1290125061 Q * dowdle Remote host closed the connection 1290125284 Q * manana Remote host closed the connection 1290125824 Q * aj__ Ping timeout: 480 seconds 1290126486 J * aj__ ~aj@d170093.adsl.hansenet.de 1290127650 Q * aj__ Ping timeout: 480 seconds 1290128494 J * aj__ aj@tmo-057-37.customers.d1-online.com 1290129899 Q * ghislain Quit: Leaving. 1290130615 M * Bertl off to bed now ... have a good one everyone! 1290130632 N * Bertl Bertl_zZ 1290131038 Q * hparker Quit: Quit 1290131043 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fee0:9872 1290132149 Q * hparker Quit: Quit 1290132154 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fee0:9872 1290146157 Q * DLange Quit: lunch 1290146829 J * DLange ~DLange@dlange.user.oftc.net 1290151227 Q * thierryp Remote host closed the connection 1290151565 Q * aj__ Read error: Connection reset by peer 1290152517 J * ncopa ~ncopa@3.203.202.84.customer.cdi.no 1290152559 Q * hparker Quit: Quit 1290153537 J * aj__ ~aj@213.238.45.2 1290154017 J * ghislain ~AQUEOS@LPuteaux-151-41-11-129.w217-128.abo.wanadoo.fr 1290154126 J * virman 58c71982@ircip1.mibbit.com 1290154200 Q * ghislain 1290154215 J * ghislain ~AQUEOS@LPuteaux-151-41-11-129.w217-128.abo.wanadoo.fr 1290154290 M * virman anyone knows if ebuild for newest util-vserver is ready ? :) 1290154345 M * Chlorek for gentoo? 1290154352 M * Chlorek sync and check by yourself 1290154361 M * virman ok. 1290154379 M * virman yestyerday it was not ready- i asked hollow , but no reply :) 1290154388 M * Chlorek but I don't think so 1290154392 M * Chlorek for the same reason :0 1290154394 M * Chlorek ;) 1290154617 M * virman nope 1290154623 M * virman still nothing 1290154750 Q * ghislain Quit: Leaving. 1290154766 J * ghislain ~AQUEOS@LPuteaux-151-41-11-129.w217-128.abo.wanadoo.fr 1290154843 M * Chlorek current is ok - I'm still using it and can't see any problems with .34 kernel 1290154854 M * Chlorek but I don't know what about higher versions 1290154856 M * virman i have 35 kernel.. 1290154866 M * virman and vss and rss is zeros 1290154883 M * Chlorek use cgrups 1290154891 M * Chlorek *cgroups 1290154929 M * virman unaware of changes i added netfilter , and iptables to kernel , and restarted.. krnel has changed from .34 to .35 and.. well is as it is :/ 1290154943 M * virman i know 1290154953 M * virman but still these would not appear 1290154972 M * PowerKe virman: have you tried just copying the previous version of the ebuild to a new file and see if it works? 1290155101 M * virman no , it's production enviroment.. and i don't know what would happen :) 1290155122 M * virman but i may find some free serv and try it :) 1290155173 J * BenG ~bengreen@cpc6-aztw22-2-0-cust100.aztw.cable.virginmedia.com 1290155213 Q * BenG 1290155326 M * PowerKe If it compiles it's probably good, if it doesn't, it won't leave the compilation sandbox anyway 1290155343 M * PowerKe If it installs you can still downgrade again to the old version 1290155382 M * PowerKe That's a risk you'll always have, especially when running masked packages on a gentoo box 1290155397 M * Chlorek downgrade kernel to previous version and next time wait till others tests newer version 1290155408 M * Chlorek this is my universal sloution for many problems 1290155515 M * Chlorek you shouldn't get problems if you use gentoo 1290156838 M * virman i know - but my colegue did sync kernel ,and i was unaware of that. 1290157077 J * petzsch ~markus@dslb-092-078-157-226.pools.arcor-ip.net 1290159232 M * PowerKe virman: you could unmask only a specific version of the kernel then it won't be automatically upgraded, unless a new stable version is released 1290159876 M * Chlorek virman: I think the problem sticks in your kernel configuration, because, if you use gentoo sources (vserver-sources in this case ofc.), kernel sources should be compatible with current util-vserver ver. 1290161215 M * virman well - i emerged util-vserver by ebuild.. 1290161226 M * virman iknow that sucks. 1290161282 M * virman btw - any ideas when new stable added to portage would be avilable ? :) 1290161335 M * Mr_Smoke virman: you can easily creat the ebuild yourself 1290161339 M * Mr_Smoke I did for pre2924 1290161359 M * Mr_Smoke just use the one for pre2910 and rename it. Works out of the box 1290161370 M * virman well - i don't have any idea what is in these ebuild (compilator stuff? ) 1290161376 M * virman ok. 1290161385 M * virman i'll try that today 1290161395 M * virman but not now - (i'm working :P ) 1290161572 M * virman http://mibpaste.com/3MJN0J 1290161641 M * Mr_Smoke virman: so you tried ;) 1290161656 M * virman yeah - have few minutes :P 1290161661 M * Chlorek be happy that ebuild is not the same as spec ;) 1290161666 M * Mr_Smoke virman: the way you should do it is copy from the pre2910 version with the new name, then run ebuild digest on it 1290161673 M * Chlorek ebuilds are simple 1290161676 M * virman ok. 1290161680 M * Mr_Smoke Then, and only then, will portage know about it 1290161690 M * Mr_Smoke you have to place it in the same folder obviously 1290161819 M * virman compiling :P 1290161861 M * Mr_Smoke There you go :) 1290161862 M * Chlorek you can put it in /usr/portage/local - it will prevent erease after next sync 1290161869 M * Chlorek *erase 1290161976 M * virman workzzz :D 1290161978 A * Chlorek @ coffee time 1290162011 M * Mr_Smoke Just started my first vserver with the latest 2.6.36-grsec and pre2924 1290162015 M * Mr_Smoke Smooth :) 1290162203 J * thierryp ~thierry@zankai.inria.fr 1290162494 M * Mr_Smoke Huh 1290162521 M * Mr_Smoke Isnt VIRT_MEM in /etc/vservers//flags supposed to tell the guest to see its own mem ? 1290162529 M * Mr_Smoke Oh, maybe setting rlimits would help :p 1290162607 M * Mr_Smoke Hm or not. 1290162609 M * Mr_Smoke O.o 1290162612 M * Chlorek ech 1290162642 M * Chlorek try /etc/vservers/guest/cgroup/memory.limit_in_bytes and memory.memsw.limit_in_bytes 1290162706 M * Chlorek yf you have mounted cgroup fs ofc. with suitable options 1290162734 M * Mr_Smoke Oh yeah, it's cgroups now 1290162738 M * Mr_Smoke Forgot all about that 1290162749 M * Mr_Smoke Huh, interesting. Vserver restart caused the host to hang 1290162769 M * Mr_Smoke it pings but doesn't answer its ssh port 1290162853 J * deepbit ~deepbit@cpc19-sgyl28-2-0-cust95.sgyl.cable.virginmedia.com 1290162931 M * Mr_Smoke Physical reboot it is then 1290162935 M * Chlorek ech, my fuckin spelling mistakes 1290162948 M * Chlorek I always see it after enter the text 1290162950 M * Mr_Smoke "yf" took me a while to decode, yeah :) 1290162962 M * Mr_Smoke 'k shower time 1290163069 M * deepbit Hi, can somebody help me about the isolation network in the kernel? How does it work exactly, is it possible to sniff pakets between containers for example? 1290163508 M * Chlorek ~NET_RAW blocks permissions to capture net devices 1290163551 M * Chlorek hm 1290163580 A * Chlorek have some day of kindness to children today... 1290163685 Q * Piet Ping timeout: 480 seconds 1290163850 J * barismetin ~barismeti@zanzibar.inria.fr 1290164036 M * deepbit ok thx ;), i'll try it 1290164048 M * Mr_Smoke Chlorek: keep it up ;) 1290164065 M * Chlorek for what? 1290164072 M * Mr_Smoke I'm actually upgrading from the old 2.6.22-2.2.0.7 stable to 2.3.* stuff 1290164081 M * Mr_Smoke So I need to get up to speed with stuff like CGROUPS etc 1290164084 M * Mr_Smoke Chlorek: kindness ;) 1290164275 J * Piet ~Piet__@7GDAAA1N9.tor-irc.dnsbl.oftc.net 1290164553 M * Mr_Smoke Woo, congratulations, I forgot Cgroups in my kernel entirely 1290164556 M * Mr_Smoke Time to make a new one then. 1290165325 Q * aj__ Ping timeout: 480 seconds 1290166019 M * Mr_Smoke Chlorek: care to enlighten me about the difference between limit_in_bytes and memsw.limit_in_bytes ? 1290166049 M * Chlorek I'm busy 1290166053 M * Chlorek try later, please 1290166054 M * Chlorek ok? 1290166084 M * Mr_Smoke Sure :) 1290166181 M * Chlorek cardinal difference is in phisical mem and swap resources 1290166194 M * Chlorek read the documentation 1290166202 M * Chlorek you will find there everything 1290166526 M * Mr_Smoke It appears limit_in_bytes is what the guest will "see" as available 1290166603 M * Mr_Smoke Chlorek: the documentation says they both indicate the "total memory limit". It is kinda confusing 1290166727 M * Mr_Smoke D'uh. cat: write error: Invalid argument 1290166797 M * Mr_Smoke This happens consistently every time I try to start a guest. Odd. 1290166803 M * Mr_Smoke Every second time, it works. 1290167080 M * Mr_Smoke And this is only if I have a cgroup set of limits defined for the guest 1290167330 M * Chlorek which limit? 1290167553 M * Chlorek limit_in_bytes is responsible for phys mem and memsw for phys mem + swap summarily 1290168346 Q * petzsch Quit: Leaving. 1290171444 M * Mr_Smoke Chlorek: i'll try each individually, let's see 1290171543 M * Mr_Smoke Chlorek: it's memsw. As soon as it is present, I get the "cat" error on startup 1290172020 M * Chlorek Mr_Smoke: compare with http://c.sed.pl/p518 1290172144 M * Chlorek # grep MEM /etc/vservers/jail/* 1290172144 M * Chlorek /etc/vservers/jail/flags:VIRT_MEM 1290172147 M * Chlorek it's all 1290172154 Q * FireEgl Quit: Leaving... 1290172447 M * daniel_hozac Mr_Smoke: then you probably don't have that cgroup available. 1290172607 J * kir ~kir@swsoft-msk-nat.sw.ru 1290172950 M * Guy- can vserver be messing with SO_PEERCRED? 1290172971 M * Guy- my problem is that since I upgraded to 2.6.36, postgres's ident authentication no longer works 1290172979 M * Guy- I run psql as uid 264 (the postgres user) 1290172992 M * Guy- but postgres sees 65534 (nobody) when it queries the PEERCRED 1290173255 M * Guy- (this has got to be a kernel bug, I can't imagine what else it could be, but maybe it's in mainline?) 1290173854 M * Guy- in fact, it's even weirder... if I connect as uid=0 and any arbitrary gid, peercred returns all zeroes (as if the gid were also 0) 1290173868 M * Guy- if I connect as any arbitrary nonzero uid, it's always 65534:65534 1290173962 M * Guy- I mean, so_peercred always reads it as 65534:65534 1290174043 M * daniel_hozac does it work on vanilla? 1290174052 M * daniel_hozac i don't see anything in the patch messing with those code paths. 1290174181 M * Guy- I can't easily test with vanilla now 1290174207 M * Guy- but I'm experimenting further with unixclient/unixserver (which are a quite minimal socket server and client) 1290174242 M * Mr_Smoke Chlorek: how come yours has CGROUP_NS ? I thought it was supposed to be set to "N" for the time being 1290174336 M * Guy- [pid 25450] getsockopt(4, SOL_SOCKET, SO_PEERCRED, "ic\0\0\10\1\0\0\10\1\0\0", [12]) = 0 1290174342 M * Guy- it works with those... 1290174344 M * Guy- weird 1290174367 M * Guy- I checked strace, it's really the getsockopt() call that returns garbage for postgres 1290174374 M * Guy- I don't understand how that's possible 1290174392 M * daniel_hozac what's the postgres one look like? 1290174471 M * fLoo <3 Bertl_zZ 1290174473 M * fLoo <3 daniel_hozac 1290174477 M * fLoo just to let you know 1290174493 M * Guy- daniel_hozac: getsockopt(7, SOL_SOCKET, SO_PEERCRED, "\t\21\0\0\376\377\0\0\376\377\0\0", [12]) = 0 1290174505 M * Guy- when connecting as uid and gid 264 1290174522 M * daniel_hozac do you have the entire strace? where is it getting that socket from? 1290174532 M * Guy- it's listening on it 1290174544 M * Guy- the client connects, the server forks, checks peercred 1290174547 M * daniel_hozac no, that would be a different socket :-) 1290174548 M * Mr_Smoke daniel_hozac: that cgroup is available, it is listed in /dev/cgroup as it should 1290174581 M * daniel_hozac the entire trace could be useful. 1290174735 M * daniel_hozac Mr_Smoke: well, then you likely either typoed it, or the values you entered are invalid. 1290174787 M * Guy- daniel_hozac: I'll go over one later, when fewer people are using the db; I can't imagine what's going on (especially seeing as this worked before) 1290174801 M * Mr_Smoke daniel_hozac: I already checked the typo, and it has 512M in it , is this wrong ? 1290174835 M * Guy- I found this: http://kerneltrap.org/mailarchive/linux-netdev/2010/6/14/6279287/thread 1290174842 M * Guy- [PATCH 4/8] af_unix: Allow SO_PEERCRED to work across namespaces. 1290174848 M * Mr_Smoke daniel_hozac: got it 1290174858 M * Mr_Smoke daniel_hozac: memsw was < total 1290174869 M * Mr_Smoke Vserver doesn't like that, rightly so :) 1290174924 M * daniel_hozac actually has nothing to do with Linux-VServer. that's mainline. 1290174928 Q * Piet Remote host closed the connection 1290174943 M * daniel_hozac Guy-: shouldn't affect you, unless you're using network namespaces? 1290174968 M * Guy- daniel_hozac: not explicitly 1290174973 M * Guy- but this appears to have been the problem 1290174980 M * Mr_Smoke daniel_hozac: yup, my bad. I'm still full of no coffee :S 1290174980 M * daniel_hozac how so? 1290174999 M * Guy- daniel_hozac: the init system of the vserver apparently runs in a different namespace than my shell that got spawned by vserver enter 1290175012 M * Guy- if I start the postgres server from the same shell, ident auth works 1290175080 M * daniel_hozac what util-vserver? 1290175086 M * daniel_hozac because that's just... wrong. 1290175110 N * Bertl_zZ Bertl 1290175119 M * Bertl morning folks! 1290175156 M * Guy- daniel_hozac: 0.30.216-pre2921 1290175201 M * Guy- daniel_hozac: I now tried stopping and starting the vserver, then entering it again, and the problem persists 1290175274 M * Bertl Guy-: and the kernel/patch is? 1290175294 M * Guy- 2.6.36-vs2.3.0.36.38 1290175337 M * Bertl and vserver-info - SYSINFO shows 0.30.216-pre2921 ? 1290175361 M * Guy- Kernel: 2.6.36-vs2.3.0.36.38-jayhawk 1290175362 M * Guy- VS-API: 0x00020307 1290175362 M * Guy- VCI: 0x0000000013001f11 1290175362 M * Guy- util-vserver: 0.30.216-pre2921; Nov 3 2010, 22:13:23 1290175384 M * fLoo hey Bertl 1290175385 M * fLoo :) 1290175435 M * Mr_Smoke daniel_hozac: funny thing btw : if/when kernel.grsecurity.audit_mount is enabled, I get vcontext: pivot_root(): Invalid argument when trying to start a guest 1290175444 M * Mr_Smoke Hello Bertl :) 1290175503 M * Bertl and most likely a message in dmesg as well :) 1290175575 J * manana ~mayday090@84.17.25.149 1290175625 M * Mr_Smoke Bertl: well, just a bunch of "mount of ...." entries actually, no error per se 1290175664 M * Mr_Smoke That's what bothering me 1290175687 M * Mr_Smoke Also, pivot_root is *not* forbidden in this case, in spite of what the error says 1290175691 M * Bertl you might then want to report it to the grsec folks or at least harry (because obviously grsec denied the pivot_root for some reason) 1290175718 M * Mr_Smoke Yeah, it seems like audit_mount is denying pivot_root, which is strange 1290175875 M * Mr_Smoke I don't mind disabling audit_mount for the time being, but it's strange indeed 1290175963 M * Guy- daniel_hozac: does this warrant a bugreport of some kind? 1290175983 M * Guy- daniel_hozac: or did I just make one implicitly? :) 1290176175 M * Mr_Smoke Bertl: apart from that, I have to say VServer is much smoother to use than before. My other box is running 2.2.0.7, they're worlds apart :) 1290176248 M * Bertl good to hear! 1290176279 M * Mr_Smoke And to use, too :) 1290176286 M * Bertl I'm also quite satisfied with 2.6.36 and probably 2.6.37 (or .38) will be a keeper 1290176313 M * Bertl most of the performance issues I have been observing since 2.6.25 are fixed 1290176315 M * Mr_Smoke Great 1290176323 M * Mr_Smoke Using 2.6.36 right now 1290177469 M * Guy- I just managed to deadlock 2.6.36 on i/o :( 1290177522 M * Mr_Smoke ew. 1290177535 M * Mr_Smoke How did you do that ? 1290177588 M * Guy- nothing special, really; I have a magic directory on a filesystem, and trying to read it blocks the process, apparently forever 1290177599 M * Guy- (I have one blocked on that dir since two days) 1290177614 M * Mr_Smoke Ouch. 1290177640 M * Guy- it could be an inconsistency, but still 1290177676 M * Bertl define deadlock and what filesystem? 1290177696 M * daniel_hozac Guy-: i'll look into it... 1290177772 M * Guy- daniel_hozac: thanks 1290177789 M * Guy- Bertl: hung tasks (although they're not reported as such); on xfs 1290177816 M * Guy- one is stuck in sync_page, the other in down 1290178234 M * Mr_Smoke daniel_hozac: hi there again. Do you have any idea about that audit_mount causing pivot_root to fail ? 1290178300 M * Bertl Mr_Smoke: grsec issue I guess 1290178630 M * Mr_Smoke Bertl: you reckon I should mail the grsec team directly ? 1290178677 M * Bertl I'd contact harry first, especially if it is reproduceable 1290178713 M * Bertl then, you could construct a test case for grsec and contact the grsec folks with it 1290178767 M * Mr_Smoke I'll try and get in touch with harry then 1290178778 M * Mr_Smoke You think posting to the VSML is enough ? 1290178858 M * Bertl probably, cc harry and it should be perfect 1290178920 M * Mr_Smoke Will do, thanks. 1290178995 M * daniel_hozac pivot_root isn't really meant to be used the way we use it. 1290179005 M * daniel_hozac you get that on an unpatched kernel too. 1290179015 M * daniel_hozac probably just something harry missed in the merge. 1290179064 M * Mr_Smoke daniel_hozac: I see, thanks. 1290179068 M * Mr_Smoke And indeed, 1290179083 M * Mr_Smoke on my old box (2.2.0.7) chroot_deny_pivot is *enabled*, and vserver works nicely 1290179145 M * daniel_hozac we don't use pivot_root on such an old kernel. 1290179149 M * daniel_hozac that relies on the barrier. 1290179178 M * Mr_Smoke Ok. Way over my head I think, but duly noted :) 1290179186 M * Mr_Smoke I'll post to the ML 1290179207 M * daniel_hozac hopefully it's not, that's what keeps your guests in the guest... 1290179249 M * Mr_Smoke erm ? 1290179255 M * Mr_Smoke What is not what ? Sorry I got lost 1290179622 Q * deepbit Quit: Leaving 1290179944 J * Piet ~Piet__@7GDAAA1UE.tor-irc.dnsbl.oftc.net 1290180170 J * FireEgl FireEgl@2001:470:e056:9:9496:faab:ed1e:42b2 1290181342 J * derjohn_mob aj@80.187.234.249 1290181827 Q * derjohn_mob Ping timeout: 480 seconds 1290182116 J * nkukard ~nkukard@41-133-198-30.dsl.mweb.co.za 1290182432 J * dowdle ~dowdle@scott.coe.montana.edu 1290182935 M * Bertl Guy-: hmm, haven't tested xfs for a long time, might even be Linux-VServer related, do you have a stack trace for those tasks? 1290183119 J * derjohn_mob aj@tmo-087-6.customers.d1-online.com 1290183225 Q * thierryp Ping timeout: 480 seconds 1290183623 P * kir Leaving. 1290184378 Q * barismetin Remote host closed the connection 1290184393 Q * derjohn_mob Ping timeout: 480 seconds 1290185185 Q * Piet Read error: Connection reset by peer 1290185381 J * Piet ~Piet__@7GDAAA1WU.tor-irc.dnsbl.oftc.net 1290185508 M * Guy- Bertl: can I obtain one somehow? they're still stuck 1290185546 M * Bertl yes, with the magic sysrq you should be able to dump a stack for all processes on the machine 1290185559 M * Bertl you just have to pick out those of interest then 1290185570 M * Bertl (you can trigger the magic sysrq via /proc) 1290185644 M * Guy- oh right 1290185744 M * Guy- oh, I just found the hung task message in the kernellog, so it was logged after all 1290185750 M * Guy- 2010-11-17 06:31:10.657279500 kernel: INFO: task find:10108 blocked for more than 120 seconds. 1290185758 M * Guy- I'll pastebin the stacktrace 1290185831 M * Guy- Bertl: http://paste.linux-vserver.org/18377 1290185982 M * Guy- strangely, the other process's PID doesn't occur in the kernel log 1290186009 M * Bertl could you nevertheless get a current trace of both processes? 1290186050 M * Guy- sorry, I was imprecise: I used the 't' sysrq trigger and there is no mention of my other hung task 1290186086 M * Guy- or even this one... 1290186092 M * Bertl interesting ... 1290186108 M * Bertl could it be that the dmesg buffer is not large enough? 1290186128 M * Guy- I suppose 1290186147 M * Guy- is there a way to enlarge it during runtime? 1290186151 M * Bertl could you try to increase it with 'dmesg -s'? 1290186154 M * Guy- or should I just somehow try to read it faster? 1290186199 M * Bertl ah, no, that feature was only temporary 1290186209 M * Bertl but you can read the current size this way 1290186293 M * Guy- I can read 245530 from dmesg if I specify dmesg -s 1048576 1290186359 M * Bertl okay so the buffersize is roughly 256k 1290186409 M * Bertl that means 4k lines of log, with 50-60 lines per process, that then makes 70 processes 1290186443 M * Bertl so assuming that you have a little more than 70 processes, you are not going to get all of them logged properly 1290186452 M * Guy- I have a lot more than 70 1290186457 M * Guy- I suspect I see the end of the list 1290186468 M * Guy- kern.info: Nov 19 18:06:10 kernel: e/0x10 1290186474 M * Guy- this is how the listing starts 1290186477 M * Guy- obviously a truncated line 1290186483 M * Bertl yep 1290186501 M * Bertl you might try to dump it in a close loop to a fast filesystem 1290186502 M * Guy- (1072 processes in total) 1290186512 M * Bertl it might pick up the relevant data 1290186522 M * Guy- like dd if=/proc/kmsg of=/path/to/somefile? 1290186561 M * Bertl more like while true; do dmesg >>/path/to/log; done 1290186588 M * Guy- OK 1290186620 M * Bertl but be quick, because once the I/O kicks in, you will get quite slow 1290186688 M * Guy- I'll try to pipe it into buffer(1) 1290186711 M * Bertl or write into a tmpfs 1290186752 M * Mr_Smoke harry: in case you drop by, I'm the one who just mailed the ML about audit_mount and audit_chdir 1290186853 M * Guy- Bertl: still no luck (with the tmpfs) 1290186928 M * Bertl let's try a different approach, what about: 1290186940 M * Bertl cat /proc//task//stack ? 1290186968 M * Bertl (you might need to enter the spectator context if they are inside a guest) 1290186977 M * Guy- bingo 1290187013 M * Guy- Bertl: this is the other process (zsh): http://paste.linux-vserver.org/18378 1290187063 M * Guy- Bertl: and this is the same find(1) that hung first: http://paste.linux-vserver.org/18379 1290187217 M * Bertl okay, doesn't look like Linux-VServer is involved there 1290187245 M * Bertl it might even be a false positive of the hung task checker, causing this ... 1290187260 M * Bertl (interestingly this feature was disabled in the 2.6.36 default :) 1290187261 M * Guy- well, the find process has been sitting there, motionless, for two days 1290187300 M * Guy- there was some pretty heavy i/o going on when it started (including a raid6 resync) 1290188035 Q * snooze 1290191266 Q * virman Quit: http://www.mibbit.com ajax IRC Client 1290192607 Q * FireEgl Quit: Leaving... 1290192842 J * harobed ~harobed@arl57-1-82-231-110-14.fbx.proxad.net 1290192876 Q * harobed 1290194542 J * FireEgl FireEgl@2001:470:e056:8:24f2:7e69:1531:44c0 1290194867 J * thierryp ~thierry@home.parmentelat.net 1290194971 Q * thierryp Remote host closed the connection 1290196460 M * Bertl off to bed now ... have a good one everyone! 1290196479 N * Bertl Bertl_zZ 1290197775 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fee0:9872 1290200855 J * bonbons ~bonbons@2001:a18:1:1402:2c0:9fff:fe2d:39d 1290201336 J * niki ~niki@94.145.207.11 1290202670 Q * ntrs Ping timeout: 480 seconds 1290203125 J * ntrs ~ntrs@vault08.rosehosting.com 1290205700 J * petzsch ~markus@p4FF44B8C.dip.t-dialin.net 1290205898 M * Mr_Smoke huh 1290205908 J * ktwilight_ ~keliew@91.176.12.226 1290205917 M * Mr_Smoke I get a "Device or resource busy" for an LVM volume that's unmounted... 1290205918 M * Mr_Smoke This is new 1290205932 M * daniel_hozac is it unmounted in all of your guests? 1290205976 M * Chlorek ech 1290206127 Q * ktwilight Ping timeout: 480 seconds 1290206141 M * Mr_Smoke daniel_hozac: all guests are stopped 1290206161 M * Mr_Smoke No errors anywhere 1290206405 M * Mr_Smoke I ended up rebooting the host (test system) but I don't like it one bit 1290206626 M * Mr_Smoke AHA 1290206657 M * Mr_Smoke Apparently, the LVM is "locked" that way as soon as I start the vserver on it, and even if I shut it down later 1290206747 M * Mr_Smoke I had a similar behaviour with /dev/cgroup earlier 1290206753 M * Mr_Smoke Must be a pax thing maybe 1290208081 M * Mr_Smoke daniel_hozac: weird thing is, if the filesystem is mounted but the vserver not started, everything is ok 1290208091 M * Mr_Smoke but if I start the vserver then shut it down, I get pain 1290208690 M * Mr_Smoke And now it's working. 1290208701 M * Mr_Smoke <= cursed. 1290210839 Q * petzsch Quit: Leaving.