[00:12] kyencer: okay, will ned a little time to test this, but if you are interested you could do some testing with vs0.09.6 (which doesn't include the memfake yet but is the abse for the next patch) [00:12] s/abse/base/ [00:17] http://vserver.13thfloor.at/Experimental/delta-2.6.3-vs0.09-vs0.09.6.diff [00:21] ILON (~ILON@65.124.79.172) joined #vserver. [00:21] hi ILON! [00:21] Hello [00:21] Topic changed on #vserver by Bertl!~herbert@MAIL.13thfloor.at: http://linux-vserver.org/ || latest stable 1.26, devel 1.3.7, exp 0.09.6 [00:22] I have been playing with vserver and though I would stop by and see what is going on [00:22] and, what do you see? ;) [00:23] ;) [00:24] Where is the best place to get help? Here or the mailing list? [00:24] quick help, here, detailed info maybe ml ... [00:24] k [00:24] do you need help? [00:24] I will. I am sure. [00:25] I want to put together vlist for hosting. just trying to make sense of the docs. [00:25] vlist? [00:26] there are many differing methods for working with vserver [00:26] sorry.. vserver [00:26] ah, okay ... [00:27] linux-vserver is very flexible ... [00:27] There does not seem to be any official detailed installation guide or procedures [00:28] well, there are, but most docs are outdated ... [00:28] Yea.. [00:28] is there a forum ? [00:29] but the support on the channel seems good, at least what I heard ... [00:29] forum like 'post messages to?' nope that function has the ml [00:30] gothca [00:55] _shur1 (~shushushu@vserver.electronicbox.net) left irc: Ping timeout: 480 seconds [01:00] monrad (~monrad@213083190235.sonofon.dk) left irc: Remote host closed the connection [01:03] _shur1 (~shushushu@vserver.electronicbox.net) joined #vserver. [01:19] bengrimm (~ben@bengrimm-host225.dsl.visi.com) left irc: Quit: Trillian (http://www.ceruleanstudios.com) [02:03] micah (micah@adsl-68-78-111-196.dsl.emhril.ameritech.net) joined #vserver. [02:03] hi micah! [02:04] howdy! [02:04] I'm having a weird problem [02:04] vgdisplay /mnt [02:04] vgdisplay -- ERROR: VGDA in kernel and lvmtab are NOT consistent; please run vgscan [02:04] :p [02:04] anyone know how to fix that? [02:04] vgscan ? [02:04] tried that [02:04] and? [02:04] it seems to work, but then vgdisplay still says that [02:04] you are trying this in a vserver? [02:05] no, this is actually outside [02:05] did you remove your /dev/* entries (maybe accidentially)? [02:06] I dont think so [02:06] there are lots of /dev entries :) [02:06] okay, what does vgscan report? [02:07] it says it finds the active volume groups and that: [02:07] vgscan -- "/etc/lvmtab" and "/etc/lvmtab.d" successfully created [02:07] vgscan -- WARNING: This program does not do a VGDA backup of your volume groups [02:07] and how about lvscan? [02:07] gives me the same error: [02:07] lvscan -- ERROR: VGDA in kernel and lvmtab are NOT consistent; please run vgscan [02:08] hmm, okay, then let's check pvscan ... [02:08] pvscan looks good [02:08] okay, then try the vgscan/lvscan again [02:08] same thing :p [02:09] vgscan looks fine, but lvscan is not happy [02:10] I can take one of the vg's offline maybe? [02:10] it is not being used [02:11] surriel (~riel@imladris.surriel.com) joined #vserver. [02:11] surriel (~riel@imladris.surriel.com) left irc: Remote host closed the connection [02:14] lvmdiskscan looks fine too [02:14] but lvscan gives that same error [02:14] hmm, sounds strange ... did you try a reboot? [02:14] I'm afraid to ;P [02:15] but I am tryingthat now :) [02:19] arg [02:19] still a no go [02:19] it rebooted fine, but vgdisplay still has that error [02:20] hmm, strange ... [02:20] what did/does the pvscan report btw? [02:21] pvscan is normal [02:22] yeah, what does it report? [02:22] pvscan -- reading all physical volumes (this may take a while...) [02:22] pvscan -- ACTIVE PV "/dev/md2" of VG "rootvg" [5.53 GB / 3.04 GB free] [02:22] pvscan -- ACTIVE PV "/dev/sdc1" of VG "webvg01" [149.04 GB / 0 free] [02:22] pvscan -- total: 2 [154.59 GB] / in use: 2 [154.59 GB] / in no VG: 0 [0] [02:22] webvg01 is empty right now, so I would like to remove it and recreate it maybe that will fix the problem [02:23] if I can figure it out [02:23] guess this is a question for the lvm guy ... [02:24] where do I find him? [02:24] Heinz Mauelshagen [02:24] but I would try the ml [02:24] right [02:24] linux-lvm@redhat.com [02:24] everything seems to be working otherwise [02:24] do you know how to remove a VG? [02:25] I want to remove it and re-create it [02:25] vgremove [02:25] if there is no lv associated, but that will probably choke on the lvtab too [02:26] yeah, it vgremove gives me that error [02:26] there is a lv associated [02:27] but lvremove gives the same reason [02:28] that's what I meant ... [02:28] I mean if you do not care about the data on /dev/sdc1 [02:28] I dont [02:28] its totally empty [02:28] you could try to zeroing out the disk with dd [02:29] at least the first 8k or so ... [02:29] hm, how do I do that? [02:29] rebooting or rerunning pvscan should update this then ... [02:29] dd if=/dev/zero of=/dev/sdc1 bs=1k [02:29] let it run for 10 secs and abort it with CTRL-C [02:29] running [02:30] ok, now what? [02:30] run pvscan again [02:30] no webvg01 anymore :) [02:30] now run the vgscan [02:30] and then the lvscan [02:30] lvscan -- ERROR: VGDA in kernel and lvmtab are NOT consistent; please run vgscan [02:30] :( [02:31] okay, something different to try, move the /etc/lv*tab file/dir somewhere else ... [02:31] and start with pvscan ... [02:32] mv /etc/lvmtab /tmp [02:32] pvscan [02:32] vgscan [02:32] vgdisplay [02:32] vgdisplay -- ERROR: VGDA in kernel and lvmtab are NOT consistent; please run vgscan [02:32] :( [02:32] surriel (~riel@imladris.surriel.com) joined #vserver. [02:32] surriel (~riel@imladris.surriel.com) left irc: Remote host closed the connection [02:33] the only time I had this (or something similar ;), was as the userspace tools did not see all the devices ... [02:34] did you experiment with lvm2/ device mapper? [02:34] no [02:34] this is 2.4 with lvm1 [02:34] was just an idea ;) [02:34] no sorry, you should ask on the lkm mailing list ... [02:34] s/lkm/lvm/ [02:35] right, I'll dot hat [03:05] iprone (~iprone@216.89.233.129) left irc: [03:12] bertl: still working on alpha boxes [03:12] bertl: they are kind of a pain [03:12] hmm, how so? [03:12] bertl: my skills with the SRM console and aboot are a lil rusty [03:13] i've got a debian one up [03:13] just gotta "modernize" it [03:13] paul (~irssi@muedsli-wan081.citykom.de) left irc: Quit: leaving [03:13] great! [03:13] working on the mandrake one [03:14] a cooker install work for you? [03:14] sure ... [03:15] and I don't need the userspace stuff (X, etc) just minimal system +ssh [03:15] yeah [03:15] bertl: i like those :-) [03:15] bertl: where are you located, btw? [03:15] and if possible an urpmi setup which can download from network etc [03:15] I'm from Austria/Europe [03:16] Bertl: Kevin and I are located in Atlanta [03:16] Oual (~val@valzone.zbla.net) joined #vserver. [03:17] Nick change: Oual -> Val_ [03:17] hi Val! [03:17] hi :) [03:17] Bertl: What debian tree do you prefer? [03:17] hi all [03:17] kyencer: one that works ;) [03:17] huhu, speeking about debian ? :) [03:17] so probably sarge is [03:17] s/is/it is/ [03:18] 2.4.25 arrived as debian source package [03:18] wow, how is this possible? [03:18] did somebody from the future send a message? [03:18] :) [03:19] seriously, sounds good, so you need/want a port of 1.26/1.3.8 right? [03:20] 1.26 right [03:20] I do tarball of the rejects [03:20] Bertl: i have a stupid question, in the aboot conf, what args to i pass it in mandrake for text only install? [03:20] nope, please don't [03:20] (for val) [03:20] Bertl: s/to/do/ [03:20] Val_: please make the patch available (tor the patches) to get a debian 2.4.25-1 ;) [03:21] kyencer: usually you insert the CD and get a menu ... [03:21] there you select expert or text ... [03:21] bertl: well i'm booting from a floppy [03:21] hmm, IIRC that was always text only ... [03:21] bertl: no cdroms for this server :-) [03:22] yeah, did some network installs too ... [03:22] bertl: it's not prompting me for that though [03:22] Bertl : http://valzone.linuxfr.org/val/debian-kernel-source-2.4.25-against-patch-2.4.25-vs1.26.diff.tar.bz2 [03:23] Val_: read my messages ;) [03:23] ok :) [03:23] thank you! [03:23] wait shity term [03:23] ok [03:24] and what about 2.6.3 does debian have this? [03:24] yes they have [03:24] is the 2.6.3 vserver patch stable ? :) [03:24] :p [03:25] well, it is (at least here ;) [03:26] seriously, we'll get stable very soon (with 2.6.3) so I'd appreciate to check with debian ... [03:26] ok, i make the path to get a debian 2.4.25 source & test the last one for 2.6.3 [03:26] netrose (john877@SP2-24.207.231.2.charter-stl.com) left irc: Ping timeout: 480 seconds [03:26] and I only see you around, when they release a new kernel 8-) [03:27] when mailing list doesn't contain any links to a new patch for last debian sources, yes :) [03:31] so I've got a question about vserver [03:32] I'm setting up a webserver to provide hosting for a number of different non-profit organizations [03:32] sounds good [03:32] and I've been looking at different options to jail apache etc. and so far have only found hackish, resource intense ways of doing it [03:32] so now I'm looking at the ctx kernel patch and what vserver offers [03:33] (hopefully you are looking at the vs patch not the ctx one) [03:36] (ctx is outdated by now, the new patches are securer, and provide more features) [03:36] hmm, no [03:36] I didn't know of the vs pathc [03:36] http://www.linux-vserver.org/index.php?page=Linux-VServer [03:37] or simple www.linux-vserver.org [03:37] I guess what I am trying to figure out is... [03:38] what the best path is to take [03:38] some say to do jail, some say debootstrap apache instance, some recommend, xen/UML, but those are slow and a resource hog [03:39] we have a community webserver now, but it is all running under the www-data group and someone installed a old phpnuke with security holes [03:39] well, apache works fine inside vserver, you can jail it, and constrain the resources it uses ... [03:39] so people upload via php and get www-data shells, its a mess [03:40] Bertl : http://valzone.linuxfr.org/val/patch-linux-to-debian-2.4.25.diff.bz2 [03:40] and you can give away ether user or root shells ... [03:40] Val_: tx, this is called 2.4.25-1 ? [03:40] yes [03:41] i test 2.6.3 and do patch if needed [03:42] okay [03:42] Bertl: I guess I need to read up on vserver, I just dont really know enough about what it offers [03:43] micah: it's simple, you have 3 tools to work with: [03:43] chroot, chcontext and chbind [03:43] the first you already know, the second is a chroot for process space and the third for network space [03:44] in addition to that, we modified the folowing: [03:44] - barrier to make chroot secure [03:44] similar to grsecurity? [03:44] - fair scheduler/ token bucket scheduler (in ck,2.6) [03:45] 2.6.3 debian source patched, no rejects [03:45] micah: I don't know how grsec hardens the chroot ... we use a flag (combination of flags) to do that ... [03:45] we also added some virtualization and some limits ... basically that's it ... [03:46] Val_: okay ... you tried with 0.09? [03:46] yes [03:46] let's compile it :) [03:46] there is a 0.09-0.09.6 patch out there ... [03:46] I'd try that too ... [03:46] http://vserver.13thfloor.at/Experimental/delta-2.6.3-vs0.09-vs0.09.6.diff [03:46] ? [03:46] ok [03:47] well done [03:47] dilox (~dilox@host97-8.pool8249.interbusiness.it) joined #vserver. [03:47] hi dilox! [03:48] hi [03:48] I've got some question about vserver.. can I? [03:48] sure ... [03:49] ok.. I just patched kernel and install vserver on debian woody, it seems to work fine [03:49] Bertl: so its different from a UML instance because you don't have the overhead/resource requirements needed to install a whole new instance? [03:50] dilox: thats what I'm thinking of doing here :) [03:50] but.. [03:50] i tried to run apt-get having a private ip, it cannot connect [03:51] ExpiryJames (~james@h24-71-63-164.ok.shawcable.net) left irc: Quit: Leaving [03:51] dilox: let me sumarize this: [03:51] i set public ip... works fine! why? [03:51] sorry for bad english... i'm italian [03:51] you have a machine, with one public ip, and you setup a vserver with a private ip [03:51] right [03:52] i can ping.. but connection time out [03:52] then you try to conenct from the private ip/vserver to the outside, and think that your MASQ rules will catch this, right? [03:52] (the MASQ part is optional ;) [03:52] broo (~broo@host30-5.btbx.net) joined #vserver. [03:52] hi broo! [03:53] hi Bertl [03:53] hi broo [03:53] Bertl: It will probably be monday before i have the alpha's ready for ya [03:53] hi dilox [03:53] Bertl: sorry about that [03:53] dilox: if you want the private network ip to connect to the outside, you have to use SNAT to use the public ip for those conenctions ... [03:53] Bertl: Kevin and I are going to setup a domain for all of our stuff that were working on from now on, it will be at unixfirst.com [03:54] sounds good ... [03:54] i can ping external ip from vserver [03:54] but no connection [03:54] yes, I'll explain [03:54] ok [03:55] be patient ;-) [03:55] np [03:55] the vserver concept doesn't give you a virtual interface (for now) [03:55] ok [03:55] it uses the same interfaces and addresses present on the host like a normal server [03:55] eth0:vserv [03:55] exactly ... [03:56] so while there apply the same rules to this interface, the access to them is limited [03:56] a vserver can only access (bind/use/send/receive) on those addresses which where assigned [03:57] dilox: were you following a document on how to do this in woody? I would like to also [03:57] the icmp (which requires CAP_NET_RAW) is a different issue, as it doesn't use addresses [03:57] on a secure server CAP_NET_RAW is disabled, and ping from inside not possible [03:58] so if you want to reach the outside from a 'private' vserver, you have to do some network address translation before the packet leaves the hsot ... [03:58] s/hsot/host/ [03:59] mmm [03:59] this, for example, can be done with 'iptables -t nat -A POSTROUTING ... -SNAT --to [04:00] dilox: did you use this: http://www.linux-vserver.org/index.php?page=DebianVserver ? [04:00] micha.. i patched kernel, i follow an how to on linux-vserver.org then i used the debian-newvserver.sh [04:00] bertl i try right now [04:01] iptables -t nat -A POSTROUTING -o eth0 -j SNAT --to 1.2.3.4 [04:01] micha.. to install debian into vserver, it works very well [04:01] 1.2.3.4 is private ip of vserver right? [04:02] nope, the public ip [04:02] ok [04:03] there is also a toolset called util-vserver, which is in active development, and the alpha branch of this tool, can create various servers out of thin air ... [04:03] (just for your information) [04:03] i can't install it.. i got vserver it says broken dipendencies [04:03] ok, one last question, then I'll go try myself... I can't find it on the web -- what does a vserver actually enail? Is it a full OS installation? How is it less resource intensive than a UML? [04:04] you might want to put a --source on it as well to prevent outside people bouncing packets off of your SNAT rule [04:04] yeah, I know this happens on woody ... [04:05] bertl sorry.. perhaps i didn't understand.. doesn'works [04:05] micah: simple, how does UML access files? [04:05] Bertl: well, it is a totally new OS basically [04:06] okay, so an open/read does what? [04:06] eth0 is the network public interface [04:06] right? [04:06] dilox: yes [04:06] ok [04:06] dilox: did you add the above rule? [04:06] Bertl: I dont know... I know that you have to have created separate filesystems, and install a whole new debian [04:06] micah: okay, I'll explain: [04:07] the userspace process in uml want's to read a file ... [04:07] it accesses the UML kernel, which in turn accesses the filesystem [04:07] this filesystem is a file on the host [04:07] so the uml calls the host kernel [04:07] which accesses this file for the uml process [04:08] heh [04:08] after that, the data traverses the same way back [04:08] now we have something called inode cache [04:08] this eases the 'normal' linux behaviour by caching accessed inodes [04:08] where did we encounter inodes in the example? [04:08] at two places, the file inside and the file outside [04:09] and both files require a filesystem management ,,, [04:09] so how does this look like in vserver? [04:09] arghh... i have to restart something after iptables etc... (i think no)?? [04:09] :) [04:09] dilox: did you enter this to a file or set with iptables? [04:10] micah: we get a request for the file, and the kernel satisfies it ;) [04:10] set with iptables [04:10] Bertl: I'm following you, I suppose I am wondering how the actual vserver looks, is it a couple of large files on the host system that make up the filesystems of the vserver that you then install a new OS into? [04:10] look at a vserver as a chroot installation ... [04:11] without chroot holes :) [04:11] dilox: okay, get another console/terminal [04:11] ok [04:11] start tcpdump -s 1000 -vvnei eth0 on that term [04:11] Bertl: so you only install what is needed to run a service in the vserver? [04:12] basically yes, no kernel no hardware stuff [04:12] i got eth1 on external ip [04:12] i use eth1? [04:12] right? [04:12] okay, in that case eth1, you used eth1 for the iptables too? [04:13] yes [04:13] okay [04:13] tcpdump not found [04:13] apt-get install tcpdump? [04:13] Bertl: ok, thats nice :) [04:13] yes [04:13] sorry [04:14] ares:~# iptables -L -t nat [04:14] Chain PREROUTING (policy ACCEPT) [04:14] target prot opt source destination [04:14] Chain POSTROUTING (policy ACCEPT) [04:14] target prot opt source destination [04:14] MASQUERADE all -- localnet/24 anywhere [04:14] SNAT all -- anywhere anywhere to:82.89.7.29 [04:14] SNAT all -- anywhere anywhere to:10.0.0.81 [04:14] Bertl: I'm going to give it a whirl on our dev box [04:14] SNAT all -- anywhere anywhere to:10.0.0.81 [04:14] SNAT all -- anywhere anywhere to:82.89.7.29 [04:14] SNAT all -- anywhere anywhere to:10.0.0.83 [04:14] micah: make that, and if you have any questions, feel free to ask ... [04:14] what's this? [04:14] a lot of writing [04:15] dilox: I'd say you should flush your current tables [04:15] kyencer (~kyencer@216.89.233.129) left irc: Ping timeout: 480 seconds [04:15] iptables -t nat MASQUERADE -F [04:15] iptables -t nat -F MASQUERADE [04:15] I guess that was F [04:16] iptables: No chain/target/match by that name [04:17] Bertl: where do I get the ADDON quota patch? [04:18] dilox: POSTROUTING instead of MASQUERADE [04:18] did you read the news section on linux-vserver.org? [04:18] ah I was reading; http://www.linux-vserver.org/index.php?page=Step-by-Step+Guide [04:18] GREAT it works... [04:19] I have to flush iptables first [04:19] okay ... [04:19] patch-2.4.24-vs1.24-q0.12.diff.bz2 [04:19] there was masquerading on it... [04:19] Action: broo is away: hey look over there -> [04:19] but i need [04:19] micah: http://www.13thfloor.at/vserver/s_addons/quota/ [04:19] masquerading [04:19] Bertl: got it [04:19] dilox: you can use masquerading too, but only for non vservers [04:20] ok [04:20] perfect [04:20] btw, masquerading entries belong to the FORWARD chain [04:21] broo: thanks for the help, got confused ... [04:21] ok.. i'll set a particular range of ip for vserver so i can separate the rules [04:21] thanks a lot [04:22] Bertl: what should I set the Persistent Context ID for files to? [04:22] hehe [04:22] :) [04:22] depends on what you need ... usually UID24/GID24 is a good choice [04:22] bertl.. are you part of the group that made vserver? [04:22] Bertl: not sure what I need :) [04:23] micah: some addon-info for that, you need a separate partition for the vservers if you use context tagging ... [04:23] dilox: guess I am ... [04:24] micah: and if you switch from one tagging format to the other, this might give unexpected results ... [04:24] Bertl: you need that for quotas? [04:24] depends ... [04:24] heh [04:24] if you put each vserver on a separate partition, you do not need it .. [04:25] hmm [04:25] well, I guess I dont know if I am going to do that [04:25] I probably wont [04:25] yes I think you are [04:25] if vservers share a partition, then they can even share files ... [04:25] i dont want to make a new partition for each virtual host for each vserver... [04:25] do you live in states boys? (if i can ask) [04:25] i want to limit the overkill factor :) [04:26] dilox: yes, I do [04:26] united states i mean [04:26] i'm italian.. [04:26] mesa is from Austria/Europe [04:26] what city? just courius [04:26] near Vienna ... [04:26] Bertl: so in taht case I need Persistent Context IDs, quota support, VFS v0 quota format support? [04:26] nice [04:27] micah: if you want per vserver quota or disk limits, yes [04:27] cool [04:27] making my kernel now [04:28] enable the vroot device ... for security [04:28] bertl.. i'm glad.. it's the first time somebody help me on IRC [04:28] well, it's a good atmosphere here ... [04:29] yes [04:31] Bertl: got the vroot device [04:31] where are the other people in the chat? [04:31] no talking? [04:32] dang, gotta go, will play with vserver more later :) [04:32] thanks bertl! [04:32] bye [04:32] cu [04:32] micah: np [04:33] so.. now vserver works fine, i solved ip problem... [04:33] wonderful ... [04:33] ehehe [04:34] give me some other usefull tips [04:34] please [04:34] what do you want to know? [04:35] don't know.. if i run 2 or more vserver..can i run for example 2 apache? [04:35] sure, that's the idear behind it ... [04:35] ok [04:35] but you also need two ips for that ... [04:36] ok.. i set two private ip [04:36] well, unless you want to do some load balancing ... [04:36] how can I? [04:36] it isn't very useful to balance on the same host ... [04:36] but it is possible ... [04:37] using iptables? [04:37] for example ... [04:37] or a proxy, forwarding the requests to different vservers [04:38] ok [04:38] the program i install.. tcpdump let me see the packet iptables filters? [04:39] it let's you see all packet leavin/entering an interface [04:40] ok.. infact i see my machine on ssh port 22 [04:40] ok [04:41] you can also filter what you see, (have a look at the man page ;) [04:41] i'm doing [04:41] what's the time in Us? [04:42] here is 3 am [04:42] i'm going to sleep [04:43] don't know ... have a good sleep ... cya [04:43] i'm very gratefull with you.. i hope to meet again if havin some problems [04:43] I'm sure we will ... [04:43] do you live here???? [04:44] somewhat ... [04:44] on #vserver? [04:44] yeah, I'm some advanced info bot (AI-class) living here ;) [04:44] eheh [04:44] ok.. [04:45] good night Bertl.. thanks again [04:45] you're welcome! [04:45] dilox (~dilox@host97-8.pool8249.interbusiness.it) left #vserver. [05:17] kyencer (kyencer@c-24-98-145-233.atl.client2.attbi.com) joined #vserver. [05:26] bye all :) [05:26] Val_ (~val@valzone.zbla.net) left irc: Quit: zZz [05:37] kyencer: still around? [05:59] Berlt: Yeah i'm home now [06:00] hehe, and still with vserver ... [06:01] yup [06:01] well, it took me a little longer, had to add some stuff (for the future ;) [06:01] how do you find time to do all this coding? [06:01] but I'm almost finished with the memory fake info ... [06:02] well simple, I take the time I need ... [06:02] from time to time I have to work as consultant to earn my living ... [06:02] don't we all ;-) [06:03] the rest is free to dispose ... [06:05] well i spend about 8-10 hrs a day @ eapps [06:05] 3 hours of day in class [06:05] and the rest free to dispose [06:07] i'm currently studying management at Georgia Tech [06:07] interesting studies? [06:07] not really [06:07] my passion is in computing [06:07] easy degree to obtain [06:08] then why not computer sciences? [06:08] i was more interesting the practical stuff, not too interested in theories [06:08] i was Comp Sci for about 2 years [06:09] but in order to continue my studying also, i had to start working full time [06:09] which didn't leave me the time for CS [06:09] ic [06:09] i do like business and management, but will always work in computers [06:10] well I finished my CS master, and currently working on my doctoral thesis [06:15] that's cool [06:16] kevin and i would really like to become more involved [06:16] no problem there ... [06:16] just to give you a little background our daily operations really stress out the vservers, we have 32 per a server all running either Tomcat or JBoss [06:16] hmm, SMP machines? [06:16] as eApps is in the business of Application Server hosting [06:16] yeah [06:16] Dual Xeon's [06:16] with 4 gigs of RAM [06:16] well, that is pretty good testing ... [06:17] oh yeah [06:17] hehe [06:17] we're doing a major upgrade this weekend [06:17] you should definitely post something on the ml ... [06:17] rolling all of them up to vs1.26 [06:17] last time matt was asking about SMP stability for example .. [06:18] well we've setup a VServer for unixfirst.com, so we'll be posting stuff there [06:18] good, just make sure to put a link on the wiki ;) [06:18] wilco [06:18] (and back of course ;) [06:21] well i'm off to enjoy my evening [06:21] ttyl [06:21] cya [06:21] i will have some stuff for you on monday with the ml stuff [06:21] ok great [06:21] keep me posted if any updates roll out [06:21] kyencer@unixfirst.com [06:21] will get on the ml, you should subscribe ... [06:22] i have [06:22] i'll look there for updates [06:22] okay, in this case, you are informed ;) [07:49] broo (~broo@host30-5.btbx.net) left #vserver (Client exiting). [07:53] Nick change: cdub -> cgone [10:27] ensc (~ircensc@ultra.csn.tu-chemnitz.de) left irc: Quit: Terminated with extreme prejudice - dircproxy 1.0.5 [10:27] Nick change: Bertl -> Bertl_zZ [13:06] JonB (~NoSuchUse@129.142.112.33.ip.tele2adsl.dk) joined #vserver. [14:15] monrad (~monrad@213083190235.sonofon.dk) joined #vserver. [14:44] w [14:46] z [14:48] sorry [14:53] no problemo [14:55] when i run vserver-stat what is VSZ & RSS ? [14:56] Action: monrad should learn to use man [14:56] what does the man say that they are ? [14:56] VSZ Number of pages of virtual memory [14:56] RSS Resident set size [14:57] from the debian man page [14:58] okay [16:19] Mcleod (~altec@202.9.60.199) left irc: Read error: Connection reset by peer [17:01] monrad (~monrad@213083190235.sonofon.dk) left irc: Quit: Leaving [17:30] serving (~serving@213.186.190.121) left irc: Read error: Connection reset by peer [17:55] loger joined #vserver. [17:55] yes? [17:55] Bertl: i've also been rethinking some of my old ideas [17:55] you know, intel has released the ia32e, which is basically amd64 [17:56] in linux, both is called x86-64 ... [17:56] the code is somewhat similar to i386, so adding support to QEMU shouldn't be that hard [17:57] Bertl: and noone else is already working on adding that support ? [17:57] (and it basically uses the same hardware) [17:57] dont know ... you have to google,but it doesn't look lile [17:57] s/lile/like/ [17:57] the next level of complexity, as I suggested, would be ia64 [17:58] you know, there is a simulation, but no qemu support ... [17:58] and the third level actually is what I suggested last time ... [18:00] no, the last time you suggested "add support to QEMU, add support for that faster simulation, and add support for writing targets smarter, faster, easier" [18:01] yeah, the faster simulation part is the third level ... [18:01] simple: x86-64, medium: ia64, hard: bintrans+qemu [18:02] yeah well, adding x86-64 might be too little, and after doing that adding IA64 might be looked upon as easy [18:02] i dont know [18:03] i was planning to look more into ideas and write something down about them [18:04] today [18:07] hmm, how does it come, that we do not list you in the hall of fame yet? [18:08] (I mean as providing hardware for testing) [18:08] maybe because i dont care enough for fame myself [18:08] and, are you actualy using it ? [18:09] well, I'm going to install 1.3.8 on it ... for testing ... [18:09] okay [18:10] so could you please provide me with a small text, you'd like to see there ... [18:11] [Dariush Pietrzak's Bastard Patchset] ?? what do they do ? [18:12] i cant find the url [18:12] don't know ... [18:12] looks like a kernel patchset ... based on ctx17 or older [18:12] http://eyck.forumakad.pl/Projects/bsd/ [18:13] monrad (~monrad@213083190235.sonofon.dk) joined #vserver. [18:13] found the hall of fame [18:13] hi monrad! [18:13] jes (~jes@cpc1-leed5-3-0-cust196.ldst.cable.ntl.com) joined #vserver. [18:13] hi jes! [18:13] afternoon all [18:13] heya Bertl [18:14] how's things? [18:14] fine, please help my weak mind, what did we test last time? [18:14] errrr.....well last time it was all pretty much working ;) [18:14] Mcleod[Zzz] (~altec@202.9.60.199) joined #vserver. [18:15] jes: but we where testing/adapting something, was it the x86_64 stuff? [18:15] ahhh yes [18:15] one "issue" was the fact that the lib64 directory wasn't being setup when using build legacy [18:15] Bertl: why not just mention what the machine is ? [18:16] Jon, did you see the 'Rose Hosting' part? [18:16] yes, that is in the top [18:16] something like this, the machine, of course, is a part of it ... [18:16] hmmm another quick question, I'm trimming out things from my vserver that I don't need [18:17] and in the /proc dir (of the vserver) there is kcore [18:17] which consumes 4Gb of disk space....the question is, is that really needed from within the vserver? [18:17] hehe ... that is a good one, sorry ;) [18:17] Bertl: okay [18:17] or....is it "fibbing" and it's not really consuming disk space [18:17] okay /proc is a totally virtual filesystem [18:18] and the kcore is actually the kernel memory [18:18] *nods8 [18:18] *nods* even [18:18] ok...so I don't worry about it then [18:18] tanjix (Termin4t0r@pD904A892.dip.t-dialin.net) joined #vserver. [18:18] hi together [18:18] hi tanjix [18:18] jes: you should more worry about showing this inside a vserver, which should be avoided by proc security [18:18] hi tanjix! [18:19] Bertl....but I don't think vproc works in alpha yet does it? [18:19] we talked about it the other day [18:19] jes: which kernel patches/tools do you use atm? [18:19] Action: micah is still waiting for his machine to be rebooted [18:20] [root@cerberus tmp]# ./testme.sh [18:20] Linux-VServer Test [V0.07] (C) 2003-2004 H.Poetzl [18:20] New security context is 1 [18:20] chcontext is working. [18:20] chbind is working. [18:20] Linux 2.6.3 x86_64/0.29.196/0.29.196 [Ea] [18:20] does that answer the question Bertl? [18:20] Bertl: I think I screwed up my kernel build last night :p [18:20] jes: okay, just use setattr to manipulate the /proc entries ... there is a page on the wiki describing the logic ... [18:20] micah: how so? [18:21] ty Bertl [18:21] Bertl...I do setattr from the host, or within the vserver? [18:21] Bertl: not sure, made the new kernel with the patches, installed it and then rebooted, still waiting for it to come up. [18:21] jes: host, always host [18:21] Bertl: actually, its a remote machine so I'm waiting for someone there to reboot it and fix the problem [18:21] ok, ty Bertl [18:22] micah: hmm, (s)he should have a look at the console, might give some info [18:22] if I make changes using setattr, does the vserver need to be bounced to pick them up, or will it happen automagically? [18:23] automagically, and for all vserver, but it is lost on a reboot ... [18:23] ahhh ok [18:26] woohoo....well I have my vservers serving http transparently now from my host ;) [18:27] e.g. domain1.com -> vserver1, domain2.com -> vserver2 [18:28] now I just need to get my head around LDAP ;) [18:28] Bertl: http://www.linux-vserver.org/index.php?page=Hall+of+Fame [18:29] Errr I think there's a link on there that shouldn't be there....allows you to modify the listing [18:30] JonB: hmm, well, not what I had in mind, but if you prefer it this way ... [18:30] jes: you mean the wiki? [18:31] I mean the link that JonB just posted Bertl [18:31] okay, you know what a wiki is? [18:31] Bertl: i dont have any particular preferings, you are welcome to change it to your words. [18:31] yeah, I just didn't think non-logged in people could modify? [18:31] damn i booted my remote firewall and it did not come up again [18:32] ick@monrad [18:32] how remote is "remote" monrad? [18:32] SeerHome (~theseer@c168174.adsl.hansenet.de) joined #vserver. [18:32] monrad: use the same serial/remote reset as i do [18:32] heya ;) [18:32] hiya SeerHome [18:32] well about 15 km but at my parents home [18:32] hey SeerHome! [18:32] erm.. maybe stupid question.. but how do i REMOVE a vserver? [18:32] rm -rf /vservers/foo [18:33] SeerHome: yeah [18:33] gets me a bunch of permission denieds [18:33] SeerHome: and the .conf file [18:33] SeerHome: you should stop it first ;) [18:33] so i got a remote "serial" dad he just have to be home :) [18:33] *G* [18:33] lol@monrad [18:33] oh..g ood point actually ;> [18:34] monrad...call him to say the house is on fire ;) [18:34] okay, who is interested in testing the newest 2.6.x features? [18:35] jes: well i dont know if that helps if i am going to do it again [18:35] lol monrad [18:35] true [18:35] Bertl...what features? [18:35] grmpf ;> now that was stupid.. [18:35] Action: JonB stares at his feet, and suddently remembers that he has to go shopping, so he can not test ;-P [18:35] SeerHome: you can use vkill to terminate he processes, and umount to get rid of the /proc and /dev/pts [18:36] already got rid of it *g* [18:36] JonB: you ahve enought work with thinking about your project anyway! [18:37] what features Bertl? I'm up for some testing [18:37] well, it has now a fake memory feature ... [18:37] hehe, but i do have to shopping, they close in 20 minutes [18:37] so long as it's x86-64-ified ;) [18:37] hmm i got to drive out to do it myself [18:37] jes: of course [18:38] Bertl...fake memory? [18:38] well, if you set the VM/RSS limits, it reports them in the meminfo [18:38] ahhh [18:38] (it's the first approach, so it might be buggy as hell ;) [18:39] lol [18:39] also we now have per context capabilities *TADA* [18:39] oh, I guess I should be hiding /proc/virtual too then? [18:39] jes, that is/will be automagically ... [18:39] really? mine isn't [18:39] yeah, I know ;) [18:40] lol [18:40] just remind me Bertl...quota's aren't working in alpha yet are they? [18:40] hmm, JonB did you try 2.6 on the dual yet? [18:40] jes: nope, but I'm working on that too ... [18:41] ty Bertl [18:41] 2.6 works fine on my dual Bertl [18:41] well after you helped me ;) [18:44] okay, jes, could you do a test compile on x86_64 of the latest 1.3.8rc and 2.6.3-vs0.09.7, and provide the warnings, errors? [18:45] ok Bertl will do, as soon as I can....got some users logged into that box at the mo [18:46] no test booting, just compiling ;) [18:46] ahhh ok [18:46] one sec then ;) [18:47] errr where's the 1.3.8rc? [18:47] one sec ... [18:49] will 1.3.8 work with my kernel (2.6.3) ? [18:49] or is that what the patch is for? [18:49] yeah, it's just compiling ;) [18:50] k [18:50] Bertl: no, i havent tested it yet [18:50] jes: but anyway, let's start with 2.6.3 ... [18:50] http://vserver.13thfloor.at/Experimental/delta-2.6.3-vs0.09-vs0.09.7.diff [18:50] this is ontop of vs0.09, you know where to get that? [18:51] http://www.13thfloor.at/vserver/e_patches/vs-26x/patch-2.6.3-vs0.09.diff ? [18:51] yep! [18:51] ok...I already have that diff applied I believe [18:52] Bertl: but you're welcome to put 2.6 in it [18:52] so...get a fresh kernel tree....apply the patch...then apply the delta? [18:52] yep [18:52] to simplify that [18:52] k one mo [18:52] the following procedure can be used: [18:52] tar xjf linux-2.6.3.tar.bz2 [18:53] cp -la linux-2.6.3 linux-2.6.3-vs0.09 [18:53] ( cd linux-2.6.3-vs0.09; patch -p1 <../patch-2.6.3-vs0.09.diff ) [18:53] cp -la linux-2.6.3-vs0.09 linux-2.6.3-vs0.09.7 [18:54] ( cd linux-2.6.3-vs0.09.7; patch -p1 <../delta-2.6.3-vs0.09-vs0.09.7.diff ) [18:54] ok, delta applied cleanly [18:54] making kernel now [18:56] still going Bertl.... [18:57] I didn't do a make -j3, just so you get the warnings etc in the right order [18:57] np [19:01] ok Bertl done [19:01] http://www.daedalus-solutions.co.uk/jes/output.txt [19:02] brb [19:02] okay [19:09] paul (~irssi@195.202.59.230) joined #vserver. [19:09] hi paul! [19:11] hi [19:12] tanjix (Termin4t0r@pD904A892.dip.t-dialin.net) left irc: [19:20] netrose (john877@SP2-24.207.231.2.charter-stl.com) joined #vserver. [19:20] hi bobi! [19:21] Hi [19:23] Bertl: if you wanted to change the mentioning you are welcome [19:23] okay, I'll do that on the next edit ... [19:26] serving (~serving@213.186.190.121) joined #vserver. [19:26] serving? [20:01] okay, dinner time ... [20:01] Nick change: Bertl -> Bertl_oO [20:10] micah (micah@adsl-68-78-111-196.dsl.emhril.ameritech.net) left irc: Quit: leaving [20:22] matta (matta@tektonic.net) joined #vserver. [20:23] Nick change: Bertl_oO -> Bertl [20:23] hi matta! [20:25] hi! [20:27] the fake mem is read for testing ;) [20:28] yeah? [20:28] just fakemem/vm limits now? [20:29] or did you manage to do rss? [20:29] hrm... still need to get qemu working [20:29] rss is accounted, I can add a check with a few lines, but for now it's not active ... [20:29] but it is used for the fake mem [20:29] cool [20:30] ok i'm gonna try to get this working... [20:30] what are your issues with QEMU? [20:30] the image [20:30] i build an image from one of my hosting UML images [20:30] and it gives the bios error [20:30] same command line but using their default "linux.img" boots [20:31] how did you do it? [20:31] the image [20:31] oh yeah [20:31] or just copied over? [20:31] remember, I need to fdisk/partition [20:31] forgot that was the problem [20:31] I have Virtual PC on my windows workstation [20:31] perhaps that will work :) [20:32] i just hate doing that as it's a PII-400 and windows and virtual pc are both slow [20:32] well, partitioning with fdisk isn't that ahrd ;) [20:32] it wasn't saving the sectors for me [20:33] okay, how larg is your UML disk? [20:33] +e [20:33] 10GB [20:33] is the one I made [20:33] give me a second ... [20:33] sparse though [20:34] how much do you use for vserver and for the system? [20:41] Doener_zZz (~doener@pD9588176.dip.t-dialin.net) joined #vserver. [20:43] matta: how much does your / consume? [20:43] hrm [20:43] (the system part of your UML disk) [20:43] about 800MB [20:43] so a 2GB / and 8GB /vservers should be fine [20:43] si 2GB part for system should be sufficient? [20:44] *G* [20:48] Gardener (Gard@cm79.sigma236.maxonline.com.sg) joined #vserver. [20:48] Gardener (Gard@cm79.sigma236.maxonline.com.sg) left #vserver. [20:48] Doener (~doener@pD9588C58.dip.t-dialin.net) left irc: Ping timeout: 480 seconds [20:49] matta: http://vserver.13thfloor.at/Stuff/MATTA.bs [20:49] dd if=/dev/zero of=MATTA.img bs=1M count=10240 [20:49] then ... [20:50] overwrite the first block (1k) with the MATTA.bs [20:50] of course you can also do [20:51] dd if=/dev/zero of=MATTA.img bs=1k skip=1 count=10M [21:20] . [21:21] netrose_ (john877@SP2-24.207.231.2.charter-stl.com) joined #vserver. [21:21] netrose (john877@SP2-24.207.231.2.charter-stl.com) left irc: Read error: Connection reset by peer [21:23] hmm [21:23] the install-*.sh scripts seem to be broken ;/ [21:23] or i'm to stupid to use /adobt them.. [21:24] hmm, which install scripts? [21:24] [root@core util-vserver]# ./install-fc1.sh trouble minimum [21:24] where fc1 is a modified version of the rh9.0 script [21:24] (just replaced RedHat by Fedora ) [21:24] rpmdb: /vservers/trouble/var/lib/rpm/Packages: No such file or directory [21:25] error: cannot open Packages index using db3 - No such file or directory (2) [21:25] 147 packages to install [21:25] that's the "result" [21:25] hmm, okay not using those scripts ... [21:25] then i get a trillion times [21:25] error: cannot open Packages database in /vservers/trouble/var/lib/rpm [21:25] once i ctrl-c the script, the dir /vservers/trouble is not removeable [21:26] rm -f tells me permission denied [21:26] i'm a bit confused ;> [21:26] probably the /proc and /dev/pts is already mounted ... [21:27] mount | grep trouble [21:27] -> nothing [21:28] grep trouble /proc/mounts [21:28] nothing [21:28] okay use lsof +d /vservers/trouble [21:29] lsof +d /vservers/trouble [21:29] [root@core util-vserver]# rm -rf /vservers/trouble [21:29] rm: cannot remove directory `/vservers/trouble/dev/pts': Operation not permitted [21:29] hmm, check the permissions ... [21:30] rm: cannot chdir from `/vservers/trouble/dev/.' to `null': Not a directory [21:30] very strange ;> [21:30] ls -lad /vservers /vservers/trouble [21:31] d--------- 14 root root 4096 Feb 28 19:24 /vservers [21:31] drwxr-xr-x 5 root root 4096 Feb 28 19:24 /vservers/trouble [21:31] lsattr -d /vservers /vservers/trouble [21:31] -----------t- /vservers [21:31] -----------t- /vservers/trouble [21:32] hmm, and what about /vservers/trouble/dev/pts [21:32] lsattr -d /vservers/trouble/dev/pts [21:32] -----------t- /vservers/trouble/dev/pts [21:33] i can chmod .../dev/pts to 777 [21:33] but still cannot delete it [21:33] soor_ (~as@pD951ADE1.dip.t-dialin.net) left irc: Ping timeout: 480 seconds [21:33] ls -lad /vservers/trouble/dev/pts [21:33] drwxrwxrwx 2 root root 4096 Feb 28 19:24 /vservers/trouble/dev/pts [21:33] okay please try 'umount /vservers/trouble/dev/pts' [21:33] umount: /vservers/trouble/dev/pts: not mounted [21:34] that's only logical since mount | grep trouble didn't show anything [21:34] okay, please run the testme.sh script and let me have a look at the output ... [21:34] http://vserver.13thfloor.at/Stuff/testme.sh [21:35] any parameters? [21:35] nope, just run it on the host [21:35] k.. paste the output? [21:35] i don'T want to get killed for flooding ;> [21:35] in private ... [21:36] jes (~jes@cpc1-leed5-3-0-cust196.ldst.cable.ntl.com) left irc: Quit: Leaving [21:36] hmm, okay looks good so far ... [21:36] yeah.. [21:36] did you apply the quota patches? [21:37] no [21:37] it's plain v1.26 on a kernel.org 2.4.25 kernel [21:37] okay, try the rm again on the pts and if it still fails, strace it ... and let me have a look at that [21:37] soor (~as@pD951A78C.dip.t-dialin.net) joined #vserver. [21:38] rm -rf /vservers/trouble/dev/pts [21:38] rm: cannot remove directory `/vservers/trouble/dev/pts': Operation not permitted [21:38] okay.. so strace it is.. [21:43] rmdir("/vservers/trouble/dev/pts") = -1 EPERM (Operation not permitted) [21:43] check the permissions of /vservers/trouble/dev [21:44] ls -lad /vservers/trouble/dev [21:44] drwxrwxrwx 3 root root 4096 Feb 28 19:24 /vservers/trouble/dev [21:44] and with lsattr? [21:44] lsattr -d /vservers/trouble/dev [21:44] -----------t- /vservers/trouble/dev [21:45] hmm, the t flag isn't supposed to be set for subdirectories . [21:45] try to remove the 't' flag recursively from /vservers/trouble [21:46] chattr -R -t ? [21:47] yup [21:47] okay.. that helped a lot [21:47] rm: cannot chdir from `/vservers/trouble/dev/.' to `null': Not a directory [21:47] rm: cannot chdir from `/vservers/trouble/dev/.' to `zero': Not a directory [21:47] rm: cannot chdir from `/vservers/trouble/dev/.' to `full': Not a directory [21:47] those are the only lines left.. [21:48] pwd [21:48] [root@core dev]# lsattr [21:48] lsattr: Operation not supported While reading flags on ./null [21:48] pwd==/vservers/trouble/dev [21:49] try to remove the entries in /vservers/trouble/dev manually ... [21:49] Nick change: Doener_zZz -> Doener [21:49] rm -f null [21:49] rm: cannot remove `null': Operation not permitted [21:49] [19:47] hmm, the t flag isn't supposed to be set for subdirectories . [21:49] SeerHome: and the permissions of that files [21:49] [root@core dev]# ls -lsad null [21:49] 0 crwxrwxrwx 1 root root 1, 3 Feb 28 19:24 null [21:50] strange ... [21:50] hmm... afaik +t is set for every directory being created if the parent dir has the flag set... [21:51] dunno about files... [21:51] yes, but the tools take care about removing that flag ... [21:51] but it might be that 0.28 didn't support this ... [21:51] [root@core dev]# rpm -qa | grep vserver [21:51] util-vserver-0.28-0 [21:51] it's 0.28 ;> [21:51] nope, they don't support this [21:53] at least up to 0.28, never used 0.29 [21:53] okay, we are at 0.29.2 now ;= [21:54] so i'll upgrade later today? [21:54] s/i'll/should i/ [21:54] not that this is going to fix my current problem i guess... [21:55] might be worth a try, but it doesn't explain the things you describe ... [21:55] the only stuff left are those few files [21:55] full, null, ptmx,random, tty, urandom and zero [21:56] in the /vservers/trouble/dev dir [21:56] can you unmount and remount the fs? [21:56] I mean /vservers [21:56] hmm.. no [21:56] that would mean i have to stop all other vserver [21:57] hmm, okay ... [21:57] if that's the only way i'll have to do that.. [21:57] I don#t know, never had something like this ... [21:57] me neither ;> [21:57] what does vserver-stat tell? [21:57] at first glance, I'd say the filesystem is corrupt ... [21:58] i _think_ i had something similar, but i cannot remember exactly what is was... [21:58] Doener: nothing unusual.. the "trouble" (guess i should have choosen a different name ;> ) is NOT started [21:58] i was trying to build it [21:58] using the install-*.sh scripts from /usr/lib/util-vserver [21:58] hmm... hidden easter-egg in the tools? ;) [21:59] a pretty foul egg ;> [22:01] the install* scripts are legacy stuff, left over from jacks tools ... [22:01] enricos util-vserver use a different build scheme ... [22:01] you can read about this, and try it with the alpha-tools [22:02] alpha sounds bad on a production server ;> [22:02] i can try on my home box later this weekend.. [22:04] lstat64("full", {st_mode=S_IFCHR|0777, st_rdev=makedev(1, 7), ...}) = 0 [22:04] access("full", W_OK) = 0 [22:04] unlink("full") = -1 EPERM (Operation not permitted) [22:04] i don't get it ;/ [22:05] *streeeeetch* [22:05] gooooood mornin' [22:05] eum, evening. [22:05] whatever [22:05] :) [22:11] morning click! [22:30] hmm [22:30] i just did a vserver-copy [22:30] same problem ;/ [22:30] so there defintifly is a problem somewhere [22:31] problem is, that the /vserver is part of the / fs [22:31] so i have to reboot to run fsck on it [22:31] hrm, bad setup [22:31] yeah.. i noticed [22:31] try to cp -va the ../dev/pts somewhere else ... [22:32] would be interesting to know if this is also denied [22:32] k.. wait a sec.. [22:33] no [22:33] no problem [22:33] and you can remove the copy now? [22:33] neither copy nor the rm -rf on the copied version [22:33] okay, what about moving the /dev/pts ? [22:34] mv dev/* /tmp/test [22:34] mv: can't stat source dev/* [22:35] and moving the /dev ? [22:35] that seemed to have worked [22:36] rm -rf /tmp/test [22:36] rm: cannot chdir from `/tmp/test/.' to `null': Not a directory [22:36] *lol* [22:36] what filesystem is this? [22:36] ext3 [22:38] doesn't make sense to me ... [22:41] i don't even get why it's trying to chdir tot hat file [22:47] hmm, maybe because the +t flag was inherited to the file [22:48] chattr -t doesn't change anything [22:49] hmm.. those flags aren't supported on special files... [22:54] vserver-copy's rsync failed [22:54] and left me with a bunch of .something-something files [22:54] i had to chattr -R -t those to get them removed [22:54] now the only files left again are those in dev [22:54] ;> [22:55] actually "reboot" made it to those files [22:55] I'd say, the +t is inherited to block/char devices too [22:55] but why can't i chattr that 'off' ? [22:57] don't know ... [22:58] hmm ... would you consider rebooting the server with an additional patch applied? [22:59] sure.. [22:59] but not before mondey [22:59] monday [22:59] since i cannot reach the box without making a lot of trouble if it doesn'T come back [22:59] okay ... [23:03] so how do i create a new verserver now? [23:03] *g+ [23:04] vserver-copy template test [23:04] didn't work ( rsync failed as described ) [23:04] updating to recent util-vserver (0.29 or 0.29.2) should solve that ... [23:04] k.. [23:04] available from your site? [23:04] enricos place ... [23:05] ah.. [23:05] k [23:06] http://www-user.tu-chemnitz.de/~ensc/util-vserver/util-vserver-0.29.tar.bz2 [23:06] ? [23:06] for example ... [23:06] i can't find a 0.29.2 [23:06] http://www-user.tu-chemnitz.de/~ensc/util-vserver/ [23:06] oh it's in pre [23:07] k.. getting the 0.29.2 then.. [23:23] okay.. that seems to have worked :) [23:23] I'm pretty sure now that this is an unwanted sideeffect of the +t propagation ... [23:26] sounds reasonable... [00:00] --- Sun Feb 29 2004