1239848670 Q * geb Remote host closed the connection 1239849119 Q * nenolod Remote host closed the connection 1239849147 J * nenolod nenolod@petrie.dereferenced.org 1239849260 Q * Piet Quit: Piet 1239849943 J * cehteh ~ct@pipapo.org 1239854864 J * emcepe ~mcp@wolk-project.de 1239854864 J * padde_ ~padde@patrick-nagel.net 1239854865 Q * FireEgl Remote host closed the connection 1239854932 Q * mcp Ping timeout: 480 seconds 1239854932 N * emcepe mcp 1239854941 J * meebe583 meebey@booster.qnetp.net 1239854947 J * tokkee_ tokkee@ssh.faui2k3.org 1239854949 Q * thermoman Remote host closed the connection 1239854951 J * thermoman ~thermoman@84.201.90.210 1239854957 J * n01101111x ~nox@host.noxlux.de 1239854963 Q * Marillion Remote host closed the connection 1239854964 J * Marillion ~dirk@hetzner3.127011.net 1239854965 J * DelTree_ ~deplagne@goldorak3.eric.deplagne.name 1239854990 Q * Hollow Read error: Connection reset by peer 1239854994 J * Hollow ~bene@shiva.xnull.de 1239854997 Q * meebey Read error: Connection reset by peer 1239855013 Q * nox Remote host closed the connection 1239855013 N * n01101111x nox 1239855017 Q * DelTree Read error: Connection reset by peer 1239855062 Q * tokkee Ping timeout: 480 seconds 1239855107 Q * padde Ping timeout: 480 seconds 1239855111 N * padde_ padde 1239855695 J * FireEgl FireEgl@173-16-9-10.client.mchsi.com 1239856701 Q * nenolod Quit: my eyes cannot compute this misery. 1239856714 J * nenolod nenolod@petrie.dereferenced.org 1239857112 Q * balbir_ Ping timeout: 480 seconds 1239857248 N * Bertl_zZ Bertl 1239857252 M * Bertl morning folks! 1239857417 Q * nenolod Quit: my eyes cannot compute this misery. 1239857431 J * nenolod nenolod@petrie.dereferenced.org 1239858654 M * cehteh *yawn* 1239859600 M * Bertl okay, off to do some work .. bbl 1239859604 N * Bertl Bertl_oO 1239859718 J * imcsk8 ~ichavero@189.155.84.22 1239860844 J * balbir_ ~balbir@59.145.136.1 1239861183 J * sharkjaw ~gab@149-240-82.oke2-bras6.adsl.tele2.no 1239861455 Q * imcsk8 Quit: This computer has gone to sleep 1239864128 J * thierryp ~thierry@zircon.inria.fr 1239865225 J * davidkarban ~david@193.85.217.71 1239866184 J * xdr ~xdr@62.88.128.188 1239866515 J * cga ~weechat@62.196.2.6 1239867226 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net 1239867717 Q * scientes Ping timeout: 480 seconds 1239868841 Q * sharkjaw Remote host closed the connection 1239868902 Q * thierryp Remote host closed the connection 1239869020 J * kir ~kir@swsoft-msk-nat.sw.ru 1239869070 J * sharkjaw ~gab@149-240-82.oke2-bras6.adsl.tele2.no 1239869095 J * thierryp ~thierry@zircon.inria.fr 1239869172 Q * infowolfe Ping timeout: 480 seconds 1239869228 J * infowolfe ~infowolfe@97-120-129-143.ptld.qwest.net 1239869252 M * ktwilight am trying to redirect text into guest via host, e.g. vserver exec echo "text" > /path/in/guest/file, but the redirect always end up in /path/in/host/file, is there any other way? 1239869323 J * esa bip@62.123.8.70 1239870067 J * esa` bip@62.123.8.129 1239870075 Q * esa Ping timeout: 480 seconds 1239870417 M * ghislainocfs2 is : vserver exec 'echo "text" > /path/in/guest/file' 1239870421 M * ghislainocfs2 working ? 1239870557 M * transacid yap you have to quote that. how should the shell know that you dun want that output of that command stored? 1239870600 M * ktwilight i double quoted and that didn't work 1239870602 M * ktwilight lemme try single 1239870678 M * ktwilight nope, it throws a vcontext: execvp() No such file or directory error 1239871182 M * ghislainocfs2 shell redirections are tricky really 1239871263 J * Pazzo ~ugelt@reserved-225136.rol.raiffeisen.net 1239871433 N * Bertl_oO Bertl 1239871458 M * Bertl ktwilight: lack of shell knowledge on your side, try : 1239871469 M * ktwilight :( 1239871477 M * Bertl ktwilight: vserver exec bash -c ' ...... ' 1239871487 M * ktwilight o 1239871493 M * ktwilight that's a nice alternative 1239871522 M * Bertl well, _any_ > < redirect is handled by the current shell, so no chance to do that with escaping 1239871633 N * meebe583 meebey 1239871764 M * ktwilight ok, back to drawing board :) thanks all 1239871909 Q * arthur Ping timeout: 480 seconds 1239872283 N * tokkee_ tokkee 1239872378 Q * infowolfe Read error: Operation timed out 1239872392 J * infowolfe ~infowolfe@c-76-105-242-186.hsd1.or.comcast.net 1239872852 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1239872979 J * esa bip@62.123.8.70 1239872985 Q * esa` Ping timeout: 480 seconds 1239873011 Q * thierryp Ping timeout: 480 seconds 1239873065 J * jsullivan ~jsullivan@cpe-74-75-231-235.maine.res.rr.com 1239873134 M * jsullivan Good day, everyone. 1239873148 M * jsullivan I'm afraid I'm stuck with a runaway process in a vserver . . 1239873153 M * jsullivan and cannot kill it. 1239873167 M * Bertl vkill? 1239873175 M * jsullivan Strangely, this is the second time I have seen this and it is a bit worrisome. 1239873182 M * jsullivan ah - as opposed to kill 1239873192 M * jsullivan I'll give it a try right now. 1239873215 M * Bertl well, there are certain situations where processes become unkillable 1239873231 M * Bertl (some of them might get fixed in the future though) 1239873298 M * jsullivan Yes, this is a bit strange. 1239873304 M * jsullivan When couldn't stop the process . . 1239873312 M * jsullivan I tried to shutdown the guest . . 1239873316 M * jsullivan and now I get 1239873320 M * jsullivan 40035 2 0 0 0m05s26 0m02s57 29m10s40 vdbase-compiler 1239873326 M * jsullivan from vserver-stat 1239873335 M * Bertl that looks like two processes there 1239873360 M * Bertl well, what state are they in (vps auxwww) and what is your guest setup in regard of scheduling 1239873626 M * jsullivan I show: 1239873630 M * jsullivan root 16888 40035 vdbase-compiler 0.0 0.0 0 0 ? Ds 04:50 0:00 [bash] 1239873645 M * jsullivan root 17844 40035 vdbase-compiler 0.0 0.0 0 0 ? Z 04:56 0:00 [elinks] 1239873652 M * jsullivan That's the one which started the mess. 1239873655 M * jsullivan and . . 1239873668 M * jsullivan root 18259 40035 vdbase-compiler 0.0 0.0 0 0 ? D 04:59 0:00 [rc] 1239873686 M * jsullivan vkill did not seem to do the trick. 1239873691 M * tanjix hey Bertl! 1239873733 M * jsullivan e.g., vkill -c 40035 -s KILL 16888 1239873771 M * ktwilight jsullivan, try chcontext --xid 1 -- kill 16888 1239873773 M * Bertl jsullivan: you process is in 'D' state, so waiting for a device 1239873787 M * Bertl jsullivan: check the kernel log (dmesg) for a kernel trace 1239873794 Q * bourgeau Quit: bourgeau 1239873887 M * jsullivan thanks ktwilight but it returns . . 1239873892 M * jsullivan unrecognized option --kill 1239873900 M * jsullivan ah space 1239873907 M * ktwilight :) 1239873923 M * jsullivan no such process 1239873925 M * ktwilight though what Bertl spotted is valid, D state can't be killed, even with -9 1239873960 M * ktwilight jsullivan, does vps show the process? 1239873966 M * jsullivan yes 1239873984 M * jsullivan I do not see anything seemingly significant in /var/log/dmesg 1239873992 M * jsullivan Am I looking in the right place? 1239874000 M * jsullivan Please pardon my kernel ignorance. 1239874007 M * ktwilight try vkill with chcontext, instead of kill 1239874031 M * jsullivan Operation not permitted 1239874049 M * Bertl jsullivan: check with 'dmesg' 1239874132 M * jsullivan That doess give me piles of information but I'm not sure what to do with it :( 1239874136 M * jsullivan BUG: soft lockup - CPU#1 stuck for 61s! [bash:15115] 1239874146 M * Bertl see, that is your problem there 1239874167 M * Bertl what kernel is that? 1239874172 M * jsullivan but how does one fix it without rebooting the entire host? 1239874178 M * jsullivan it is 2.6.28.7 1239874180 M * jsullivan I believe 1239874186 M * fb jsullivan: this may not make you happy, but the same happens for me without vserver even compiled in into the kernel 1239874199 M * Bertl no chance, you can change the context id and start the guest with a new one 1239874207 M * Bertl yes, that is a mainline issue 1239874219 M * jsullivan Ouch - thankfully we are in test . 1239874229 M * jsullivan but I do plan on using these with very heavily loaded servers . .. 1239874235 M * jsullivan potentially hundreds of guests. 1239874247 M * Bertl you can turn off the lockup test or make it check for a longer period of time 1239874255 M * jsullivan It would be very nasty to take them all down because someone caused such a problem. 1239874275 M * Bertl it seems to be related to the filesystem issues in 2.6.28 and should be fixed in 2.6.29 1239874283 M * Bertl (although I couldn't verify that yet) 1239874306 M * jsullivan Argh! and I chose 2.6.28 because I thought it would be a bit more reliable :) 1239874313 J * thierryp ~thierry@zircon.inria.fr 1239874333 M * jsullivan So how do I change the context id and start the guest with a new one? 1239874333 M * fb Bertl: the same issue happens to be in 2.6.27 too 1239874348 M * jsullivan Funny - that's the other time I had the problem. 1239874357 M * Bertl jsullivan: just change the entry in /etc/vservers//context to an unusued one 1239874359 M * fb (but this can also be some backport made by ubuntu kernel team) 1239874361 M * jsullivan I picked just the right two kernels! 1239874373 M * Bertl fb: nope, it already was in 2.6.26 and 2.6.27 1239874380 M * fb ah :) 1239874399 M * jsullivan I noticed the vserver patch for 2.6.29 was pre4. 1239874403 M * jsullivan That's why I avoided it. 1239874408 M * Bertl pre5 :) 1239874414 M * jsullivan Is it stable yet (as if the one I'm using is!)? 1239874435 M * Bertl nope, we aren't finished yet with basic port testing 1239874455 M * jsullivan Sorry to tie you up . . 1239874462 M * Bertl it seems to work, but e.g. tanjix already reported some 'unusual' behaviour 1239874470 M * jsullivan but what will starting the hung guest in a different context do? 1239874500 M * jsullivan that is to help kill this errant process? 1239874524 M * Bertl nothing, the process will 'hang' around till the machine gets rebooted 1239874535 M * jsullivan LOL and 1239874539 M * Bertl but the new guest will work independantly from that 1239874548 M * jsullivan OK - thanks very much. 1239874560 M * jsullivan It looks like it's a reboot for now and an upgrade as soon as 2.6.29 is stable. 1239874581 M * Bertl it doesn't consume many resources, so unless you have a real problem with that process, you can leave it as is 1239874591 M * fb jsullivan: you really don't have to worry about this hanging process, really :) 1239874598 M * jsullivan It's chewing up 100% of one of the 8 processors. 1239874616 M * jsullivan Actually . . . 1239874623 M * jsullivan it's not the hung client . . 1239874631 M * jsullivan its a host session where I tried to kill the process. 1239874646 M * jsullivan So I suppose those are really two issues. 1239874665 M * fb maybe you could try to strace it 1239874680 M * fb or gdb 1239874702 M * jsullivan from top . . 1239874706 M * jsullivan 15115 root 20 0 66060 1544 1180 R 98.4 0.0 39:28.10 bash 1239874715 M * jsullivan but I can't kill 15115 either. 1239874735 M * jsullivan That's the one killing the cpu. 1239874755 M * jsullivan Any options besides kill -9 to stop it? 1239874768 M * Bertl in which context is it? 1239874786 M * Bertl kill -9 should work on that one, because it is running 1239874794 M * jsullivan I believe it is in the host. 1239874797 M * fb Bertl: i think it's the host process 1239874798 M * jsullivan I'll double check. 1239874802 M * Bertl check with vps auxww 1239874831 M * jsullivan root 15115 0 MAIN 66.4 0.0 66060 1544 ? Rs 04:37 41:39 -bash 1239874862 M * Bertl daniel_hozac: ping? 1239874994 M * jsullivan Interestingly . . . I didn't think it would help but I had two ssh sessions open to the host. 1239875003 M * jsullivan I exited both and they both have hung on the way out . . 1239875011 M * jsullivan as if they have an open process. 1239875051 M * jsullivan So I have an immortal CPU hog! 1239875056 M * Bertl daniel_hozac: guest build fails on recent kernel + latest util-vserver .. any idea why? http://paste.linux-vserver.org/12836 1239875061 M * jsullivan The other time we saw this was running OSSEC on a vserver. 1239875067 M * jsullivan The rootkit check caused the same problem. 1239875102 M * Bertl yeah it seems that you somehow trigger a loop in the kernel code 1239875112 M * Bertl don't do that! :) 1239875141 M * Bertl okay, off for now .. bbl 1239875149 N * Bertl Bertl_oO 1239875207 M * jsullivan Alas, looks like it's time for a host reboot :( 1239875274 M * jsullivan Ouch! it will not reboot - needs to be powered off! 1239875377 M * fb jsullivan: smells like broken hardware 1239875479 M * jsullivan Nah - this is the same thing we saw with OSSEC. 1239875486 M * jsullivan It cannot stop the hung process . . 1239875495 M * jsullivan so the shutdown just sits there trying to halt the guest. 1239876017 Q * gdm Remote host closed the connection 1239876122 J * gdm ~gdm@pistol.redetoile.net 1239876902 Q * esa Ping timeout: 480 seconds 1239876904 J * esa bip@62.123.8.129 1239877243 J * bourgeau ~bourgeau@euclide.rsr.lip6.fr 1239877684 J * Piet ~piet@asteria.debian.or.at 1239877710 J * arthur ~arthur@pan.madism.org 1239877907 J * esa` bip@62.123.8.70 1239877926 Q * esa Ping timeout: 480 seconds 1239878497 Q * jsullivan Quit: using sirc version 2.211+KSIRC/1.3.12 1239878662 Q * esa` Ping timeout: 480 seconds 1239878857 J * esa bip@62.123.8.129 1239878964 M * nwmcs459 hmm is there a specific patch i should use for 2.6.29? seems there is pre5 and then testing with later updates. Does pre5 or testing have bugfixes? 1239880157 Q * xdr Ping timeout: 480 seconds 1239880562 J * pmenier ~pmenier@ACaen-152-1-50-105.w83-115.abo.wanadoo.fr 1239880883 M * pmjdebruijn nwmcs459: it's listed as "experimental" for a reason 1239880904 M * pmjdebruijn nwmcs459: pre generally means, it's not entirely complete/tested yet 1239881103 J * saulus_ ~saulus@d026020.adsl.hansenet.de 1239881379 N * Bertl_oO Bertl 1239881398 M * Bertl nwmcs459: what do you mean by 'bugfixes'? 1239881515 Q * SauLus Ping timeout: 480 seconds 1239881515 N * saulus_ SauLus 1239881586 M * fb Bertl: first attempt at yesterday's samba issue, it works perfectly here 1239881621 M * Bertl fb: means? 1239881672 M * fb Bertl: it has an option to bind to specific interfaces, but even if I select only physical interface, it still tries to talk to 127.0.0.1 when updating self-related entries with wins 1239881721 M * Bertl which shouldn't be a problem as long as it talks 'ip' (as in udp and tcp :) 1239881741 M * fb imho it's a bug in samba, in a docs they point, when you don't include lo (or 127.0.0.1) in interfaces, some utils (smbpasswd for example) will not work, because yhey use 127.0.0.1 by default 1239881758 M * Bertl (that's what the lback virtualization is for, after all) 1239881781 M * fb and you must provide which IP to use in that case 1239881850 M * fb Bertl: yeah. But for some reason it doesn't work here :) 1239881894 M * Bertl could you do a tcpdump -vvnei on the host for the 'failing' request, together with an strace -fF of the daemon? 1239882016 M * fb Bertl: afai can understand, it's nmbd talking to itself (nmbd provides wins server when enabled) 1239883800 Q * nwmcs459 Remote host closed the connection 1239884733 M * Bertl nap attack ... bbl 1239884739 N * Bertl Bertl_zZ 1239885494 J * geb ~geb@AOrleans-253-1-69-40.w92-140.abo.wanadoo.fr 1239885649 M * geb hi 1239886076 Q * balbir_ Remote host closed the connection 1239887075 P * kir Leaving. 1239887983 J * kir ~kir@swsoft-msk-nat.sw.ru 1239888416 J * wellsp ~Unknown@des77-2-78-224-216-36.fbx.proxad.net 1239888422 M * wellsp hey 1239889029 M * ktwilight 'lo wellsp 1239889254 Q * sharkjaw Quit: Leaving 1239889291 M * wellsp i have a little problem with vserver... :) 1239889332 M * wellsp i would like to do hard links to do like vunify 1239889419 M * wellsp i have linked all files, exept some configuration files in /etc and files in /var/log, and /dev 1239889461 M * wellsp but i must have forgotten something 1239889471 M * wellsp so, here is the problem : http://pastebin.com/f52e3e3a2 1239889485 M * wellsp "vlogin: openpty(): No such file or directory" 1239889505 J * ghislainocfs21 ~Ghislain@LPuteaux-151-41-11-129.w217-128.abo.wanadoo.fr 1239889546 M * wellsp i have googled, and discussions talks about /dev directory 1239889585 M * wellsp so i re-created it like this (in "base" vserver) : http://readthefuckingmanual.net/error/1633/ 1239889596 M * wellsp but it still doesn't work 1239889775 M * wellsp To summary, when i start the two vservers, i can't enter in one of them. 1239889791 M * wellsp Can someone help me to solve the problem ? 1239889902 Q * ghislainocfs2 Ping timeout: 480 seconds 1239891167 J * esa` bip@62.123.8.70 1239891182 Q * esa Ping timeout: 480 seconds 1239891666 M * ktwilight wellsp, vhashify? 1239891825 Q * Piet Quit: Piet 1239892038 M * wellsp ktwilight: i don't use it 1239892054 M * ktwilight but sounds like that's what you want? 1239892077 M * wellsp hum, yes, but without indexing files 1239892171 J * esa bip@62.123.8.129 1239892185 P * ghislainocfs21 1239892192 Q * esa` Ping timeout: 480 seconds 1239892610 M * wellsp ktwilight: do you have an idea to help me ? 1239892639 M * ktwilight no idea 1239892686 J * ghislainocfs2 ~Ghislain@LPuteaux-151-41-11-129.w217-128.abo.wanadoo.fr 1239892818 J * esa` bip@62.123.8.70 1239892836 Q * esa Ping timeout: 480 seconds 1239893274 J * Piet ~piet@asteria.debian.or.at 1239893612 M * ryker For the first time, I installed snmp on vserver guest instead of just the host. I finally got it working, but I keep getting 'No page line in /proc/stat' in the message log every 5 seconds. 1239893649 M * ryker I googled it, and it's supposedly fixed quite a few years ago, and I have the latest net-snmp 1239893656 M * ryker has anyone had this problem and fixed it? 1239893785 M * geb hum 1239893794 M * geb you want monitor the vserver or the host ? 1239894361 M * sid3windr I have the same problem ryker 1239894373 M * sid3windr but not on every machine, which makes it ExtraVague(tm) 1239894447 M * geb sid3windr, same question: you want to monitor the host or just a vserver ? 1239894542 M * geb if it is for monitoring the host, i think it is related to vprocunhide 1239894655 M * sid3windr the vserver 1239894661 M * sid3windr snmpd is running on the vserver 1239894666 M * sid3windr but there is no /proc/stat etc 1239894679 M * sid3windr also can't find /proc/swap 1239894687 M * sid3windr but I only got that again after I upgraded to debian lenny I think 1239894694 M * sid3windr they probably broke the snmpd there again :| 1239894936 M * ryker sry, i was away for a bit 1239894946 M * ryker i'm running snmpd on the vserver guest 1239894959 M * ryker the guest is where i get these messages 1239895734 M * wellsp found. 1239895743 M * wellsp i just forgotten the fstab file in /etc/vservers 1239895747 M * wellsp thanks anyways 1239895750 M * wellsp cya 1239895753 Q * wellsp Quit: :) 1239896960 J * ghislainocfs21 ~Ghislain@adsl2.aqueos.com 1239897201 J * dowdle ~dowdle@scott.coe.montana.edu 1239897220 Q * thierryp Remote host closed the connection 1239897307 Q * ghislainocfs2 Ping timeout: 480 seconds 1239897561 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1239898361 Q * esa` Quit: Coyote finally caught me 1239898591 N * Bertl_zZ Bertl 1239898597 M * Bertl back now ... 1239898663 Q * davidkarban Quit: Ex-Chat 1239898698 J * scientes ~scientes@75-165-65-163.tukw.qwest.net 1239898784 M * Bertl ryker: what's the plan with snmp on the guest? 1239899103 M * trippeh Did the hardlink hashify stuff change recently again? 1239899111 M * trippeh Getting operation not permitted in guests 1239899195 M * trippeh (on ln) 1239899195 M * Bertl not really, but, if you are using debian, all flags are 'changed' in 2.6.26-X and changed back in later kernels 1239899211 M * trippeh I use debian, but with my own kernels. 1239899225 M * Bertl i.e. you have to convert back and forth if you use the 2.6.26 kernel 1239899252 M * Bertl check that the flags are correct for the files in question 1239899268 M * trippeh How? 1239899280 M * trippeh A while since I fixed this last time :) 1239899315 M * Bertl showattr and lsxid 1239899441 M * trippeh 1# showattr ./usr/share/keyrings/debian-archive-keyring.gpg 1239899441 M * trippeh -----uI- ./usr/share/keyrings/debian-archive-keyring.gpg 1239899458 M * trippeh # lsxid ./usr/share/keyrings/debian-archive-keyring.gpg 1239899459 M * trippeh !!ERR!! ./usr/share/keyrings/debian-archive-keyring.gpg 1239899462 M * trippeh Hm :) 1239899673 M * Bertl see, broken 1239899684 M * Bertl should be -UI- for a unified file 1239899723 M * Bertl your files are 'just' immutable .. which is what causes the error 1239899752 Q * kir Quit: Leaving. 1239899795 M * trippeh Right. 1239899803 M * trippeh But this machine never booted a debian kernel :) 1239899823 M * Bertl well, maybe it booted a 2.6.26 kernel/patch 1239899830 M * trippeh Yeah, it did 1239899860 M * Bertl there you go, debian didn't specifically break it, they just took the only broken (or different) patch 1239899866 M * trippeh =) 1239899889 M * Bertl you have three options to clean that up 1239899927 M * Bertl 1) boot the 2.6.26 kernel, dump the flags, keep a list of unified files, boot the new kernel, add the proper flags there 1239899974 M * Bertl 2) remove the -I- flag from all files (in all guests) and re-unify all of them 1239900007 M * Bertl 3) do a wild guess which files are wrong (probably all -I- files with a link count > 1) and add the -U- to them 1239900026 M * trippeh 2 sounds simplest 1239900044 J * kir ~kir@swsoft-msk-nat.sw.ru 1239900074 M * trippeh I do have a 'save-vsdata.sh" script from daniel last time this broke (probably when upgrading to .26? :P) 1239900116 M * Bertl yep, that sounds about option 1 1239900124 M * trippeh Guess that would need a reboot yeah. 1239900135 M * Bertl you could run that, if you know you didn't change the flags on any files since 2.6.26 1239900189 M * trippeh No changes. Didn't unify any since either. 1239900200 M * trippeh Looks like 2) is the option I'll go for tho 1239900208 M * Bertl option 2 is probably the safest version, where option 3 is likely to be the fastest 1239900220 M * trippeh Hate reboots. :) 1239900245 M * trippeh 2) safe to do while guests are running? 1239900365 M * Bertl IIRC, vunify requires them to run, so it should be 1239900403 M * Bertl but removing the -I- flag might allow the guests to mess with the data, so be careful 1239900438 M * trippeh yeah, I control them and its just their /usr etc - no user or variable data. 1239900444 M * trippeh Should be pretty safe. 1239900463 M * trippeh Only thing messing with that data is dpkg. 1239900680 M * trippeh (which also means this doesn't break much; so I'll poke this later, probably in a few days.) 1239900683 M * trippeh Thanks for the pointers 1239900758 M * ghislainocfs21 hello bertl, i saw that cgroup can limit memory usage 1239900769 M * ghislainocfs21 what do you think of it compared to the vserver method ? 1239900796 M * Bertl hmm, I think I answered that a day ago :) 1239900808 M * ghislainocfs21 oh sorry i issed it then ! 1239900818 M * ghislainocfs21 i will check the archives :) 1239900852 Q * kir Quit: Leaving. 1239900942 M * Bertl basically I think it isn't perfect yet, and it might cause some overhead, but once it is done, it should be fine to use instead of the Linux-VServer limits 1239900978 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1239900982 M * ghislainocfs21 ok, i saw that it add a page overhead to account each for the cgroup 1239900997 M * ghislainocfs21 i wonder how they count hashified files, got to test this 1239901030 M * ghislainocfs21 and for vserver the memory of a library that is hashified is accounted in the first vserver to use it i guesss ? 1239901033 M * Bertl yeah, that would be interesting ... if you plan to do that, feel free to ask me for ideas on the test setup 1239901067 M * ghislainocfs21 oh yes i can just tell me i am in the 'i test those thing' mood 1239901081 M * Bertl also differences between the current (Linux-VServer accounting) and the cgroups accounting would be nice 1239901100 Q * harobed Ping timeout: 480 seconds 1239901118 M * Bertl first step is to get a test setup with some guests using both accounting methods 1239901126 M * ghislainocfs21 if i active both of them it will not blow up ? 1239901133 M * ghislainocfs21 ok 1239901138 M * Bertl best to do it at the same time for the first bunch of tests 1239901169 M * Bertl get a guest which has some controlled/controllable tasks running, like e.g. apache, postfix, etc 1239901198 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fe60:79d4 1239901299 M * Bertl make clones which start only one of those services automatically 1239901306 M * ghislainocfs21 yes apache /mysql would be my take 1239901322 M * Bertl i.e. so that you can start/stop it without entering the guest or such (which probably messes up the cgroups) 1239901332 N * DoberMann[ZZZzzz] DoberMann[PullA] 1239901370 M * ghislainocfs21 hum ok 1239901388 M * ghislainocfs21 lets create 2 guest then 1239901613 J * BenG ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1239901680 M * ghislainocfs21 ok, i'll eat and then create them :D 1239901686 M * ghislainocfs21 i will tell you what i find 1239901697 M * Bertl k 1239901722 Q * bourgeau Ping timeout: 480 seconds 1239901957 J * thierryp ~thierry@home.parmentelat.net 1239902244 Q * BenG Quit: I Leave 1239902249 Q * thierryp Remote host closed the connection 1239902457 M * vasko hi. is it safe to online resize ext3 partition while having several vserver guests started from that partition? or should i stop guest first? i am running vs2.3.0.34, no quota or hashify used... 1239902482 M * Bertl should be fine (at least as fine as online ext3 resizing is) 1239902502 M * vasko thanks :) 1239902616 Q * geb Ping timeout: 480 seconds 1239903187 J * geb ~geb@AOrleans-157-1-87-110.w90-20.abo.wanadoo.fr 1239903658 J * imcsk8 ~ichavero@148.229.1.11 1239903693 J * thierryp ~thierry@home.parmentelat.net 1239903762 Q * gnuk Quit: NoFeature 1239904001 J * dreamind ~dreamind@et-1-10.gw-nat.bs.ka.oneandone.net 1239904005 M * dreamind Hi 1239904040 Q * dreamind 1239904082 J * dreamind ~dreamind@et-1-10.gw-nat.bs.ka.oneandone.net 1239904085 N * dreamind Guest701 1239904127 P * Guest701 1239904132 J * Guest701 ~dreamind@et-1-10.gw-nat.bs.ka.oneandone.net 1239904164 Q * Guest701 1239904304 J * doener_ ~doener@i59F57CB5.versanet.de 1239904357 Q * geb Ping timeout: 480 seconds 1239904405 Q * doener Ping timeout: 480 seconds 1239904613 M * fb Bertl: this is virtnet in question: http://paste.linux-vserver.org/12840 1239904855 M * Bertl well, not unexpected then, disable the 'single IP' specialcasing 1239904869 M * Bertl i.e. add ~single_ip to the nflags 1239904895 J * geb ~geb@AOrleans-157-1-127-15.w90-20.abo.wanadoo.fr 1239904986 Q * thierryp Remote host closed the connection 1239905362 M * fb Bertl: ah, so this is "single IP special case" that bite my ass... thanks for pointing this out. Btw, maybe it could be auto-disabled when selecting loopback virtualisation? 1239905389 M * Bertl yeah, we'll probably do that at some point 1239905711 M * fb thanks for yous continuous, highest-grade support! 1239905726 M * Bertl you're welcome! thanks for your support and testing too! 1239905825 J * bourgeau ~bourgeau@tomsoieur.fr 1239905874 M * fb i wish i had a time to contribute with something real :( 1239906784 J * cga ~weechat@82.84.131.229 1239906858 Q * cga 1239907147 J * cga ~weechat@82.84.131.229 1239907228 Q * cga 1239908182 J * thierryp ~thierry@home.parmentelat.net 1239908579 J * cga ~weechat@82.84.131.229 1239909311 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1239909510 M * tanjix Bertl, any update for me? 1239909911 M * Bertl nothing yet, but we should have some data soon 1239910135 M * tanjix ok, i will then use the old kernel until we have a solution ;) 1239910174 J * keith ~keith@89.100.88.169 1239910246 M * keith Hi 1239910349 M * Bertl hey! 1239910365 Q * thierryp Quit: ciao folks 1239910390 J * cga ~weechat@82.84.131.229 1239910390 M * keith Can anyone here help me with setting up networking with VServer guests? 1239910407 M * Bertl probably :) what's the problem? 1239910542 M * keith I have a machine on a university network which resolves its IP automatically for the moment. I have set up a guest on the machine but not really sure about how to go about connecting it to the network 1239910589 M * Bertl Linux-VServer uses IP isolation, which means, you do not need to consider guests for the networking setup at all 1239910629 M * Bertl if you have several 'public' IPs, and you want to use some of them for guests, just assign them to the guests and make sure that the host can 'work' with them 1239910661 M * Bertl if you have a single public IP, you want to use private IPs for the guests, and S/DNAT or masquerading to get internet access 1239910699 M * keith Well I have 6 or 7 public ips and the guest I have set up as been assigned one of these 1239910752 M * Bertl good, that's basically it, depending on your (guest) config, those IP(s) will be brought up/taken down on guest start/stop 1239910794 M * keith How do I check that my guest config that that is the case? 1239910815 M * Bertl what's in your /etc/vservers//interfaces dir? 1239910816 M * keith The IPs are belonging to old machines not in use anymore. 1239910835 M * keith I have been informed that they are not associated with a MAC address but with a hostname 1239910872 M * keith So basically I set up the guest with the hostname and its associated ip. 1239910886 M * Bertl IPs are not really tied to MAC addresses or hostnames, though filters can check for combinations and resolving does the mapping between names and IPs 1239910925 M * Bertl the interfaces section should contain a directory for each IP you want the guest to 'use', like e.g. '0' 1239910957 M * keith it is just 0 in that directory 1239910962 M * Bertl with entries, 'dev' containing the interface name (e.g. eth0), 'ip' containing the IP, and 'prefix' or 'mask' containing the netmask 1239910994 M * keith I am missing a mask 1239911002 M * Bertl is a prefix there? 1239911009 M * keith Just an ip 1239911028 M * Bertl what is your netmask/prefix? 1239911035 M * keith hmm let me check 1239911134 M * keith Fairly sure its 255.255.255.0 1239911149 M * Bertl okay, so just put '24' in 'prefix' 1239911158 M * keith thank you 1239911181 M * Bertl start the guest, check with 'ip addr ls' that the guest IP is up and correct 1239911238 M * keith Its there 1239911262 J * harobed ~harobed@arl57-1-82-231-110-14.fbx.proxad.net 1239911282 M * Bertl good, now try 'ping -I on the host 1239911488 M * keith No, no ping. 1239911598 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1239911886 M * keith Its an issue with using the odl ip addresses 1239911899 M * keith A random one seems to work fine 1239912334 M * Bertl odl? 1239912334 Q * ktwilight Read error: Connection reset by peer 1239912395 J * ktwilight ~ktwilight@91.178.148.143 1239912402 M * keith Sorry old 1239912415 M * Bertl where do you use an 'old' ip? 1239912418 M * keith Sorry I'll be more specific 1239912436 M * keith I am a Sys Admin of a University's Internet Society. 1239912480 M * keith We have a load of old machine not in use anymore and their hostnames are still associated with ips on the network 1239912496 M * Bertl via DNS entries, I presume 1239912510 M * keith Ah I have no control over those :) 1239912535 M * Bertl doesn't matter, IPs do not care much about DNS :) 1239912547 M * keith Thats why I am trying to use the hostnames and IPs of the old machines for the guests 1239912569 M * Bertl should be fine, no problem to be expected there 1239912578 M * keith Thats what I thought. 1239912604 M * keith But if I use those hostnames and ips the guest doesn't see the network and vice versa 1239912620 M * fb Bertl: can this nflag be changed on a running guest? 1239912625 M * keith I am thinking maybe I have the netmask wrong? 1239912627 M * Bertl is the 'old' machine still on? 1239912633 M * keith Nope 1239912637 M * keith They are actually all dead 1239912639 M * Bertl fb: sure 1239912682 M * Bertl keith: okay, maybe you can paste me the output of 'ip addr ls' somewhere (could be in private, if you are concerned about pasting your IPs) 1239912724 M * keith Yeah private please, where? 1239912735 M * Bertl just /msg me for example 1239912759 M * Bertl (or put it on an encrypted pastebin or so) 1239912801 M * keith Ah for a sys admin I am not sure how to use IRC :) 1239912864 M * keith link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 1239912864 M * keith inet 127.0.0.1/8 scope host lo 1239912864 M * keith inet6 ::1/128 scope host 1239912864 M * keith valid_lft forever preferred_lft forever 1239912867 M * keith 2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000 1239912870 M * keith link/ether 00:02:a5:d5:9c:9b brd ff:ff:ff:ff:ff:ff 1239912872 M * keith inet 137.43.102.117/24 brd 137.43.102.255 scope global eth0 1239912875 M * keith inet 137.43.4.14/32 scope global eth0 1239912877 M * keith inet 137.43.4.14/24 brd 137.43.4.255 scope global eth0 1239912880 M * keith inet 137.43.102.118/24 brd 137.43.102.255 scope global secondary eth0 1239912882 M * keith inet6 fe80::202:a5ff:fed5:9c9b/64 scope link 1239912885 M * keith valid_lft forever preferred_lft forever 1239912887 M * keith 3: eth1: mtu 1500 qdisc noop state DOWN qlen 1000 1239912890 M * keith link/ether 00:02:a5:d5:9c:9a brd ff:ff:ff:ff:ff:ff 1239912892 M * keith bugger :S 1239913103 M * Bertl well, now that the world knows .. no point in being secret :) 1239913131 M * Bertl anyway, the .4.14 is listed twice, once with /24 and once with /32 1239913163 M * Bertl the latter one is probably from the first start (without the prefix/mask) and you didn't stop the guest before changing that, right? 1239913178 M * keith Nope, how do I fix it? 1239913197 M * Bertl just remove that IP manually, with 'ip addr del 137.43.4.14/32 dev eth0' 1239913243 M * Bertl now, the other thing is, those are in a different network .4.x than your host .102.x, which means, that you might need a different gateway too 1239913290 M * Bertl which brings us to the question, what the gateway for the .4.x network would be 1239913304 M * keith I can't tell 1239913319 M * keith Or is there a way of finding out? 1239913320 M * Bertl well, that is something you need to find out from the network admins 1239913339 M * Bertl it should also be configured in the 'old' machines network setup 1239913352 M * Bertl (otherwise it wouldn't have worked in the first place :) 1239913376 M * keith I thought of that, but it doesn't seem that any of the current admins know the passwords for them 1239913432 M * Bertl hehe, well, do you have physical access to the machine? 1239913444 M * keith The guest or the old? 1239913449 M * Bertl the old one 1239913455 M * keith I will tomorrow. 1239913483 M * Bertl okay, so you are preparing for transferring the data from the old machines? or just reusing the IPs? 1239913504 M * keith Just reusing the IPs 1239913513 M * keith The old machines are in various forms of decay. 1239913555 M * Bertl okay, then I would go for IPs in the .102.x range for now, and postpone the dual network setup for when you know the gateway address (either from the network admins or via inspection of one of the older machines) 1239913635 M * keith 'host machinename' still gives me the ip 1239913659 M * keith is there some chance that the network will tell me what the gateway address is? 1239913671 M * Bertl that is easily changed, just add an entry in /etc/hosts for now 1239913695 M * Bertl keith: no, it could be any address in the /24 range 1239913778 M * keith Thing is I am not even sure that is the netmask 1239913943 M * Bertl all that should be available from the person administrating the .4.x network range 1239914006 M * keith Thats grand then. 1239914013 M * keith I'll sort it all out tomorrow. 1239914021 M * keith Thanks for all that. 1239914037 Q * bonbons Quit: Leaving 1239914090 M * Bertl you're welcome! 1239914095 Q * keith Quit: leaving 1239914320 M * fb huh.. this vista seems to live its own life :> 1239914684 M * Bertl how would I know? :) 1239915247 Q * Pazzo Quit: Ex-Chat 1239915464 M * fb good night everyone 1239915696 J * ghislainocfs2 ~Ghislain@adsl2.aqueos.com 1239916012 Q * ghislainocfs21 Ping timeout: 480 seconds 1239916765 Q * Piet Quit: Piet 1239917287 Q * geb Ping timeout: 480 seconds 1239917592 M * Bertl tanjix: it seems I cannot recreate you problem ... I have the following test setup now (please correct me where I deviate from your setup): 1239917602 M * Bertl - host has one public IP 1239917619 M * Bertl - three guests with separate public IPs, each one running sshd 1239917642 M * Bertl - I can ssh into each guest via IP from outside 1239917654 M * Bertl - I can ssh into each guest via IP from the host 1239917681 M * Bertl - host is running a restricted sshd (host IPs) too 1239917741 M * Bertl all works fine with 2.6.29.1-vs2.3.0.36.9-pre5, and util-vserver 0.30.216-pre2833, I never end up in the wrong guest or connect to the wrong sshd 1239917919 J * Pazzo ~ugelt@sadsl-246059.rol.raiffeisen.net 1239918219 Q * Pazzo Quit: Ex-Chat 1239918906 J * yakker ~yakker@aegis.CS.Princeton.EDU 1239919097 Q * yakker 1239919101 J * er ~yakker@aegis.CS.Princeton.EDU 1239919133 M * Bertl off to bed now .. have a good one everyone! 1239919140 N * Bertl Bertl_zZ 1239919566 N * DoberMann[PullA] DoberMann[ZZZzzz] 1239919881 Q * fb Ping timeout: 480 seconds 1239922255 Q * imcsk8 Quit: This computer has gone to sleep 1239923997 Q * harobed Ping timeout: 480 seconds 1239925926 Q * bourgeau Quit: bourgeau 1239926333 J * bourgeau ~bourgeau@tomsoieur.fr 1239926350 Q * bourgeau Remote host closed the connection 1239926361 J * bourgeau ~bourgeau@tomsoieur.fr