1126915226 M * Bertl yeah, added it two lines higher right after the sendfile 1126915229 M * Aiken sendpage added to struct file_operations reiserfs_file_operations 1126915238 M * Aiken :) 1126915242 M * Bertl excellent work! 1126915272 M * Bertl Aiken: guess you tested on alpha? 1126915276 M * Aiken yes 1126915285 M * Bertl okay, so the 64bit case is verified too 1126915314 M * Bertl xfs fails with similar issues 1126915747 M * Greek0 oh cool. while forward porting vserver I've found a bug upstream 1126915756 A * Greek0 rocks :) 1126915792 M * Bertl Greek0: forward to what? 2.6.14-rc1? 1126915836 M * Greek0 hmm more or less 1126915848 M * Greek0 git checkout from today 1126915853 M * Greek0 did you do -rc1 already? 1126915920 M * Bertl nope, you're way ahead here :) 1126915984 M * Greek0 good to hear, kind of.. would have slapped myself if you did it already.. 1126916364 M * Bertl Aiken: okay, jfs is working, I guess xfs will have to wait :) 1126916664 M * Bertl hmm, generic_page_sendpage seems to do the trick there too 1126916688 M * Bertl (no idea what the implications are) 1126916889 M * Aiken I just added xfs support and vnmlinux is still below 4 meg 1126916943 M * Bertl wow, good job! 1126917226 M * Bertl Aiken: do you want to follow up on the ML (with the solution and a promise that the next release will fix it)? 1126917317 M * Aiken ok 1126917366 M * Aiken I'll mention it in an answer to chuck's question about the other reiserfs options 1126917379 M * Bertl ok, excellent, thanks! 1126917488 M * Bertl Greek0: which version are you porting btw? 1126917577 M * Greek0 Bertl: 2.1.0-pre7 1126917608 M * Bertl k, still happy porting? or everything done by now? 1126917610 M * Greek0 I dumped quilt again and looked some more into git now.. seems to fit me better 1126917648 M * Greek0 mostly done. still have to redo the tcp/inet stuff I had already done for my last checkout (redoing it because of the quilt->git switch).. should be easy 1126917685 M * Greek0 and some changes to the handling of security_vm_enough_memory() in the kernel.. 1126917690 M * Bertl okay, will update to pre14 or soo pretty soon ... 1126917715 M * Bertl so you probably prefer a delta, right? 1126917718 M * Greek0 pre14? 1126917732 M * Bertl yeah, well, actually pre15 or 16 1126917751 M * Greek0 whoops. did I miss something? 1126917757 M * Greek0 the last one I saw was -pre7 1126917768 M * Bertl nono, you missed nothing yet 1126917778 A * Bertl had a coding rush ... 1126917822 M * Greek0 ah ;) 1126917843 M * Greek0 well, deltas would be nice. otoh I can easily do them myself or use interdiff or something 1126917871 M * Bertl right, sorry for insulting your abilities ... 1126917976 M * Greek0 don't worry, I didn't feel insulted 1126918069 M * Bertl good! 1126918469 M * Bertl Greek0: hmm, did take some time to realize, that you indeed missed up to pre11 :) 1126918738 M * Greek0 mm. was kind'a sleepy this week and before a bit busy with getting my kernel setup pretty with my .deb-from-built-tree thing (kdeb) and al viro's kmk 1126918762 M * Greek0 I'm pretty proud of it now, in case you haven't noticed yet ;) 1126918780 M * Bertl excellent work! :) 1126919045 M * Bertl well, kernel build #42 if that isn't a lucky number :) 1126919224 J * monrad ~monrad@213083190134.sonofon.dk 1126919231 M * Bertl welcome monrad! 1126919508 M * monrad evening 1126919562 M * Bertl monrad: how's it going? 1126919662 M * monrad fine 1126919666 M * monrad how are you? 1126919697 M * Bertl good, a lot of new ideas and inspirations ... 1126919702 M * Greek0 I'm off to bed, cu 1126919710 M * Bertl Greek0: good night! 1126919803 M * monrad i been a bit too busy to play around with vserver 1126919904 M * monrad and i would also like to test some things out with xen and ATA over ethernet 1126919920 M * monrad so many thing to play with so little time 1126919927 M * Bertl talk to Aiken he is using the latter one IIRC 1126919955 M * monrad aoe? 1126919963 M * Bertl yep 1126919988 M * monrad it looks nice on paper 1126920007 M * monrad and maybe with some GFS on top 1126920042 M * monrad and some live migration with xen and it begins to be fun :) 1126920057 M * monrad and some vserver inside the xen's 1126921537 Q * monrad Remote host closed the connection 1126923597 T * Bertl http://linux-vserver.org/ | latest stable 2.0, 2.0.1-pre2, 1.2.10, 1.2.11-rc1, devel 2.1.0-rc1 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;) 1126923619 M * Bertl okay, off to bed now ... have a nice whatever everyone! 1126923644 N * Bertl Bertl_zZ 1126924430 A * Aiken just noticed his name mentioned 1126928589 J * Aiken_ ~james@tooax6-129.dialup.optusnet.com.au 1126928913 Q * Aiken Ping timeout: 480 seconds 1126942199 J * dddd44 debian-tor@tor-irc.dnsbl.oftc.net 1126942535 Q * kas_3 Ping timeout: 480 seconds 1126942626 J * kas_3 tor@tor-irc.dnsbl.oftc.net 1126942860 Q * dddd44 Ping timeout: 480 seconds 1126942884 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126942983 J * vip-vs ~vip-vs@cc521104-d.ensch1.ov.home.nl 1126943042 M * vip-vs hi folks 1126943170 Q * kas_3 Ping timeout: 480 seconds 1126943185 J * ag-_ ag@caladan.roxor.cx 1126943215 Q * ag- Read error: Connection reset by peer 1126943240 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net 1126943385 Q * dddd44 Ping timeout: 480 seconds 1126943963 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126944035 Q * kas_3 Ping timeout: 480 seconds 1126944533 Q * ag-_ Quit: leaving 1126944541 J * ag- ag@caladan.roxor.cx 1126946047 J * menomc ~amery@200.75.27.114 1126946154 Q * mnemoc Ping timeout: 480 seconds 1126946154 N * menomc mnemoc 1126946887 J * Aiken__ ~james@tooax8-091.dialup.optusnet.com.au 1126947205 Q * Aiken_ Ping timeout: 480 seconds 1126949049 N * Bertl_zZ Bertl 1126949053 M * Bertl morning folks! 1126949222 J * maharaja maharaja@ip52.ipax.at 1126949303 M * Doener morning 1126949409 M * Doener Bertl: sunday won't work for me, we move on sunday and phone/internet will be installed on monday, so no connection at that time 1126949485 M * maharaja re 1126949509 M * Bertl Doener: okay, will not be the last discussion round I guess ... 1126949540 M * Doener yeah, just wanted to let you know 1126949728 M * Bertl thanks! 1126949777 M * maharaja Bertl: im going to buy a new serial cable today 1126949783 M * maharaja and will set up another vserver 1126949790 M * maharaja as the crashes still persist 1126949794 M * maharaja maybe with ext3 instead of xfs 1126949810 M * maharaja yesterday, the server (again) crashed 1126949827 M * maharaja and i wasnt able to use the magic sysrq 1126949866 M * Bertl very strange ... no oops, no nmi watchdog, no sysrq ... but hang 1126949884 M * Bertl in my book this would be the classical hardware bug 1126949903 M * maharaja mhm 1126949912 M * maharaja i allready exchange the hardware 2 times 1126949913 M * Bertl (but maybe we will discover a new kind of kernel issue) 1126949932 M * maharaja 1. time from p4 1.5ghz to p4 3.2ghz 1126949938 M * maharaja second thime back to 1.5ghz 1126949955 M * maharaja everything except the harddisks has been replaced 1126949963 M * maharaja im going to buy two new harddisks today 1126949967 M * Bertl yeah, you already told me ... 1126949971 M * maharaja and again setup the 3.2ghz machine 1126949976 M * maharaja any suggestions 1126949981 M * maharaja like using ext3 instead of xfs? 1126949985 M * maharaja on the second machine 1126949986 M * Bertl I'd look into power supply and housing enironment 1126949989 M * maharaja or should the setup be the same? 1126950004 M * maharaja we've got 10 other servers on the same power supply 1126950007 M * maharaja they work flawlessly 1126950035 M * maharaja (no vserver patches) 1126950046 M * Bertl well, based on this argument, 100 hosting companies have linux+ linux-vserver running and they work flawlessly :) 1126950070 M * maharaja hehe 1126950073 M * maharaja i agree on that 1126950105 M * Bertl I'm not saying it's not software related ... 1126950144 M * Bertl but it _looks_ very hardware specific, unless you managed to get all the other things terribly wrong, which I do not assume ... 1126950160 M * maharaja is there any special "code" a serial cable has got? 1126950170 M * maharaja or a "correctly" wired serial cable has got 1126950178 M * maharaja cause im not getting any good results from geizhals/ditech 1126950187 M * Bertl hmm, it is called null-modem cable .. 1126950230 M * Bertl and it's not terribly hard to make one yourself ... 1126950270 M * maharaja does this sound right: Laplink Kabel 2x9polig, 2x25polig, 4x Buchse, 2.5 m 1126950276 M * maharaja mhm 1126950276 M * maharaja no0 1126950278 M * maharaja :) 1126950455 M * Bertl http://www.directshopper.de/axis-serial-null-modem-kabel-2100-2120-2420_netzwerkkabel-kabel-lt-80v_p (don't know their wiring though) 1126950464 M * Doener http://www.preisroboter.de/ergebnis182091.html -- shipping costs 500-700% of the cable *g* 1126950569 M * Bertl maharaja: do you know somebody who can tell the hot side of the soldering iron from the handle? 1126950654 J * yarihm ~yarihm@84-74-16-246.dclient.hispeed.ch 1126950770 M * Bertl maharaja: if so, the easiest way is to go to Conrad.de, get two 9 pin d-sub female connectors, 2-4 meter 8 wire screened cable, and build one yourself: http://vserver.13thfloor.at/Null-Modem/nullmodem9to9.html 1126950797 M * maharaja Bertl: what do you mean with your first sentence? 1126950802 M * maharaja i do not understand it :P 1126950809 M * maharaja could you tell it in german? 1126950845 M * Bertl transl: "kennst du jemanden der weiss wo die heisse seite beim lötkolben ist?" 1126950954 M * Greek0 hehe 1126950959 M * maharaja :P 1126950961 M * maharaja not me ;) 1126950988 M * Greek0 the first time I read it I thought you ment "heise seite" (as in heise.de) .. 1126950991 Q * yarihm Read error: Connection reset by peer 1126951021 J * yarihm ~yarihm@84-74-16-246.dclient.hispeed.ch 1126951081 M * Bertl Greek0: yeah, well, 'heisse ende' would have less ambiguous 1126951088 M * Bertl *been 1126951109 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net 1126951140 M * Greek0 Bertl: what would have helped too was putting on my glasses after getting up ;) 1126951221 Q * yarihm Quit: 1126951234 J * yarihm ~yarihm@84-74-16-246.dclient.hispeed.ch 1126951271 M * Bertl wb yarihm! 1126951299 M * Bertl Greek0: btw, 2.1.0-rc1 is out ... 1126951306 M * yarihm yo Bertl 1126951445 Q * dddd44 Ping timeout: 480 seconds 1126951510 M * Greek0 thanks, saw it already 1126951532 M * Greek0 merge conflicts resolved so far, I'm now at getting the thing to compile again 1126951543 M * Greek0 file_struct was splitted, and that's breaking some stuff 1126951638 M * Bertl interesting ... 1126951832 M * Bertl okay, off for now ... back later in the evening ... 1126951836 N * Bertl Bertl_oO 1126951932 M * maharaja Bertl: have you got a proper minicom rc? 1126951948 M * maharaja mhm - brb - rebot 1126951951 M * maharaja reboot 1126952045 M * vip-vs Is it possible to dynamicly assign an IP address to a VS, like dhclient does? 1126952107 M * vip-vs I haven't found any documentation about it... 1126952237 Q * yarihm Read error: Connection reset by peer 1126952246 J * yarihm ~yarihm@84-74-16-246.dclient.hispeed.ch 1126952315 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126952605 Q * kas_3 Ping timeout: 480 seconds 1126952744 M * AndrewLee Bertl_oO: Hi, I tested sarge can not pass the test of testfs.sh. 1126952840 M * AndrewLee Bertl_oO: I am using kernel-source-2.4.27+kernel-patch-vserver in sarge, would that be matter? 1126953850 Q * yarihm Read error: Connection reset by peer 1126954062 J * yarihm ~yarihm@84-74-16-246.dclient.hispeed.ch 1126956602 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net 1126956764 Q * Aiken__ Quit: Leaving 1126956970 Q * dddd44 Ping timeout: 480 seconds 1126957140 Q * kas_3 Ping timeout: 480 seconds 1126957227 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126959384 Q * Doener Ping timeout: 480 seconds 1126959390 J * Doener ~doener@p548748D0.dip.t-dialin.net 1126959689 M * maharaja re 1126962225 J * renihs ~renihs___@193.170.52.70 1126966407 J * liquid3649 ~inet@p54974A06.dip.t-dialin.net 1126966708 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net 1126966995 Q * dddd44 Ping timeout: 480 seconds 1126966998 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126967179 M * nebuchadnezzar hello, i try to patch a 2.6.13.1 with the v2.1.0-rc1 and I get an error 1126967186 M * maharaja which 1126967187 M * maharaja error 1126967188 M * nebuchadnezzar 1 out of 2 hunks FAILED -- saving rejects to file arch/ia64/kernel/perfmon.c.rej 1126967188 M * nebuchadnezzar 1126967197 M * nebuchadnezzar Hunk #2 FAILED at 2349. 1126967215 Q * kas_3 Ping timeout: 480 seconds 1126967215 M * maharaja look at those rej files 1126967220 M * maharaja and try to merge the patch by hand 1126967227 M * nebuchadnezzar I make a --dry-run :-) 1126967379 M * nebuchadnezzar well, my filesytem is not consitant, forget what I said 1126967588 Q * dddd44 Remote host closed the connection 1126967745 M * maharaja ok 1126967748 M * maharaja :) 1126967790 A * nebuchadnezzar is not so happy :-/ 1126967874 M * maharaja neither am i :> 1126967878 M * maharaja whats happening? 1126967911 M * nebuchadnezzar I make a badblocks check to see 1126968162 M * maharaja what does dmesg|tail show ? 1126968333 M * nebuchadnezzar nothing 1126968380 M * nebuchadnezzar 0 bad blcks... 1126968402 M * nebuchadnezzar I need to pass my system to xfs instead of ext3... 1126968412 A * nebuchadnezzar don't really like ext3 1126969120 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126969741 Q * mugwump Ping timeout: 480 seconds 1126969761 M * maharaja nebuchadnezzar, do not enable 4k stacks with xfs 1126969789 M * nebuchadnezzar I have not 4k stack 1126969836 M * nebuchadnezzar I have no pb with xfs 1126970284 Q * renihs Remote host closed the connection 1126970575 Q * liquid3649 Read error: Connection reset by peer 1126970668 M * maharaja just a hint of mine 1126970694 M * maharaja cause my vserver used to crash very often because of 4k stacks has not been enough for xfs on lvm on mdraid 1126970694 M * maharaja ;) 1126970844 M * nebuchadnezzar I have juste XFS on LVM for now, I only use hardware RAID :-) 1126970916 M * maharaja ;) 1126970927 M * maharaja which hw raid? 1126971127 M * nebuchadnezzar I plan to buy a 3ware card 1126971139 M * nebuchadnezzar with real hardware SATA RAID 1126971152 M * nebuchadnezzar or adaptec 2410 1126971192 A * mnemoc see to money jumping around nebuchadnezzar 1126971223 M * nebuchadnezzar well, I plan to... 1126971247 M * nebuchadnezzar my server crash cause of a disk failure... 1126971251 M * nebuchadnezzar crashed 1126971863 M * maharaja i use softwareraid on all servers 1126971864 M * maharaja as on most of these systems, data reduncy is not that important 1126971864 M * maharaja its simply to save some time 1126972705 Q * nebuchadnezzar Remote host closed the connection 1126973406 J * nebuchadnezzar ~nebu@zion.asgardr.info 1126975982 J * nayco ~nayco@lns-bzn-9-nan-82-251-11-232.adsl.proxad.net 1126976010 M * nayco Hello, !!! 1126976386 M * daniel_hozac hi nayco 1126976543 M * maharaja daniel: do you know what things bertl fixed in his fix02 patch for util-vserver 208? 1126976819 Q * obi Ping timeout: 480 seconds 1126977488 J * obi ~obi@asus.saftware.de 1126978222 M * maharaja can somebody help me to get my serial console setup up and running? 1126978607 J * gndmstr ~gndmstr@ip1.pathworx.sbbsnet.net 1126978694 Q * nebuchadnezzar Quit: ERC Version 5.0 (CVS) $Revision: 1.769 $ (IRC client for Emacs) 1126978760 M * gndmstr anyone familiar with the reiserfs extended attr problem? i got a patch for file.c which fixed CoW on it, but i wonder... my clones were made without that patch. will this fix things automatically or should i reinstall the clones? 1126979056 M * maharaja ok, working 1126979504 M * daniel_hozac maharaja: fix02 updates the kernel headers and the alternative syscall implementation. 1126979543 M * daniel_hozac gndmstr: it should fix things automatically, if you had it mounted with attrs when you used setattr. 1126979564 M * daniel_hozac maharaja: fix02 vs. fix01 just fixes an issue with PIC. 1126979572 M * maharaja thank you daniel! 1126979624 M * gndmstr it was mounted with attrs, however the kernel did not have extended attributes enabled. it only has basic reiserfs support enabled 1126979673 M * gndmstr it now has full extended plus security and posix enabled plus the file.c patch 1126979683 M * gndmstr tested and proved cow works on a test setup 1126979713 M * maharaja *afk (bnc on) 1126979715 M * gndmstr but the entire template and clone system was done without all that even tho attrs were enabled in fstab 1126980212 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net 1126980505 Q * dddd44 Ping timeout: 480 seconds 1126980536 J * dddd44 ~dhb55@66.254.98.196 1126980700 Q * kas_3 Ping timeout: 480 seconds 1126980734 J * kas_3 ~dhb55@tor-irc.dnsbl.oftc.net 1126981045 Q * dddd44 Ping timeout: 480 seconds 1126981108 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126981475 Q * kas_3 Ping timeout: 480 seconds 1126981505 J * kas_3 ~dhb55@66.254.98.196 1126981590 Q * dddd44 Ping timeout: 480 seconds 1126981658 Q * gndmstr Remote host closed the connection 1126982918 J * monrad ~monrad@213083190134.sonofon.dk 1126983459 J * gndmstr ~gndmstr@ip1.pathworx.sbbsnet.net 1126983495 M * gndmstr quick question. are guests or the vserver system in general happy with NPTL? 1126983540 M * gndmstr i have a friend who wants to try veservers out but one of his servers to put in as a guest makes extensive use of it 1126984079 M * sannes it will work, don't see how vservers could affect this very much .. 1126984104 M * gndmstr ok thanks. it was one of his questions and i have 0 experience with that 1126984169 M * Hollow gndmstr: it currently does work... i don't have any other boxes than with nptl 1126984179 M * gndmstr cool thanks 1126984208 M * sannes easiest way to find out is just to test it :) .. but the rule of thumb is of course that if it works on vanilla linux it works in a vserver :) 1126984241 M * gndmstr cool thaniks 1126984566 Q * vip-vs Remote host closed the connection 1126984874 M * gndmstr are there any caveats to moving Xorg and KDE off a host onto a guest? 1126984898 M * daniel_hozac do you intend to run the X server in the vserver? 1126984932 M * gndmstr that was my intention yes. there is a vid card however i do not use it as a local workstation. it is used as a remote workstation only 1126984944 M * gndmstr xdm is running but not logged into 1126984951 M * daniel_hozac so you run the X server elsewhere? 1126984981 M * gndmstr no.. it runs x locally but i dont use that local method.. i use vnc server and log into it remotely 1126984985 M * daniel_hozac the X server requires quite a few /proc entries as well as a lot of capabilities, IIRC. 1126985004 M * gndmstr maybe im better off running it as i have it now , on the host 1126985017 M * gndmstr see i made my 2nd workstation my vserver experiment box 1126985032 M * gndmstr so i still use my workstation as i used to but i have vservers running as well 1126985063 M * gndmstr i was just thinking of trying to run it as a guest and if it coudl work, i thought i would move the 2nd workstation to my main workstation in a guest 1126985114 M * gndmstr if i can do that effectively, then i can free up an entire machine to use otherwise 1126985168 M * gndmstr taht workstation is only used for internet admin things.. mostly ssh and web. .there is almost nothing else on it 1126985198 M * gndmstr so it will never have to play a video or anything a normal workstation may need to do 1126985307 M * gndmstr the reason for that remote one is primarily for our router admin in finland. my network here is the only one with direct access into our system private net at the NOC. so she logs into my workstation to do her admin work as well.. it is impossible to give her direct access since it uses unroutable ip addys 1126985380 N * Bertl_oO Bertl 1126985385 M * Bertl good evening folks! 1126985391 M * daniel_hozac good evening! 1126985392 M * gndmstr evening 1126985444 M * micah /win 4 1126985447 M * micah oops! 1126985491 M * Bertl micah: at least no panic :) 1126985673 M * micah lol 1126985700 M * micah what a bug report that would be... mistyping /win 4 in IRC causes oops 1126985938 M * Bertl well, one of those 99% non linux-vserver related ones, as usual :) 1126985974 M * Bertl micah: btw, you 'disappeared' yesterday, did you find the util-vserver arch tests I uploaded? 1126986811 Q * gndmstr Remote host closed the connection 1126988070 Q * monrad Quit: Leaving 1126988214 M * Bertl micah: ?! 1126989579 Q * kas_3 Remote host closed the connection 1126989813 M * nayco Hello, Bertl :) 1126990343 J * Aiken ~james@tooax6-148.dialup.optusnet.com.au 1126990407 M * Bertl hey nayco! Aiken! 1126990613 M * Aiken seems to be morning, so good morning 1126991633 M * mnemoc morning where? 1126991657 M * Aiken here 1126991672 M * mnemoc australia? 1126991676 M * Aiken east coast aus 1126991682 M * mnemoc :) 1126991700 M * Aiken nearly said east coast vk 1126991914 J * waters33637 ~mike@etcb01-00-bnwkga-69-173-45-114.atlaga.adelphia.net 1126992149 Q * obi Ping timeout: 480 seconds 1126992359 M * Bertl welcome waters33637! 1126992701 J * obi ~obi@2001:6f8:13d4:0:b0fa:8fff:fee5:c429 1126993469 M * Bertl Aiken: how is 2.1.0-rc1? 1126993584 M * Aiken something is wrong 1126993604 M * Aiken I can ping that box and nothing else 1126993620 M * Aiken it was working until a couple of minutes ago with (I think) nearly 12 hour uptime 1126993649 M * Aiken all I did was 'grep -r xfs_bmap_check_leaf_extents *' in the source tree and it has stopped responding 1126993681 M * Bertl hmm, not a good sign ... 1126993804 M * Aiken just came back for the camera, a back trace on the screen :( 1126993830 M * Bertl k, better than nothing ... 1126993876 M * Aiken where that machine is, all it has to do is drop 3C and it will be the same temp as in our fridge 1126994240 M * Aiken mm/vmscan.c:206 1126994583 J * Blissex ~Blissex@82-69-39-138.dsl.in-addr.zen.co.uk 1126994632 M * Bertl Aiken: and this is with 2.6.13.1-vs2.1.0-rc1? 1126994655 M * Aiken yes, I am uploading a pic 1126994673 M * Bertl seems like a slab issue, probably mainline ... 1126994700 M * Aiken http://members.optusnet.com.au/~boddingt/crash.png 1126994705 M * Bertl tx 1126994718 M * Aiken I have 2 cow vservers running 1126994735 M * Aiken had just done a cow test with reiserfs on aoe which passed :) 1126994775 M * Aiken then loaded xfs (as a module) and then did a grep for a symbol that the xfsprogs kept claiming did not exist 1126994781 M * Aiken the grep never finished 1126994837 M * Bertl osf_select? 1126994871 M * Bertl are you running OSF binaries? 1126994873 M * Aiken (root@pebbles) addr2line -e vmlinux fffffc00003156c8 1126994873 M * Aiken arch/alpha/kernel/osf_sys.c:1026 1126994896 M * Aiken this is where I ask what is osf 1126994922 M * Bertl the 'original' operating system 1126994931 M * Bertl Alpha OSF/1 1126994937 M * Aiken no 1126994965 M * Bertl unless I'm completely wrong, this file only contains osf syscalls 1126995000 M * Aiken a quick look at the file says you are not wrong 1126995096 M * Bertl strange ... 1126995206 M * Aiken fs/select.c:93 is twice in the bt 1126995269 M * Bertl which one (address) is that? 1126995282 M * Aiken fffffc0000360608 1126995290 M * Aiken no 1126995348 M * Aiken fffffc00003a7590 and fffffc00003a75f0 1126995362 M * Aiken __pollwait 1126995368 M * Bertl ah, okay, different parts ... 1126995382 M * Aiken trying to do too many things at once :( 1126995654 M * Aiken just did all of the backtrace http://pastebin.com/366671 1126995680 M * Bertl tx again! 1126995928 M * Bertl hmm, I assume you already rebooted the machine, right? 1126995954 M * Aiken yes 1126996248 Q * janra_ Ping timeout: 480 seconds 1126996632 M * Aiken looks like I just crashed the machine again 1126996665 M * Aiken it was fine until I modprobe xfs 1126997370 M * Bertl maybe the xfs module does something weird? 1126997389 M * Bertl this time, try to do a task/memory dump if possible via magic sysrq ... 1126997409 M * Bertl (not sure how much you will be able to capture though) 1126997429 M * Bertl a serial console might become useful ... 1126997486 M * Aiken I had the silly idea of trying an old dot matrix printer as the serial console 1126997516 A * Aiken ducks for cover, he might not have the magic keys compiled in 1126997629 M * Bertl lol 1126997640 M * Bertl well, if it is reproduceable *shiver* :) 1126997682 M * Aiken nearly finished in a net on the radio so can go reset the alpha shortly 1126997698 M * Bertl maybe just do a *cold* boot ... and try to get it to *freeze* again ... 1126997725 M * nayco :D 1126997833 J * dddd44 ~dhb55@tor-irc.dnsbl.oftc.net 1126997948 M * Bertl welcome dddd44!? 1126998279 M * Aiken compiling the magix key support first 1126998288 M * Aiken magic even 1126998467 M * nayco Bertl: To implement Mandriva support in vserver-build, what do you think is easier : making a .distribution/ subdirectory for use with "vserver-build.rpm", or create a "vserver-build.urpmi" script ? 1126998564 M * nayco ...These scripts are not that easy to understand, given the number of included files, which make a pretty long code to read to fully understand how it works... 1126998607 M * daniel_hozac i think fully understanding the utils is reserved to enrico. ;) 1126998649 M * nayco Yes, I guess so, but... if anyone here knows their internals, I could spare some time ;) 1126998673 P * waters33637 Leaving 1126999019 J * linc ~lms@bl5-117-74.dsl.telepac.pt 1126999023 M * linc hello :o) 1126999043 M * Bertl welcome linc! 1126999048 M * linc bertl ;o) 1126999049 M * linc hello dewd 1126999060 M * linc pvt or here? 1126999060 M * Bertl so friends? 1126999063 M * linc off course! 1126999064 M * linc ;o) 1126999073 M * linc how can i stay mad with my favourite developer he he he 1126999074 M * linc :oP 1126999085 M * Bertl k, let's keep it here, unless it is absolutely non-vserver related :) 1126999094 M * linc nopes :o) 1126999101 M * linc well, so what do you need to know extra? 1126999106 M * Bertl (channel is logged, and folks might search the logs ...) 1126999112 M * linc yeaps, better :o) 1126999123 M * linc what kind of info would you like to know regarding my problem? 1126999133 M * Bertl did you try the suggested ping/tcpdump? 1126999137 M * linc yeaps 1126999142 M * linc that was the way 1126999147 M * linc i found out i needed to SNAT 1126999147 M * Bertl could you paste/upload the output? 1126999152 M * linc sure..w8t 1126999184 M * linc hmm..allthough right now im not near the server..so if i reboot i cant change the cablez on the nics :o7 1126999202 M * linc ...and quite franqly i dont want to be near them 1126999210 M * linc i allready did that the other weekend :D 1126999212 M * linc hehehe 1126999233 M * linc ill take away the SNAT rules and show you the dump 1126999241 M * Bertl good ... 1126999571 M * linc something really strange is happening here bert, i commented out the SNAT and it is still working 1126999583 M * linc but i swear that i keep having problems with this :o| 1126999620 M * linc just like what i mentioned on the other email, all the configurations where alike, the only thing that happened was the extra NIC 1126999637 M * Bertl do you see a change in the tcpdump if you enable/disable the SNAT rule? 1126999653 M * linc and when i booted with that extra NIC the rules that said $IPTABLES -A INPUT -d vserver.ip --dport X -j ACCEPT did not work! :o| 1126999670 M * linc [allthough the NIC itself was working perfectly cause i sshd to it] 1126999686 M * nayco Bertl: What do you think about [01:07] ? is there only Enrico that can help me ? 1126999740 M * Bertl well, easier ... I already opted for the urpmi approach, no? 1126999762 M * Bertl (guess using rpm is the easier one) 1126999808 M * nayco What do you mean by " I already opted for the urpmi approach" ? 1126999832 M * linc with SNAT: 1126999838 M * Bertl nayco: last time, IIRC, I said: it would be cool to add urpmi support (like yum and apt-get) no? 1126999874 M * linc root@newton-root ~# tcpdump -vni eth0 | grep 192.168.3.81.80 1126999874 M * linc tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 96 bytes 1126999874 M * linc 00:30:52.259075 IP (tos 0x10, ttl 64, id 11388, offset 0, flags [DF], length: 60) 192.168.2.25.52838 > 192.168.3.81.80: S [tcp sum ok] 2986325566:2986325566(0) win 5840 1126999874 M * linc 00:30:52.259369 IP (tos 0x0, ttl 63, id 0, offset 0, flags [DF], length: 60) 192.168.3.81.80 > 192.168.2.25.52838: S [tcp sum ok] 1780755637:1780755637(0) ack 2986325567 win 5792 1126999874 M * linc 00:30:52.259387 IP (tos 0x10, ttl 64, id 11389, offset 0, flags [DF], length: 52) 192.168.2.25.52838 > 192.168.3.81.80: . [tcp sum ok] ack 1 win 5840 1126999874 M * linc 00:30:53.334666 IP (tos 0x10, ttl 64, id 11390, offset 0, flags [DF], length: 57) 192.168.2.25.52838 > 192.168.3.81.80: P [tcp sum ok] 1:6(5) ack 1 win 5840 1126999874 M * linc 00:30:53.334866 IP (tos 0x0, ttl 63, id 5286, offset 0, flags [DF], length: 52) 192.168.3.81.80 > 192.168.2.25.52838: . [tcp sum ok] ack 6 win 5792 1126999877 M * linc without SNAT: 1126999907 M * linc root@newton-root ~# tcpdump -vni eth0 | grep 192.168.3.81.80 1126999907 M * linc tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 96 bytes 1126999907 M * linc 00:31:35.798152 IP (tos 0x10, ttl 64, id 35832, offset 0, flags [DF], length: 60) 192.168.2.25.52839 > 192.168.3.81.80: S [tcp sum ok] 3044221116:3044221116(0) win 5840 1126999907 M * linc 00:31:35.798443 IP (tos 0x0, ttl 63, id 0, offset 0, flags [DF], length: 60) 192.168.3.81.80 > 192.168.2.25.52839: S [tcp sum ok] 1841066557:1841066557(0) ack 3044221117 win 5792 1126999907 M * linc 00:31:35.798461 IP (tos 0x10, ttl 64, id 35833, offset 0, flags [DF], length: 52) 192.168.2.25.52839 > 192.168.3.81.80: . [tcp sum ok] ack 1 win 5840 1126999908 M * linc 00:31:36.561758 IP (tos 0x10, ttl 64, id 35834, offset 0, flags [DF], length: 57) 192.168.2.25.52839 > 192.168.3.81.80: P [tcp sum ok] 1:6(5) ack 1 win 5840 1126999908 M * linc 00:31:36.561959 IP (tos 0x0, ttl 63, id 15148, offset 0, flags [DF], length: 52) 192.168.3.81.80 > 192.168.2.25.52839: . [tcp sum ok] ack 6 win 5792 1126999912 M * Bertl linc: I doubt that my example ping/tcpdump will log that much! 1126999914 M * linc ops 1126999932 M * Bertl why not just use that instead? 1126999937 M * linc bertl: but i was having problems with connections to other vservers, not to the internet 1126999960 M * linc [mail-nss vserver] on [newton-root] -> gw <- [leonardo-root] with [ciisp vserver] 1126999973 M * linc i could connect from mail-nss to EVERYWHERE except to the ciisp vserver 1126999990 M * linc and i didnt make any change on my firewall script dewd :o| 1127000000 M * linc i just recompiled the kernel 1127000001 M * Aiken I just had kswapd, grep, syslogd oops all at the same time and can not rmmod xfs 1127000051 M * linc bertl: the problem im having is very tricky to explain because it only happens sometimes. our other admin said he THOUGHT i saw some netfilter'vserver patch 1127000059 M * Bertl linc: first mail: Sometimes i need to had SNAT rules for my vservers to route outside the root server and, someother times, i cant access my vservers from the outside 1127000064 M * linc i just wanted to ask if there is any known bug with iptables/vservers ? 1127000072 M * Bertl does this now mean inside or outside or what? :) 1127000100 M * micah Bertl: sorry, my connection keeps dropping 1127000103 M * linc bertl: like i said, it is not very easy to explain....written....especially in english hehehe 1127000110 M * linc at least not to me :oP 1127000114 M * Bertl linc: I can assure you, _all_ issues you are seeing are simple linux-networking issues ... 1127000123 M * linc bertl: oh k dewd! thanks :o) 1127000129 M * linc that is just what i needed to know 1127000139 M * linc i needed to know if there was any kind of patch 1127000150 M * Bertl linc: but, this doesn't mean that I can/will not explain it to you :) 1127000166 M * linc btw, since iam here, in the past i kept using the 2.4.x series because your support for 2.6.x wasnt stable yeat... 1127000170 M * nayco Bertl: Oh, ok ;) ! Yes, that would be cool, but looking at the vserver-build* scripts, I wondered if urpmi wasn't only a special case of rpm; and more, if using urpmi could cause a portability issue, dunno if urpmi exists on other distros... But, in fact, I too prefer creating a full-urpmi script, because as urpmi handles a lot of things by itself, it could simplify the process. 1127000204 M * linc since you have completly "ported" to 2.6.x, will the 2.4.x series stop getting support? i noticed you havent released a 2.4.31 patch [but i also noticed that 2.4.30 patch applies cleanly with 2.4.31] 1127000227 M * Bertl nayco: I'd consider the urpm* similar to yum, or apt-rpm which is supported right now 1127000241 M * Bertl micah: np, you got/saw the test results? 1127000283 M * micah Bertl: no, i can't seem to keep my connection open :P 1127000289 M * Bertl linc: no, 2.4 is still 'maintained', although no new development will happen there (btw, we are at 1.2.11-rc1 (see topic) 1127000307 M * Bertl micah: http://vserver.13thfloor.at/Stuff/Debian/ 1127000315 M * Bertl micah: last few .txt files 1127000324 M * linc bertl: what is the SYS_CAP needed for icmp? 1127000353 M * Bertl linc: what kernel are we talking about? :) 1127000359 M * linc 2.4.30 1127000364 M * linc :o) 1127000373 M * Bertl CAP_NET_RAW 1127000387 M * linc [dear god, didnt know the CAPS names changed...] 1127000388 M * linc thanks! 1127000411 M * Bertl linc: no, they didn't but 2.6 doesn't require that anymore 1127000421 M * linc no? 1127000422 M * nayco Bertl: You're, I gonna take a closed look to the corresponding scripts. 1127000431 M * linc where can i read some info on it? 1127000437 M * nayco +You're right 1127000448 M * Bertl linc: it's all over linux-vserver.org *G* 1127000449 M * nayco *closer 1127000468 M * Bertl nayco: okay, great! 1127000617 M * Bertl linc: btw, what do you need the CAP_NET_RAW for? 1127000671 M * linc so i could ping from the vserver 1127000682 M * Bertl hmm, did I suggest that? 1127000690 M * linc i guess pinging from the root server wouldnt be a good test 1127000710 M * Bertl on the contrary ... it would :) 1127000718 M * linc oh kie, youre the boss :o) 1127000761 M * Bertl that's why you use the -I option :) 1127000827 M * linc i know, to specify the interface 1127000835 M * linc but...if the problem with routing was on the vservers code 1127000843 M * linc that test would be irrelevant 1127000845 M * linc [i think] 1127000857 M * linc you're testing the kernel implementation :oP 1127000861 M * linc not your own 1127000863 M * linc but w8t 1127000865 M * Bertl trust me, a) there is no problem with routing , and it isn't in the vserver code :) 1127000869 M * linc hehehe 1127000883 M * linc you're damn'confident about your code ;o)) 1127000885 M * Bertl (out of b) error ) 1127000886 M * linc GOOD 1127000904 M * linc thats why we trust you to develop the great vservers! ;o) 1127000907 M * linc either way 1127000935 M * linc root@newton-root ~# tcpdump -vvnei eth0 icmp 1127000935 M * linc tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 96 bytes 1127000935 M * linc 00:48:48.178811 00:0d:61:01:35:b3 > 00:05:5d:e6:1e:22, ethertype IPv4 (0x0800), length 98: IP (tos 0x0, ttl 64, id 0, offset 0, flags [DF], length: 84) 192.168.2.25 > 216.239.59.147: icmp 64: echo request seq 1 1127000935 M * linc 00:48:48.342505 00:05:5d:e6:1e:22 > 00:0d:61:01:35:b3, ethertype IPv4 (0x0800), length 98: IP (tos 0x0, ttl 237, id 0, offset 0, flags [DF], length: 84) 216.239.59.147 > 192.168.2.25: icmp 64: echo reply seq 1 1127000949 M * linc [sorry for the delay but i had to open ICMP forwarding on other gateways] 1127000961 M * linc now with SNAT: 1127000987 M * linc root@newton-root ~# tcpdump -vvnei eth0 icmp 1127000988 M * linc tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 96 bytes 1127000988 M * linc 00:49:41.852256 00:0d:61:01:35:b3 > 00:05:5d:e6:1e:22, ethertype IPv4 (0x0800), length 98: IP (tos 0x0, ttl 64, id 0, offset 0, flags [DF], length: 84) 192.168.2.2 > 216.239.59.99: icmp 64: echo request seq 1 1127000988 M * linc 00:49:42.015956 00:05:5d:e6:1e:22 > 00:0d:61:01:35:b3, ethertype IPv4 (0x0800), length 98: IP (tos 0x0, ttl 237, id 0, offset 0, flags [DF], length: 84) 216.239.59.99 > 192.168.2.2: icmp 64: echo reply seq 1 1127000993 M * linc all seems correct 1127001022 M * Bertl now get an ip on a different vserver 1127001037 M * Bertl (you said you had issues there, no?) 1127001064 M * linc im just rebooting the server :o) 1127001082 M * linc i would love to get that SNAT issue again so i could show you :o( 1127001095 M * Bertl me2, me2 :) 1127001118 M * linc ill try to explain again [while the server is rebooting] 1127001127 M * linc i have, for example, a root server named newton-root 1127001136 M * linc with some different vservers [ex: mail-nss, dns1] 1127001147 M * Bertl why not draw an ascii diagram and upload it? 1127001157 M * linc from those vservers i can connect to *ANYWHERE* 1127001182 M * Bertl (exercices ascii art drawing and gives us a good picture :) 1127001186 M * linc *except* to another vserver on another machine [ciisp vserver on leonardo-root] FROM mail-nss on newton-root 1127001210 M * linc so...to correct the problem, i needed to SNAT packets from mail-nss to the outside world 1127001210 M * Bertl so the ping you want to try is: 1127001235 M * Bertl ping -c 1 -I mail-nss ciisp (an that on newton-root) 1127001239 M * Bertl or 1127001251 M * linc W8T 1127001258 M * linc IT IS HAPPENING AGAIN!!!!!!!!!!!! 1127001260 M * linc ;o))))))))))))))))))) 1127001265 M * linc never been so happy about a problem lol 1127001266 M * Bertl ping -c 1 -I ciisp mail-nss (on leonardo-root) 1127001291 M * linc [allthough i was confusing the ip address of the remote vserver :oP] it isnt called ciisp but ad! forget everything i said! 1127001293 M * linc here is the deal: 1127001343 A * Bertl needs something to eat ... 1127001348 A * linc needs to sleep 1127001359 A * linc hands over a sandwich to herb 1127001444 M * linc hmm.....i miss'spelled the ip address 1127001447 M * linc ...... 1127001459 M * linc dewd, im too tired and i keep fucking'up 1127001466 M * linc im not giving you correct info 1127001475 M * linc we'll better talk on a different day 1127001482 M * linc im *really* sorry for the trouble