regarding this ... [00:59] I jsut hope they are not as stubborn as the parisc folks ... [01:00] thanks [01:04] ok the new kernel with his patches doesnt seem to work either. im going to try quotas without context and then im going to try your new pre8 patch. [01:04] okay ... [01:05] guess until then, I'll have the context quota stuff updated too .. [01:07] anything in particular you want me to try? [01:07] well, you did try for the 'host' context right? [01:08] im doing that now. [01:08] with the new kernel. [01:08] okay, let me know how this goes ... [01:08] im just goign to try inode quotas on teh host context 0 and see if it does teh same thing. [01:09] Filther: any tcpdumps yet? if yes, please upload them somewhere ... [01:10] yes [01:11] hmm looks like i am using vfsv0 for the context but forcing it for ext3 filesystems isnt kosher. [01:12] hmm, works for me here .. what is your problem? [01:13] (my tests where on ext2 & ext3 ) [01:13] http://akarkivagyok.freeweb.hu/temp/dump1_1 [01:15] hmm, illusions ;) [01:15] it just gives me no such devices or address when it tried quotaon. [01:16] did you add a quota hash for context 0 ? [01:16] no. that could do it. [01:16] the page is very old :) [01:16] i didnt realiz ei needed to do that for teh host context. [01:17] it was my first html site [01:17] okay, you ahve to search for the reply packet ... [01:17] those are only the nmap? trace packets ... [01:18] so either there is a icmp reply to that, telling that ther is no such port open, or an udp connect reply ... [01:18] this should be something originating from the host/vserver [01:19] talon: is this William, or really Bill? [01:19] crap. i have to mount anything i want to use quotas on with tagctx? after i have teh context quota patch applied ? [01:20] hmm, well yes ... [01:20] Bertl: the name is really william but i go by bill. [01:20] Bertl: even for stuff on the host not having to do with vservers at all? [01:20] yes, because the quota hashes are modified for that purpose ... [01:21] it might be possible (in a future version) to have both, but atm, it is not ... [01:21] serving (~serving@213.186.189.107) joined #vserver. [01:21] ok, wasnt aware of that. that would be why it didnt work. [01:21] at least for testing outside of a context. [01:22] was trying ot use normal quotas on teh root fs. [01:22] well, that don't work as you already found out ;) [01:22] s/don't/soesn't/ [01:22] forget it ;) [01:22] here's a line that I definetly don't get from the vserver, but only from the main server [01:22] 00:26:18.265109 vpserver.27888 > heimdall.2034: udp 457 (DF) [01:22] I should look for something like this, which is absent [01:23] do you get that with the same'client' (nmap or whatever?) [01:23] yes [01:23] but not when the server runs on v1.vpserver, of course [01:23] :\ [01:23] just because it might be the answer of the server to a client ... [01:24] (I've run the server on the working, main server now) [01:24] okay, silly idea, does the server write some logs? [01:24] something like connected ... [01:24] doesn't seem to write logs [01:25] what server is that, if you don't mind me asking? [01:25] Action: talon updates the rc scripts again. [01:25] alien vs predator 2, I believe [01:26] hmm, well I guess that doesn't do much logging ... [01:27] seems to me when I run the avp2 server on one of the vservers, only the packets from the client searching for the server arrive, but there's no answer [01:28] they don't get 'noticed' [01:28] well, there are many possible reasons for that ... [01:29] for example, if the server decides to 'reverse' the client ip after some connect ... you can lock waiting for a nameserver query [01:29] nameserver works fine [01:29] or some security check decides I don't like that ip, I do not answer ... [01:29] there's no security here, I believe [01:29] :) [01:30] it's hard to tell, I assume you do not have any source of that server, right? [01:30] no, none at all [01:30] udp is enabled by default, no? [01:30] well, we can try with strace ... [01:30] so it should be working fine [01:30] could it be just with the routing tables? [01:30] yes, udp is enabled and should be working, everything else is a bug ... [01:31] do you ahve special routing tables? [01:31] no [01:31] 192.168.1.0 * 255.255.255.0 U 0 0 0 eth0 [01:31] you know hping2 ? [01:31] I don't [01:31] : [01:31] ) [01:31] okay, next round googling ... [01:31] ;> [01:31] install it on the host and the vserver ... [01:32] and make sure that tracepath and traceroute is installed too (on the vserver) [01:32] ahh, by the way, you are not using ipv6, are you? [01:32] the client is on windows, if that matters [01:32] no ipv6 here [01:32] not even enabled in the kernel? [01:32] nope [01:32] good [01:33] strange thing is that all other tcp services work fine [01:33] well, tcp and udp is different ... [01:34] it might easily be, that some very old udp issue has survived all tests and checks ... [01:34] so i [01:34] ok [01:34] should install it on the vserver [01:34] inode quotes work in context 0 [01:35] had to set up a user with a home dir in the /vservers fs because root is initally mounted readonly. [01:35] humm, that is very interesting ... [01:36] could you setup a quota hash for xid=0 and xid= [01:36] and see if the inode quota is accounted to the wrong hash? [01:36] hmm let me try somthign real quick. i havent tried it with the vservrs fs mounted with the quota options yet. [01:36] i was mountign it with just tagctx before. [01:37] I'll be away for some time [01:37] what should I try doing with hping2 when I'm back? [01:37] well try to send udp pings to the client to [01:38] check if that works, and verify the path with tracepath [01:39] ok [01:39] thanks [01:39] bingo [01:39] thats was the problem. [01:39] hmm, please elaborate? [01:40] on more thing [01:40] i was mountign /vservers with just rw,tagctx. [01:40] instead of tagctx,userquota,grpquota. [01:40] I've just tried running the avp2 server on the main server with chbind v1.vpserver, and it worked fine [01:40] for some reason block quotas still worked in that state. [01:40] hmm, and this makes a difference? [01:40] but not inode quotas. [01:40] does this mean something? [01:40] Bertl: aperrantly. [01:40] Filther: hmm, yes this means that the vserver udp code is working fine ;) [01:41] ;> [01:41] did you try with one ip or more than one [01:41] one ip [01:41] and do you use one or more than one in the vserver? [01:41] the one which the vserver is using [01:42] so you did chbind --ip [01:42] yes [01:42] can I have a look at your vserver config? [01:42] ok [01:43] talon: I'll investigate that anyway, because the quota options are ignored on all filesystems ... [01:43] so they should not have any effect, although they are needed by the quota utilities ... [01:43] http://akarkivagyok.freeweb.hu/temp/v1.conf [01:44] now I have to go [01:44] I'll be back soon [01:44] okay .. cu later [01:45] Bertl: i think you shoudl be able ot reproduce it by moutnign the vserver fs with just tagctx and then just using quotas normally inside your vservers. im not sure many people even bother with inode quotas. [01:45] so im not sure anyone would have noticed. [01:45] since the block quotas work fine. [01:45] I'll check that ... for sure, by the way I'm almost finished with the update of the aptches ... [01:45] yeah i will give it a shot when your done. [01:46] as well as seeing if i can get things working fine with my old 2.4.24 kernel as well. [01:46] would be interesting, if they differ from your patched up version ... [01:48] Action: talon tests again with vserevrs mounted with just tagctx just to be sure. [01:49] its a bit frustrating trying to find the right combination to make software fail but at least for me its also a little bit of fun as well. [01:50] well testing is sometimes fun, if done systematically ... [01:52] yep. [01:52] remountign /vservers without usrquota,grpquota makes inodes break again. [01:53] Bertl: really hard to keep track of everything though. even worse when you have almost zero short term memory. [01:53] I suspect this is a quota related bug, which goes unnoticed, because nobody actually is able to do that on a normal system ... [01:53] talon: what did you say where your findings? [01:53] ;) [01:54] Bertl: sounds about right to me. i dont think anyone expected quotas to be used that way. [01:54] okay just another idea, could you try to use quota on an unpatched system, by faking the quota flags in the mtab? [01:54] esp on a filesystem that wasnt mounted with quotas enabled. [01:54] Bertl: heh i could try that. [01:54] rere [01:54] shoudl be interesting at least. [01:55] not even sure if that would work. [01:55] yeah, I suspect unpatched quota will work the same ... [01:55] so, the config file looks alright (except for the CAP_ flags, I believe)? [01:55] but i see where your coming form. [01:55] ok let me see if i can get it working. [01:56] Filther: I would suspect the magic ipmask thingy .. but lets verify that ... [01:56] hmm [01:56] I think you have a point. [01:56] ;> [01:56] I'll see [01:57] try grep ipv4root /proc/self/status [01:57] with the bind --ip command and inside the vserver [01:57] Bertl: old 2.4.22 kernel with netfilter patches ok? [01:58] or yu want a specific clean kernel build? [01:58] guess so, if quota is enabled ;) [01:58] hmm come to think of it im not sure if it is. never mind i will just build a new 2.4.24 kernel. [01:58] main server (with bind): [01:58] ipv4root: 0101a8c0/00ffffff [01:58] ipv4root_bcast: ffffffff [01:58] ipv4root_refcnt: 8 [01:58] talon: well, actually a 2.4.25-pre would be better as they changed some things ... [01:59] ok i will do a clean pre8 [01:59] vserver: [01:59] ipv4root: 6401a8c0/00ffffff [01:59] ipv4root_bcast: ff01a8c0 [01:59] ipv4root_refcnt: 26 [02:00] hmm, quite a difference ... [02:00] yup [02:00] moment... [02:00] the .1 vs .100 is a minor issue [02:00] but the broadcast address differs ... [02:00] Bertl: i seem to always run into odd things like this no matter what im using. i guess i never seem to use software teh same way normal people do. :\ [02:01] well, I appreciate such people testing stuff .. because they find the corner cases ... [02:02] does anybody know if reiserfs supports quotas? [02:02] im actually working on a distro based around vserver btw. [02:03] cool! [02:04] im still taking notes on vserver. havent gotten to teh good stuff yet. other than a base linux distro based on slackware with a package manager borrowed from archlinux. [02:05] what should I set netmask to< [02:05] ? [02:05] but eventually id like to have somthing similar to the cobalt with virtual dedicated servers. [02:05] and a much saner os than cobalt linux ever was. [02:05] Filther: try to repeat it with chbind --ip --bcast 192.168.0.255 [02:05] that should give you the same as inside the vserver ... [02:06] Filther: try to repeat it with chbind --ip --bcast 192.168.1.255 [02:06] sorry ... [02:06] ok [02:06] but verify with the grep [02:07] talon: sounds good ... [02:10] if i try chbind --ip 192.168.1.100 --bcast 192.168.1.255 ./avp2linux [02:10] im mainly developing it to sell some 1u rackmount boxes. but any improvements i can make to vserver i certanly plan to contribute back. certanly in everyones best intrests to have the best vserver code possible. i saw the FreeVPS fork. its too bad alex doesnt give any of his functionality back to the original vserver tree. [02:10] the client doesn't detect the avp2 server [02:10] like withing the vserver. [02:10] im really interested in seeing how 1.3.x will turn out. [02:12] hmm, I just tried to reproduce the 'no usr/grpquota' issue on my 'newly' patched system, but it works as expected ... [02:12] mount -t ext2 -o rw,tagxid /dev/discs/disc1/part1 /mnt/part1 [02:12] mount -t ext3 -o rw,tagxid /dev/discs/disc1/part2 /mnt/part2 [02:12] /bin/sed -i 's/tagxid/tagxid,usrquota,grpquota/g' /etc/mtab [02:12] cqhadd -v -x 0 /dev/discs/disc1/part1 [02:12] cqhadd -v -x 0 /dev/discs/disc1/part2 [02:12] quotacheck -F vfsv0 -maug [02:12] quotaon -F vfsv0 -vaug [02:12] bash -c 'touch /mnt/part{1,2}/{a,b,c,d}{1,2,3,4}' [02:12] repquota -vaug [02:13] tagxid? what about tagctx? [02:13] or is that teh same thing. [02:13] well, tagxid is the 'new' option, you have to use tagctx for now ; [02:13] ive only tested it inside a vserver context when moutning with only tagctx. [02:13] not in ctx 0. [02:14] okay, will verify in xid != 0 [02:15] id disable quotas in xid 0 for good measure too. [02:16] seems the broadcast address is all that matters [02:17] Filther: funny stuff, well, you have a workaround for now, and I would ask you to post some lines on the mailing list ... [02:17] (workaround actually is specifying the broadcast in the config) [02:18] strange. [02:18] talon: okay, I can reproduce that with xid != 0 ... [02:18] it doesn't work now [02:18] look [02:18] it worked with [02:18] chbind --ip 192.168.1.100 --bcast 255.255.255.255 avp2linux [02:18] Bertl: cool im not out of my mind :) [02:19] it didn't with: chbind --ip 192.168.1.100 --bcast 192.168.1.255 avp2linux [02:19] I set (within the config file) the bcast of the vserver to 255.255.255.255 [02:19] but it's not working [02:19] that pre8 kernel is still building. [02:20] Filther: test if the broadcast is set now inside the vserver ... with the grep line ... [02:20] but [02:20] if i do chbind --ip 192.168.1.100 --bcast 255.255.255.255 avp2linux inside the vserver [02:20] (which is pretty pointless) [02:21] it works. [02:21] verify that broadcast via grep first ... [02:21] i was thinking about making a mount helper daemon for vserver with a userland tool that replaces mount. [02:21] ipv4root: 6401a8c0/00ffffff [02:21] ipv4root_bcast: ff01a8c0 [02:21] ipv4root_refcnt: 16 [02:21] so we have 192.168.1.255 again inside ... [02:22] so root users can mount nfs shares inside of vserver without any special privs. [02:22] yes [02:22] kramer (~kramer@80.86.100.172) joined #vserver. [02:22] hi all [02:22] hi kramer! [02:23] talon: that would compromise vserver security ... [02:23] I need some help, I'm a vserver newbie [02:23] aren't we all? [02:23] Bertl: well it would need to be able to set mount limitations per vserver ofcourse. [02:23] trying to set up something pretty common [02:23] cant allow anyone to mount anything anywhere. [02:24] talon: for example allowing working device nodeswould be an issue ... [02:24] just relative to the vserver root. [02:24] and im only thinking nfs mounts for now. [02:24] mounting real devices would be bad. [02:24] havent really explored the idea fully yet. [02:25] any idea how to set that bcast? [02:26] for me, it doesn't really matter now, I can supply that chbind command. but it would be 'nice' to see it working without that [02:26] well, obviously that isn't verified inside the vserver (which is a bug IMHO), so using chbind inside should work for now ... [02:26] ok [02:26] thank you for your help! [02:26] ok, here it goes: kernel 2.4.24-vs1.24, util-vserver-0.27-0.i586.rpm, vserver starts ok, but the main problem is that i have no networking whatsoever.... [02:26] depends on the tools you are using, (which version by the way?) [02:27] sorry kramer, that was for Filther! [02:27] kramer: what is your vserver config? [02:27] I'll be going now [02:27] tried to read any docs I could lay my hands on for the past 3 days, and it feels like it's something trivial [02:27] thanks again for the help. [02:27] that i'm overlooking [02:28] Filther: np, have a nice one [02:28] ;> [02:28] kramer: try that script: http://vserver.13thfloor.at/Stuff/testme.sh [02:28] bye! [02:28] Filther (Filther@dial055008.vnet.hu) left irc: Quit: Leaving [02:28] kramer: and let me know what it reports ... [02:28] the script fails at the second last line... [02:28] that is okay for stable ... [02:28] i did it some hours ago [02:28] so your chbind is working ... [02:29] you have 'normal' network setup on the host, I assume? [02:29] sure :) [02:29] okay, can you show me your vserver config file? [02:29] maybe upload it somewhere? [02:29] /etc/vservers? [02:30] the /etc/vserver/.conf yes [02:30] +s [02:30] Bertl: think you could update the bme patch for 2.4.24-vs124 plus teh context quota patch? and 2.4.25? it works ok with those. but there are a couple of rejected hunks. form what looks ot me like duplicate code between bme and the quota patch. [02:30] talon: yeah, I'll do that ... right after the last test for the q0.13 is done ... [02:31] hmm, guess now I have to update my compiler ... enrico will be really happy ... [02:32] hope i havent taken up too much of your time on all of this. i literally have all day to test this stuff since im supposed ot get vserver integrated in teh distro anyway. [02:32] well, if you feel too guilty, have a look at http://www.13thfloor.at/vserver/donate/ [02:32] 8-) [02:33] probably will if we start selling servers. [02:33] well the sequence of options _is_ intentional ... [02:34] so for example testing on that sun would be really appreciated ... [02:34] Bertl: ok, it's on ftp://gemini5.ines.ro/pub/test.conf [02:34] yeah. thats no problem i have to wait til my network guy gets back from NYC so he can set it up on a private segment. [02:35] and i can get sparclinux installed and give you root on it. [02:35] kramer: okay add an eth0: before that ip, so it looks like IPROOT="eth0:80.86.106.32" and it should work (where eth0 is just an example) [02:35] its got solaris 9 on it now. [02:36] brrr ... [02:36] ;) [02:36] id do it myself but im not comfortable playing with someone else undocumented iptables rules. [02:37] I can understand that ... np [02:39] http://vserver.13thfloor.at/Experimental/patch-2.4.25-pre8-vs1.24-q0.13pre1.diff [02:39] knowing rob he will probably slap gentoo on the sparc. [02:39] any distro you would prefer? [02:39] thanks a lot, Bertl [02:39] well, I don't care much, I know debian a little and I use mandrake for my stuff ... [02:39] np, btw, you know what's the difference? [02:40] it works, knew it was something really easy [02:40] bye all [02:40] kramer (~kramer@80.86.100.172) left irc: Quit: Leaving [02:40] hmm, probably not ... but he'll come back if he needs to know it ... [02:41] most things about vserver were pretty easy to figure out until i hit the context quota stuff. [02:41] im sure you get people confused about that a lot though. [02:42] the documentation is really an issue, I guess I'll spend some time on that in the next few month ... [02:43] well, im going to be writing some documentation of my own for my distro. id be glad to give you a copy of anything i write up. [02:43] http://vserver.13thfloor.at/Experimental/patch-2.4.25-pre8-vs1.24.diff [02:43] talon: I'd appreciate that, any vserver related docu is useful ... [02:45] i at least plan to do a more detailed context quota howto. for the way that im currently using it. [02:45] by the way, is talon farscape related? [02:45] gargoyles actually. [02:45] but i am a farscape fan as well. [02:46] ah okay ... [02:47] Bertl: im probably going to make some manpages for the vrsetup and the cq commands too. although for now it woudl only be about the options i know exist. [02:48] havent dug through the source yet. [02:48] well, gues we should update them too .. just found that the help isn't very helpful ... [02:50] that old bme0.03 still applies without any issues ... [02:50] fascinating, never thought that it would last that long ... [02:50] with the base kernel yeah. [02:50] add in teh context quota patch and theres some rejected hunks. [02:50] it still works though. [02:51] maybe teh bme patch could be integrated in with vserver? it certanly is very useful for saving disk space if you dont have rpm for the unify scripts to work on. [02:53] I was thinking about that, but there was no 'official' request for that yet, so I kept this out of the vserver patch .... [02:53] i wish bme had been integrated into 2.4.23 [02:53] what happend with that? [02:54] its such a useful feature. you can do it out of the box on any BSD distro. [02:54] with null and union mounts. [02:54] well, I posted it about 20 times on lkml, always updated ... [02:54] Al Viro had no time to respond ... and Marcelo and Andrew said, if Al Viro says it's okay, we include it ... [02:55] so it wasn't included ;) [02:55] ahh. [02:56] after some time I gave up reposting it ... especially because obviously nobody needed it ... otherwise there would have been some requests to include it ... [02:58] Bertl: i mentioned it before on some unix irc channels and had a few people quite interested. [02:58] ah I see ... the quota stuff collides somewhere with the bme ... I'll fix that tomorrow ... [02:58] they didnt even know it was possible. [02:58] well, unfortunately those people do not read lkml ;) [02:59] I guess the chances for 2.4 are gone, feature freeze ... [02:59] one guy even though moutn -oro --bind did work because mount showed it as readonly. [02:59] he never tested to see if that was really the case. [02:59] yeah, most people do ;) [03:00] i alwasy assume im being lied to until i verify it. [03:00] okay, I have to get up early today ... so I'll go to bed now ... [03:00] the bme will probably be updated for q0.13 tomorrow ... [03:00] ok i need ot grab somthing to eat anyway. [03:01] cool. thanks for all your time. [03:01] and I'll ahve a look at those accounting issues ... [03:01] np, have a good wossname ... [03:01] cu around [03:01] bye all, have a good whatever too ... [03:01] Nick change: Bertl -> Bertl_zZz [04:14] suhcoolbro (~Suh@216-161-89-245.ptld.qwest.net) left irc: Quit: NO CARRIER [05:37] hmm interesting. its not just mounting teh vserver fs without usrquota,grpquota quotas have to be enabled on teh fs in ctx 0. in order for the inode quotas to work propperly as well. very strange. [06:30] virtuoso (~shisha@ip114-115.adsl.wplus.ru) left irc: Ping timeout: 480 seconds [06:38] virtuoso (~shisha@ip114-115.adsl.wplus.ru) joined #vserver. [07:47] serving (~serving@213.186.189.107) left irc: [08:19] serving (~serving@213.186.189.107) joined #vserver. [08:30] ftp21 (mree@c-67-160-157-145.client.comcast.net) joined #vserver. [08:51] Which Linux is the best to install vservers on? debian? [08:57] noel- (~noel@pD952C364.dip.t-dialin.net) got netsplit. [08:58] noel- (~noel@pD952C364.dip.t-dialin.net) returned to #vserver. [08:59] all the links to vserver system images are broke.. anyone know any that work? [09:04] Nick change: Bertl_zZz -> Bertl [09:05] morning everyone! [09:08] noel_ (~noel@pD952C539.dip.t-dialin.net) joined #vserver. [09:11] ftp21: hmm, why do you need a system image? [09:15] noel- (~noel@pD952C364.dip.t-dialin.net) left irc: Ping timeout: 504 seconds [09:18] dunno. why not? They had a bunch of links to them.. thought i would check them out [09:19] ah, okay, you got vserver kernel working? [09:19] or just looking around? [09:56] ftp21 (mree@c-67-160-157-145.client.comcast.net) left irc: Ping timeout: 492 seconds [10:12] Nick change: Bertl -> Bertl_oO [11:14] loger joined #vserver. [11:36] Bertl: interesting. it seems that its not the mount options. its not enabling quotas in context 0 before enabling them in other contexts that triggers the behavior. with the inode quotas breaking. [11:36] as far as i can tell. [11:37] does that with 2.4.24 and 2.4.25. [11:50] but if you add a context quota for ctx 0 and run quotaon for the /vservers fs it just works as you would expect. [11:53] hmm not sure if this means anything or not. but even you have it all working in the other context with quoats enabled and everything. and without shuttign down the vserver. if you go back to context 0 and disbale quotas inode quotas break again. (and come back again in the already running context if you re-enable them) [11:55] no wait i spoke too soon. they dont came back. [11:56] sucky quotas [11:57] they do come back after you totally restart teh context though without a reboot. [11:58] i just wonder whats so special about context 0 with regards to inode quotas. [11:59] just figured id lay out a few more observations before i go to bed. [11:59] hopefully Bertl_oO can do something with it [12:00] thats what im hoping. if its not a bug it at least might lead to learning somthing not previously known about how the context quota patch interacts with the quota system. [12:02] anyway im off to bed. [12:14] nite [12:29] mcp (~hightower@wolk-project.de) joined #vserver. [12:31] noel_ (~noel@pD952C539.dip.t-dialin.net) left irc: Ping timeout: 504 seconds [13:31] Filther (Filther@dial062077235063.vnet.hu) joined #vserver. [13:31] hi [13:38] serving- (~serving@213.186.189.107) joined #vserver. [13:38] serving (~serving@213.186.189.107) got netsplit. [13:38] kestrel_ (athomas@home.swapoff.org) got netsplit. [13:38] deadguy (deadguy@bananajoe.big.du.se) got netsplit. [13:38] Zoiah (Zoiah@matryoshka.zoiah.net) got netsplit. [13:38] meebey (meebey@meebey.net) got netsplit. [13:38] mids (mids@mids.student.utwente.nl) got netsplit. [13:38] riel (~riel@riel.netop.oftc.net) got netsplit. [13:38] stupidawy (foo@198.77.239.131) got netsplit. [13:38] deadguy (deadguy@bananajoe.big.du.se) returned to #vserver. [13:39] Zoiah (Zoiah@matryoshka.zoiah.net) returned to #vserver. [13:39] Nick change: Bertl_oO -> Bertl [13:41] meebey_ (meebey@meebey.net) joined #vserver. [13:41] kestrel_ (athomas@home.swapoff.org) returned to #vserver. [13:41] mids (mids@mids.student.utwente.nl) returned to #vserver. [13:41] meebey (meebey@meebey.net) returned to #vserver. [13:41] riel (~riel@riel.netop.oftc.net) returned to #vserver. [13:41] stupidawy (foo@198.77.239.131) returned to #vserver. [13:41] meebey (meebey@meebey.net) left irc: Ping timeout: 501 seconds [13:43] hi Filther! [13:49] serving (~serving@213.186.189.107) got lost in the net-split. [13:53] Nick change: Bertl -> Bertl_oO [14:17] kestrel_ (athomas@home.swapoff.org) left irc: Ping timeout: 501 seconds [14:17] mids (mids@mids.student.utwente.nl) left irc: Ping timeout: 501 seconds [14:18] meebey_ (meebey@meebey.net) left irc: Remote host closed the connection [14:18] meebey (meebey@meebey.net) joined #vserver. [14:18] I've posted the bug report [14:19] and mentioned the problem it causes, that you have to give CAP_NET_ADMIN if you want to use chbind inside a vserver [14:20] so I have another question: how can I enable non-root users the chbind command, without getting operation not permitted messages [14:20] : [14:20] :) [14:31] later [14:31] bye [14:31] Filther (Filther@dial062077235063.vnet.hu) left irc: Quit: Leaving [14:37] Nick change: Bertl_oO -> Bertl [14:39] mids (mids@mids.student.utwente.nl) joined #vserver. [14:39] hi mids! [15:36] vat (vat@pD9E374F6.dip0.t-ipconnect.de) joined #vserver. [15:36] hoi. [15:37] hoi! [15:38] question. got problem with a courier-imapd of a customer.. it won't work on port 143..144 works.. nothing else is on port 143, no other customer on *:143... [15:38] any idea? ;) [15:39] what does it report on startup? [15:39] Starting Courier IMAP server:bind: Address already in use [15:39] ll_daemon_start: Resource temporarily unavailable [15:39] weired. nothing is listening on 143... [15:39] you are sure? [15:40] yep. very. else netstat would report anything ;) [15:40] .oO(netstat -a, netstat -l, etc.) [15:40] okay, kernel/patch version? [15:42] okay, have to leave now for about 30-40mins ... cu later ... [15:42] Nick change: Bertl -> Bertl_oO [15:43] 2.4.22ctx17c.. on another hostsystem work without problem, don't think of a vserver port, while 144 is working also. [15:58] vat: what does "chcontext --ctx 1 netstat -lpn | grep :143" say in the root vserver? [16:13] kestrel (athomas@home.swapoff.org) joined #vserver. [16:13] re [16:27] tcp 0 0 0.0.0.0:143 0.0.0.0:* LISTEN 28542/inetd [16:27] ah. [16:27] there is it. [16:28] mmmkay [16:39] :) [16:40] [13:40:34] weired. nothing is listening on 143... <-- ;) [16:44] ah :) [16:45] vat (vat@pD9E374F6.dip0.t-ipconnect.de) left irc: Ping timeout: 492 seconds [17:31] minus273 (~minus273@orange.mnet.bg) joined #vserver. [17:31] hello, guyz [17:34] 'lo [17:37] I'm currently building "linux-2.4.25-pre7-ow1-vs1.3.6". :) [17:38] is this with openwall patches? [17:41] Yeah. [17:41] fs/fs.o: In function `proc_virtual_readdir': [17:41] fs/fs.o(.text+0x23a38): undefined reference to `__cmpdi2' [17:41] fs/fs.o(.text+0x23a4f): undefined reference to `__cmpdi2' [17:41] fs/fs.o(.text+0x23a62): undefined reference to `__cmpdi2' [17:41] fs/fs.o(.text+0x23a7a): undefined reference to `__cmpdi2' [17:41] fs/fs.o(.text+0x23a91): undefined reference to `__cmpdi2' [17:41] Bah. :) [17:43] Probably related to my old woody compiler/linker. [17:43] I'll try my sarge box. [17:51] anyone with per context disk limits? [17:51] I mean with experience ;) [17:51] Nick change: Bertl_oO -> Bertl [17:52] hi! [17:52] herbert! ;) [17:52] hi 0.15K [17:52] may I ask a question (which has already been asked)? ;) [17:52] hmm, okay, but only once! [17:52] ;) [17:52] okay, here I go [17:53] I've setup a vserver with vs1.24 patches [17:53] and added patch-2.4.24-vs1.24-q0.12.diff.bz2 [17:53] but cannot activate the per context disk limits [17:54] I have tagctx on the partition in question [17:54] vrsetup is v0.14 [17:54] cq-tools 0.06 [17:54] Bertl: in case it matters, 2.4.25-pre7-vs1.3.6 does not compile with Debian Woody (gcc-2.95.4 and old linker). Debian Sarge (gcc-3.3.2 and stuff) does compile. [17:55] hmm, interesting ... [17:55] ehm, I get -> adding quota hash for /dev/vroot/0 ... failed: Function not implemented [17:55] did you use tagctx? [17:55] as mount option? [17:55] yes, but I'm using it on the same partition as root server [17:55] maybe this is a problem? [17:55] where did you specify that mount option? [17:55] in fstab [17:56] you know that fstab isn't available when root is mounted? 8-) [17:56] how's that? [17:56] chicken egg issue? [17:56] so, I need it on a separate partition? [17:56] fstab is on root fs, root fs has to be mounted, got it? [17:57] so, as a kernel option? [17:57] you can specify the flags on the kernel command line ... [17:57] yes, I got that [17:57] but I would not suggest to do that ... [17:57] thank you, I'll try it immediately [17:57] am I talking to myself? [17:57] why? [17:57] no, sorry [17:57] I'm listening [17:57] because you will be able to start 1 server if the root partition is tagctx ... [17:58] any further server start will fail ... [17:58] that is an issue with the userspace tools ... [17:58] they write to /var/run/vservers from within the context [17:58] this migrates the /var/run/vservers to that context ... [17:59] which on the second invocation gives a permission denied ... [17:59] I see [17:59] and is there any solution to this? [17:59] yes, use a separate partition ;) [17:59] yup [18:00] as a hack solution, if you insist, it would be sufficient to overmount /var/run ... [18:00] thank you for making this clear to me [18:00] with --bind? [18:00] with for example tmpfs or something like that ... [18:00] ahm, I see [18:01] or loop mount ... [18:01] but I'd suggest to split some partition for vserver use ... [18:01] and only mount that with tagctx ... [18:01] which would be easier, anyway [18:01] I see [18:01] Zoiah: what is the problem, with the kernel? [18:02] thank you, and bye for now [18:02] by [18:02] Bertl: it doesn't compile with the latest debian woody, but that's not too much of a problem, I'll just build it on sarge (since that'll be the new stable version soon). [18:03] hmm, and the vanilla 2.4.25-preX compiles? [18:03] Bertl: yeah. [18:03] [15:42:11] fs/fs.o: In function `proc_virtual_readdir': [18:03] [15:42:11] fs/fs.o(.text+0x23a38): undefined reference to `__cmpdi2' [18:03] [15:42:11] fs/fs.o(.text+0x23a4f): undefined reference to `__cmpdi2' [18:03] [15:42:11] fs/fs.o(.text+0x23a62): undefined reference to `__cmpdi2' [18:03] [15:42:11] fs/fs.o(.text+0x23a7a): undefined reference to `__cmpdi2' [18:03] [15:42:11] fs/fs.o(.text+0x23a91): undefined reference to `__cmpdi2' [18:03] That proc_virtual_readdir was added in 1.3.5 I think. [18:03] hmm, I see ... that is that div stuff ... [18:03] yeah, correct ... [18:04] but it looks like a compiler bug, funny thing ... [18:04] Well, it does work with debian sarge (testing) so don't bothert too much about it if you can spend your time on better/more interesting things. :) [18:04] minus273 (~minus273@orange.mnet.bg) left irc: Quit: Leaving [18:04] How's the IPv6 support coming btw? [18:05] hmm, did nothing on that ... [18:05] Ok, because I saw some files getting in patched in the ipv6 subdirs. [18:06] But I guess it doesn't do any mangling or anything yet. [18:06] just some fixes ... [18:06] but it is still on my todo list ... [18:07] Well, if you ever get round to it, I'm very interested in testing anything ipv6-related you produce. :) [18:13] Bertl: should I get util-vserver-0.27.199 or 0.28.190? [18:13] hmm, good question, last I tested was 0.27.199 ... [18:14] best to ask enrico ... [18:14] 0.28.190 is newer, so I'lll try that. ;) [18:15] yeah, I still don't understand his numebring .. he explained it to me some time ago, but it didn't help much ;) [18:16] Bah, 0.28.190 doesn't compile on Debian Woody either, let's try Sarge. [18:17] mhepp (~mhepp@r72s22p13.home.nbox.cz) joined #vserver. [18:18] Yup, fine on sarge again... blah. :) [18:21] frz (~frz@213.235.213.90) joined #vserver. [18:22] frz (~frz@213.235.213.90) left irc: Remote host closed the connection [18:22] frz (~frz@213.235.213.90) joined #vserver. [18:22] hello [18:22] mhepp (~mhepp@r72s22p13.home.nbox.cz) left irc: Quit: Tak ja padaaaaM [18:23] hi frz! [18:23] hi bertl [18:23] long time not seen [18:23] yup! [18:24] working very much with vservers and all looks fine, but now i experienced problems using simple quotas ;/ [18:24] how so? [18:26] i just want simple quotas for extra partition for vservers, added actual patches [18:27] hmm, what patches? [18:27] do i need this qqtools to work? [18:27] well, what did you patch into your kernel ;) [18:28] sorry a bit slow today [18:28] added pathches for vserver 1.24 and quotas 1.24 (latest addons) [18:29] hmm, what _are_ simple quotas? [18:29] ;) [18:29] disk limits , or per context quota? [18:30] dont need to limit vserver capacity only quotas for users [18:30] vservers are running in extra partition [18:30] okay, and your vservers are on a shared partiton? [18:31] or has each vserver a separate partition? [18:32] just like this an vservers are mounted under /var/lib/vservers/.. [18:32] mount /dev/vg/vs1 /var/lib/vservers/vs1 [18:32] must quota turned on on root host for examle partition 4t2 [18:32] mount /dev/vg/vs2 /var/lib/vservers/vs2 [18:33] and ctxtag? [18:33] or mount /dev/disk /var/lib/vservers [18:33] ? [18:33] which one do you use? [18:33] currently [18:33] on partition /var/lib/vservers/4t2 vserver 4t2 is running [18:34] so you have a separate partition for each vserver, right? [18:34] yes [18:34] good, in that case, you do not need any quota patches ... [18:35] you can just leave them out, and use the vroot device to protect the block device ... [18:35] ok, and how this is done? [18:35] you know losetup ? [18:35] for loopback? [18:35] yup? [18:36] you know how to use it? [18:36] i dont really know? [18:36] how to use it for quotas [18:36] hmm, okay there is a tool called vrsetup, which is very similar to losetup ;) [18:37] so you do vrsetup /dev/vroot/1 /dev/vg/vs1 [18:37] yes already compiled and used to make /dev/vroot/4t2 /dev/evms/lvm/vserver/4t2 [18:37] yes this worked [18:38] okay, that is all you need for that ... [18:38] copy the /dev/vroot/4t2 into the vserver as /dev/hdv1 [18:38] but after thsi copied to .../vservers/4t2/dev/hdv1 [18:38] yes and now comes the problem [18:38] and modify the /etc/mtab to use 'ufs' as filesystem and contain a line with usrquota,grpquota [18:39] (inside the vserver) [18:39] ~# quotaon -vaug quotaon: using //aquota.group on /dev/hdv1 [/]: No such process quotaon: Quota format not supported in kernel. quotaon: using //aquota.user on /dev/hdv1 [/]: No such process quotaon: Quota format not supported in kernel. [18:40] but i make it once more, i minute [18:40] nope! [18:40] that is _because_ you ahve the quota patches ... [18:40] for this 'simple' way of quota, which I would suggest to use, you must not have the quota patches installed ... [18:41] Bertl: 0.28.190 does not contain vproc. [18:42] whats about the name for vrootdevice? can i use vservers name [18:42] and do i have to turn quota on at rootserver? [18:42] or only at vserver? [18:45] serving- (~serving@213.186.189.107) left irc: Ping timeout: 480 seconds [18:55] # vserver-stat [18:55] vc_get_task_xid(1): Function not implementedvc_get_task_xid(2): Function not implementedvc_get_task_xid(3): Function not implementedvc_get_task_xid(4): Function not implementedvc_get_task_xid(5): Function not implementedvc_get_task_xid(6): Function not implementedvc_get_task_xid(7): Function not implementedvc_get_task_xid(8): Function not implementedvc_get_task_xid(9): Function not implementedvc_get_task_xid(10): Function not implementedvc_get_task_xid(1 [18:55] I'm probably doing something wrong. :) [18:59] quotaon -vaug quotaon: using //aquota.group on /dev/hdv1 [/]: No such device or address quotaon: using //aquota.user on /dev/hdv1 [/]: No such device or address [18:59] ? without tagctx normal mounted vservers root i get this [18:59] Hmm, and 0.27.199 vserver-stat just spams me with output... [19:02] bad thing [19:04] not really help here ;/ - already asked at mailinglist but without result [19:06] miller7 (none@213.239.180.106) joined #vserver. [19:07] hi guys [19:07] Hiya [19:07] Who's enrico btw? [19:07] hey Zoiah [19:07] he's not in [19:08] Ahh, ok, because I can't get either util-vserver-0.27.199 or util-vserver-0.28.190 to work. [19:08] sorry lost the connection ... [19:08] hey bert [19:09] frz still here? [19:09] hi miller7! [19:10] Xirzon (~Xirzon@pD9E76966.dip.t-dialin.net) joined #vserver. [19:12] Can vserver scale CPU load dynamically like Virtuozzo? That is, under heavy load, does each user get only a fixed limit of CPU share? [19:12] Xirzon: no [19:12] hmm, well that is not correct, per se ... [19:17] okay, have to leave now ... cu later ... [19:17] Nick change: Bertl -> Bertl_oO [19:18] :) [19:18] maybe later [19:30] Bertl: I know you're gone, but if I compile util-vserver-0.28.180 WITHOUT v13 api, it seems to work fine. [19:30] Filther (Filther@nilus-745.adsl.datanet.hu) joined #vserver. [19:30] hi [19:30] anyone here? [19:31] At times, yes. [19:31] no [19:31] :) [19:31] okay [19:31] :) [19:31] how can we help you Filther? :-) [19:31] quiz: how to run chbind from a non-rott account [19:31] ;> [19:32] Filther: can't? [19:32] well, can't [19:32] says I don't have privileges to do so [19:32] (first I have to give the vserver net_admin, of course...) [19:34] so, ideas? [19:35] Filther: afaik, you just can't. [19:35] why do you want to do so? [19:37] miller7: I need to do so. [19:37] yeah I asked out of curiosity [19:37] Filther: it's clear you need that, but why? [19:37] I didn't think you'd do it just cause the TV has nothing good :) [19:38] does it matter, regarding whether I can or can't do it? [19:38] ok man [19:38] sorry for asking [19:39] Filther: maybe we can find a solution that does work if you tell us what you're trying to do? [19:39] I'm trying to chbind a server in the vserver's ip, to enable the broadcast address [19:39] (because the vserver package contains a bug that I've discussed with Herbert yesterday) [19:40] And why do you need to run that chbind as a user? [19:41] if you must know: I don't want to give root privileges to the guy running the server inside the vserver [19:42] so? [19:42] Have you tried it? [19:42] yes :> [19:43] zoiah@matryoshka:~$ /chbind --ip 81.17.52.135 bash [19:43] Can't set the ipv4 root (Operation not permitted) [19:44] Ahh, hmm... but I can't use it at all in my vserver... :) [19:44] as root you could [19:44] but that doesn't matter now [19:44] matryoshka:~# /chbind --ip 81.17.52.135 bash [19:44] Can't set the ipv4 root (Operation not permitted) [19:45] But I don't have cap_net_admin and stuff. :) [19:45] ... [20:19] arekm (misiek@ikar.t17.ds.pwr.wroc.pl) joined #vserver. [20:19] arekm_ (misiek@ikar.t17.ds.pwr.wroc.pl) left irc: Ping timeout: 492 seconds [20:45] serving (~serving@213.186.189.107) joined #vserver. [20:52] Nick change: cgone -> cdub [21:21] ensc (~ircensc@ultra.csn.tu-chemnitz.de) joined #vserver. [21:21] hi [21:21] <_shuri> hi [21:31] hello again, anybody can help? after adding vrsetup and /dev/vroot/device copied to vserver /dev/hdv1 i get this errormessage: quotaon -vaug quotaon: Cannot find quota file on / [/dev/hdv1] to turn quotas on/off. quotaon: Cannot find quota file on / [/dev/hdv1] to turn quotas on/off. [21:32] what could be wrong, anybody had the same problem - really would need working quotas for vservers on separate partition [21:34] i am not really quota expert... but this seems not ok! [21:42] _shuri (~shushushu@vserver.electronicbox.net) got netsplit. [21:42] Filther (Filther@nilus-745.adsl.datanet.hu) got netsplit. [21:42] frz (~frz@213.235.213.90) got netsplit. [21:42] mids (mids@mids.student.utwente.nl) got netsplit. [21:42] meebey (meebey@meebey.net) got netsplit. [21:42] Zoiah (Zoiah@matryoshka.zoiah.net) got netsplit. [21:42] deadguy (deadguy@bananajoe.big.du.se) got netsplit. [21:42] mcp (~hightower@wolk-project.de) got netsplit. [21:42] virtuoso (~shisha@ip114-115.adsl.wplus.ru) got netsplit. [21:42] talon (talon@host-63-149-223-100.irwinresearch.com) got netsplit. [21:42] lp (~lpressl@interner.SerNet.DE) got netsplit. [21:42] cdub (~chrisw@fw.osdl.org) got netsplit. [21:42] ensc (~ircensc@ultra.csn.tu-chemnitz.de) got netsplit. [21:42] arekm (misiek@ikar.t17.ds.pwr.wroc.pl) got netsplit. [21:42] Xirzon (~Xirzon@pD9E76966.dip.t-dialin.net) got netsplit. [21:42] miller7 (none@213.239.180.106) got netsplit. [21:42] stupidawy (foo@198.77.239.131) got netsplit. [21:42] riel (~riel@riel.netop.oftc.net) got netsplit. [21:43] ensc (~ircensc@ultra.csn.tu-chemnitz.de) returned to #vserver. [21:43] arekm (misiek@ikar.t17.ds.pwr.wroc.pl) returned to #vserver. [21:43] Filther (Filther@nilus-745.adsl.datanet.hu) returned to #vserver. [21:43] Xirzon (~Xirzon@pD9E76966.dip.t-dialin.net) returned to #vserver. [21:43] miller7 (none@213.239.180.106) returned to #vserver. [21:43] frz (~frz@213.235.213.90) returned to #vserver. [21:43] mids (mids@mids.student.utwente.nl) returned to #vserver. [21:43] meebey (meebey@meebey.net) returned to #vserver. [21:43] stupidawy (foo@198.77.239.131) returned to #vserver. [21:43] riel (~riel@riel.netop.oftc.net) returned to #vserver. [21:43] Zoiah (Zoiah@matryoshka.zoiah.net) returned to #vserver. [21:43] deadguy (deadguy@bananajoe.big.du.se) returned to #vserver. [21:43] mcp (~hightower@wolk-project.de) returned to #vserver. [21:43] virtuoso (~shisha@ip114-115.adsl.wplus.ru) returned to #vserver. [21:43] talon (talon@host-63-149-223-100.irwinresearch.com) returned to #vserver. [21:43] _shuri (~shushushu@vserver.electronicbox.net) returned to #vserver. [21:43] cdub (~chrisw@fw.osdl.org) returned to #vserver. [21:43] lp (~lpressl@interner.SerNet.DE) returned to #vserver. [21:45] hello Bertl [21:45] ;) [21:55] _shuri (~shushushu@vserver.electronicbox.net) got netsplit. [21:55] talon (talon@host-63-149-223-100.irwinresearch.com) got netsplit. [21:55] virtuoso (~shisha@ip114-115.adsl.wplus.ru) got netsplit. [21:55] deadguy (deadguy@bananajoe.big.du.se) got netsplit. [21:55] Zoiah (Zoiah@matryoshka.zoiah.net) got netsplit. [21:55] meebey (meebey@meebey.net) got netsplit. [21:55] mids (mids@mids.student.utwente.nl) got netsplit. [21:55] frz (~frz@213.235.213.90) got netsplit. [21:55] Filther (Filther@nilus-745.adsl.datanet.hu) got netsplit. [21:55] lp (~lpressl@interner.SerNet.DE) got netsplit. [21:55] cdub (~chrisw@fw.osdl.org) got netsplit. [21:55] mcp (~hightower@wolk-project.de) got netsplit. [21:55] stupidawy (foo@198.77.239.131) got netsplit. [21:55] miller7 (none@213.239.180.106) got netsplit. [21:55] Xirzon (~Xirzon@pD9E76966.dip.t-dialin.net) got netsplit. [21:55] arekm (misiek@ikar.t17.ds.pwr.wroc.pl) got netsplit. [21:55] ensc (~ircensc@ultra.csn.tu-chemnitz.de) got netsplit. [21:55] riel (~riel@riel.netop.oftc.net) got netsplit. [21:56] ensc (~ircensc@ultra.csn.tu-chemnitz.de) returned to #vserver. [21:56] arekm (misiek@ikar.t17.ds.pwr.wroc.pl) returned to #vserver. [21:56] Filther (Filther@nilus-745.adsl.datanet.hu) returned to #vserver. [21:56] Xirzon (~Xirzon@pD9E76966.dip.t-dialin.net) returned to #vserver. [21:56] miller7 (none@213.239.180.106) returned to #vserver. [21:56] frz (~frz@213.235.213.90) returned to #vserver. [21:56] mids (mids@mids.student.utwente.nl) returned to #vserver. [21:56] meebey (meebey@meebey.net) returned to #vserver. [21:56] stupidawy (foo@198.77.239.131) returned to #vserver. [21:56] riel (~riel@riel.netop.oftc.net) returned to #vserver. [21:56] Zoiah (Zoiah@matryoshka.zoiah.net) returned to #vserver. [21:56] deadguy (deadguy@bananajoe.big.du.se) returned to #vserver. [21:56] mcp (~hightower@wolk-project.de) returned to #vserver. [21:56] virtuoso (~shisha@ip114-115.adsl.wplus.ru) returned to #vserver. [21:56] talon (talon@host-63-149-223-100.irwinresearch.com) returned to #vserver. [21:56] _shuri (~shushushu@vserver.electronicbox.net) returned to #vserver. [21:56] cdub (~chrisw@fw.osdl.org) returned to #vserver. [21:56] lp (~lpressl@interner.SerNet.DE) returned to #vserver. [22:17] _shuri (~shushushu@vserver.electronicbox.net) got netsplit. [22:17] lp (~lpressl@interner.SerNet.DE) got netsplit. [22:17] talon (talon@host-63-149-223-100.irwinresearch.com) got netsplit. [22:17] virtuoso (~shisha@ip114-115.adsl.wplus.ru) got netsplit. [22:17] deadguy (deadguy@bananajoe.big.du.se) got netsplit. [22:17] mids (mids@mids.student.utwente.nl) got netsplit. [22:17] frz (~frz@213.235.213.90) got netsplit. [22:17] Filther (Filther@nilus-745.adsl.datanet.hu) got netsplit. [22:17] meebey (meebey@meebey.net) got netsplit. [22:17] cdub (~chrisw@fw.osdl.org) got netsplit. [22:17] Zoiah (Zoiah@matryoshka.zoiah.net) got netsplit. [22:17] mcp (~hightower@wolk-project.de) got netsplit. [22:17] stupidawy (foo@198.77.239.131) got netsplit. [22:17] miller7 (none@213.239.180.106) got netsplit. [22:17] Xirzon (~Xirzon@pD9E76966.dip.t-dialin.net) got netsplit. [22:17] arekm (misiek@ikar.t17.ds.pwr.wroc.pl) got netsplit. [22:17] ensc (~ircensc@ultra.csn.tu-chemnitz.de) got netsplit. [22:17] riel (~riel@riel.netop.oftc.net) got netsplit. [22:17] ensc (~ircensc@ultra.csn.tu-chemnitz.de) returned to #vserver. [22:17] arekm (misiek@ikar.t17.ds.pwr.wroc.pl) returned to #vserver. [22:17] Filther (Filther@nilus-745.adsl.datanet.hu) returned to #vserver. [22:17] Xirzon (~Xirzon@pD9E76966.dip.t-dialin.net) returned to #vserver. [22:17] miller7 (none@213.239.180.106) returned to #vserver. [22:17] frz (~frz@213.235.213.90) returned to #vserver. [22:17] mids (mids@mids.student.utwente.nl) returned to #vserver. [22:17] meebey (meebey@meebey.net) returned to #vserver. [22:17] stupidawy (foo@198.77.239.131) returned to #vserver. [22:17] riel (~riel@riel.netop.oftc.net) returned to #vserver. [22:17] Zoiah (Zoiah@matryoshka.zoiah.net) returned to #vserver. [22:17] deadguy (deadguy@bananajoe.big.du.se) returned to #vserver. [22:17] mcp (~hightower@wolk-project.de) returned to #vserver. [22:17] virtuoso (~shisha@ip114-115.adsl.wplus.ru) returned to #vserver. [22:17] talon (talon@host-63-149-223-100.irwinresearch.com) returned to #vserver. [22:17] _shuri (~shushushu@vserver.electronicbox.net) returned to #vserver. [22:17] cdub (~chrisw@fw.osdl.org) returned to #vserver. [22:17] lp (~lpressl@interner.SerNet.DE) returned to #vserver. [22:26] miller7 (none@213.239.180.106) left #vserver. [22:34] __shuri (~shushushu@vserver.electronicbox.net) joined #vserver. [22:35] noel (~noel@pD9FFA553.dip.t-dialin.net) joined #vserver. [22:36] _shuri (~shushushu@vserver.electronicbox.net) left irc: Read error: No route to host [22:36] Nick change: __shuri -> _shuri [22:55] *yawn* [22:55] :) [23:20] Doener_zZz (~doener@pD9E12D35.dip.t-dialin.net) joined #vserver. [23:25] serving (~serving@213.186.189.107) left irc: Ping timeout: 480 seconds [23:28] Doener (~doener@pD9588172.dip.t-dialin.net) left irc: Ping timeout: 480 seconds [00:00] --- Tue Feb 3 2004