1157068946 M * Doener_ humm... 1157069230 M * derjohn n8 folks, I am off now 1157069365 M * matti Sleep well derjohn ;] 1157069375 J * mire ~mire@212-167-222-85.COOL.ADSL.VLine.verat.net 1157069377 M * derjohn ..ZzZZZZzz ;) 1157069978 J * comfrey ~comfrey@h-64-105-215-75.sttnwaho.covad.net 1157070348 J * shedi ~siggi@inferno.lhi.is 1157071254 M * Doener_ Bertl: Didn't find anything... Will be off to bed now, maybe I'll have some time tomorrow when I'm back home and be able to crash it myself ;) 1157071269 M * Doener_ good night everyone! 1157071272 Q * Doener_ Quit: leaving 1157071569 M * brc_ hey 1157071586 M * brc_ Unified vserver: failed to rmdir/unlink `//usr/bin/newgrp.dpkg-tmp': Operation not permitted 1157071595 M * brc_ during apt-get upgrade. I found something on irc logs 1157071600 M * brc_ Anyway to fix ? That happens on upgrades. 1157071817 M * Bertl could be a badly configured guest, i.e. barrier/immutability set where not wanted 1157071885 M * Bertl check some of those files/dirs with showattr 1157071919 M * brc_ all of them are ----UIX 1157071929 M * brc_ Some are ----uiX 1157071929 M * brc_ :) 1157071951 M * Bertl UI is fine for unified files 1157071951 M * brc_ If i run apt-get uprade twice it works. i have to do it for each error. On irc logs there was someone with this error talking with you 1157071978 M * brc_ the other ones are Ui cause are symbolic links, sorry 1157072017 M * Bertl Ui is suspicious 1157072018 M * matti Goodnight all :) 1157072032 M * brc_ Ui are the symbolic links 1157072052 M * Bertl so somebody did set the U flag on those 1157072056 M * Bertl which is wrong :) 1157072174 M * brc_ it was vhashify 1157072178 M * brc_ but the error islike this: 1157072179 M * brc_ 1115138390 M * micah ola: its a problem I worked with bertl on a while back, it only has a problem unlinking a temporary file, and if repeated, it goes away 1157072202 M * brc_ 1115138317 M * micah dpkg: error processing /var/cache/apt/archives/passwd_1%3a4.0.3-31sarge1_i386.deb (--unpack): 1157072203 M * brc_ 1115138320 M * micah failed to rmdir/unlink `//usr/bin/chage.dpkg-tmp': Operation not permitted 1157072214 M * brc_ ==== 1157072233 M * brc_ there are some other people with same problem too but i could not find a solution on those logs 1157072233 M * brc_ hehe 1157072249 M * brc_ 1139441966 M * mugwump failed to rmdir/unlink `//usr/lib/pt_chown.dpkg-tmp': Operation not permitted 1157072249 M * brc_ 1139441970 M * mugwump I hate that 1157072249 M * brc_ 1139441993 M * mugwump damn dpkg trying to chmod u-s before unlinking files for security :) 1157072317 M * Bertl that could be the case too, but IMHO the Ui on links looks more suspicious 1157072360 M * brc_ What is U? The error was not on the linked files but on the real files itself 1157072374 M * brc_ Any solution for thise case ? dpkg trying to chmod u-s before unlinking 1157072384 M * Bertl U is the immutable linkage invert flag 1157072407 M * Bertl it is given together with the I flag to make file immutable but unlinkable 1157072434 M * Bertl if given without the I (immutable) it just makes the file not deleteable 1157072452 M * brc_ ok 1157072540 M * brc_ Maybe mugwump has a solutino for that ? 1157072574 M * Bertl the chmod changes require experimental patches from daniel 1157072632 M * brc_ good to know there will be a solution. This unification thing seems really great 1157072657 M * Bertl yes, solution will be in one of the next devel kernels 1157072687 M * brc_ Yesterday a unified some vservers on a test server and it saved lot of memory 1157072691 M * brc_ a = i 1157072792 M * brc_ Wanna see before/after /proc/meminfo ? :) 1157072802 M * Bertl sure 1157072903 M * brc_ is memtotal/memfree enough ? 1157072909 M * brc_ before starting servers: 1157072910 J * blerk blerk@219.95.6.254 1157072918 M * brc_ MemTotal: 1538920 kB 1157072919 M * brc_ MemFree: 284200 kB 1157072919 M * brc_ Buffers: 216044 kB 1157072919 M * brc_ Cached: 813488 kB 1157072923 M * Bertl welcome blerk! 1157072927 M * brc_ Before unification 15 servers up: 1157072935 M * brc_ MemTotal: 1538920 kB 1157072935 M * brc_ MemFree: 82944 kB 1157072935 M * brc_ Buffers: 239940 kB 1157072935 M * brc_ Cached: 740612 kB 1157072942 M * brc_ After unification: 1157072950 M * brc_ MemTotal: 1538920 kB 1157072950 M * brc_ MemFree: 118940 kB 1157072950 M * brc_ Buffers: 263736 kB 1157072950 M * brc_ Cached: 744344 kB 1157072961 M * brc_ == 1157072967 M * blerk im trying to delete a directory where a vserver was in and i cant rm -rf it, it says its busy, its not mounted or anything, what could be keeping it busy? 1157072983 M * brc_ blerk: Have you checked if the vserver is running ? 1157072991 M * Bertl probably some processes running in that guest 1157073004 M * blerk yeah, i stoped it then unmounted it, then tried to delete it 1157073009 M * blerk hrm 1157073013 M * Bertl blerk: check with vserver-stat 1157073046 M * Bertl ah, you have a separate mount you do on the host for that guest? 1157073083 M * blerk yeah it has/had /vservers/vserver-name, vps doesnt show anything and vserver-stat too 1157073094 M * blerk :P 1157073100 M * Bertl okay, then I can explain to you waht happened :) 1157073119 M * Bertl a) you did mount the filesystem for the guest 1157073136 M * Bertl b) you did start the guest 1157073145 M * Bertl c) you started _another guest :) 1157073174 M * Bertl and here is the problem, the second guest did 'inherit' the mount from the first one 1157073192 M * blerk aha 1157073193 M * Bertl you could have unmounted the partition after starting the guest 1157073206 M * Bertl in which case that would not have had happened 1157073221 M * Bertl or, a simpler approach is to let the tools mount that for you 1157073234 M * blerk so to fix it i have to stop the other guests 1157073237 M * Bertl (by adding the fileszstem to the fstab in the config) 1157073242 M * Bertl no, not necessarily 1157073260 M * blerk i stopped the other one and i can rm -rf it now 1157073262 M * blerk hehehe 1157073269 M * Bertl you can 'enter' the namespace of those 'other' guests (with vnamespace) and umount it there 1157073287 M * Bertl but as you figured, stopping them work too 1157073302 M * blerk cool 1157073311 M * blerk im glad it was fixable without a reboot 1157073350 M * Bertl 99% of all issues are fixable without stopping a guest :) 1157073364 M * blerk next time i have it i do as you said, no stopping 1157073379 M * blerk i been playing about with some distro's in a test server 1157073391 M * blerk this stuff is sweeter than candy 1157073400 M * Bertl nice, what is the host distro if I may ask? 1157073422 M * blerk slack 1157073440 M * Bertl and how did you install the guest? via network? 1157073441 M * blerk with 2.6.17.7-grsec2.1.9-vs2.0.2-rc26 someone here provided me with 1157073483 M * blerk used a tarred up minimal debian install 1157073506 M * Bertl ah, k, you know you can debootstrap debian via network? 1157073507 Q * mire Ping timeout: 480 seconds 1157073515 M * Bertl i.e. the tools support that quite simple 1157073559 M * blerk nice 1157073573 M * blerk im not a true debian user to be honest 1157073596 M * Bertl ah, well, they can install rpm based distros too :) 1157073626 M * blerk i dont have enough aptitute to run debian ;) 1157073706 M * Bertl hehe 1157073751 M * blerk my sandbox vserver has been running since this box was placed in its rack, not a hitch, and i mess about in the sandbox a lot 1157073757 M * blerk really sweet stuff you guys made 1157073774 M * Bertl thanks! 1157073793 M * blerk 8 27 255.9M 123.9M 22m23s00 4m43s38 29d12h48 sandbox 1157073798 M * blerk thank you :P 1157073841 M * Bertl you're welcome! 1157074508 Q * gerrit Ping timeout: 480 seconds 1157074868 J * ntrs__ ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1157074868 Q * ntrs_ Read error: Connection reset by peer 1157075829 J * s0undt3ch_ ~s0undt3ch@bl7-244-1.dsl.telepac.pt 1157076149 Q * s0undt3ch Killed (NickServ (GHOST command used by s0undt3ch_)) 1157076153 N * s0undt3ch_ s0undt3ch 1157077298 M * Bertl okay folks! I'm off for tonight! have a good one everyone! cya! 1157077305 N * Bertl Bertl_zZ 1157077429 M * blerk sleepwell Bertl thanks for the assistance 1157078207 N * nokoya nokoyak 1157078857 M * blerk how can i add a custom mtab into a guest system? if i edit the one inside the guests it gets overwritten when i stop/start the guest.. 1157080062 J * blerk_ blerk@218.208.209.59 1157080065 Q * blerk Read error: Connection reset by peer 1157080163 N * blerk_ blerk 1157080442 M * blerk i cant get df to report properly :( 1157080780 M * cehteh blerk: fstab you mean? 1157080846 M * blerk adding "/dev/root / ext2 rw 0 0" to the guests fstab seems to fix it 1157080865 M * blerk someone needs to bring a strong cup of coffee in here :/ 1157080870 M * cehteh what do you try to do? 1157080879 M * blerk get df to report properly :p 1157080927 M * blerk i was trying it in /etc/vserver/View->PageStyle 1157081337 M * blerk aha 1157081352 M * cehteh anyways ... i recently had the idea to make a LD_PRELOAD library to fix broken apps in vserver 1157081391 M * cehteh would be nice for df to use /proc/mounts instead /etc/mtab (while that it not a general fix) 1157081484 M * cehteh eh and your /dev/root / ext2 rw 0 0 think of course fixes it somewhat too 1157081507 M * blerk its the noob solution :P 1157081520 M * cehteh but i dont know how meaningful the result is when one has many verservers and using the quota system 1157081529 M * blerk i tried the stoff from the weedleaf page but vserver seemed to ignore my custom mtab/fstab in there 1157081556 M * cehteh better query the quota stuff then .. and administrate hardware utilization from the root server 1157081582 M * blerk the guests that need quota get their own partition here 1157081593 M * blerk seems to work fine 1157081636 M * cehteh you cant unify/hashify if using diffrent partitions 1157081680 M * cehteh disk-space wise that might be no problem but think about that the kernel also share pages in memory, like glibc and other apps 1157081724 M * cehteh which means that unified vserver use much less ram and are much faster since the cpu has a far better cache hit ratio 1157081778 M * blerk i iwhsi would understand everythign you just said, if the vservers can rus faster and consume less memory my boss will be happy, tho i dont mind demanding dual/quad cpu systems with 2-4gb of ram ;) 1157081801 M * blerk they run fine on my dual xeons for now 1157081802 M * blerk hehehe 1157081884 M * cehteh yes but if you dont unify them they dont scale :) 1157081904 M * cehteh means adding each vserver guest needs more resources 1157081962 M * cehteh with unifying you can prolly run serveral dozen vservers where without it you can only a handful 1157081969 M * blerk wow 1157081989 M * blerk that is nice, the max amount of vservers we run per box are about 3-5 now 1157082006 M * cehteh .. but that actually depends heavily on the usage and what you run on them 1157082014 M * blerk we use it more for security/seperation than selling virtual servers 1157082048 M * blerk its also handy to setup new boxes quickly by making it d/l a image on boot wich we can pick (dns/mail/web) 1157082054 M * cehteh unifying should have no impact on security (else it would be a vserver bug) 1157082077 M * cehteh what do you run? webservers? 1157082080 M * blerk i see, i dont mind testing/trying it on a sandbox here 1157082083 M * blerk yea mostly 1157082111 M * cehteh if the installations are very similar then they prolly benefit among the best from unifying 1157082169 M * cehteh i thing some guy even ran some houndred vservers on a server just as proof of concept .. dunno exactly 1157082194 M * blerk lol 1157082229 M * blerk we have some boxes that have to be migrated to linux/vserver they run freebsd/jail now 1157082239 M * cehteh but if your cpu's are mostly unloaded with 3-5 vservers then you can likely pack 20 of them with unification and still good performace 1157082255 M * blerk that would be sweet 1157082265 M * cehteh depending on the usage i would say soemthing between 5-30 are sane numbers 1157082274 M * blerk means we can also run several vservers on slightly older hardware 1157082331 M * cehteh of course you loose the benefits of unification if each server is either high cpu intensive, or the installations differ alot, or they are io-intensive 1157082375 M * cehteh but some entry-level web hosting with predefined images for the os scale very well 1157082382 M * blerk we use mostly the same os/image just the job of the server differs, some do mail only other just web 1157082432 M * cehteh well what matters is what the kernel needs to load into memory .. where the biggiest common thing is likely the glibc 1157082475 M * cehteh if you can arrage it that one machine does the webserver jobs and another the mailserver and so on .. prolly better but not really required 1157082537 M * cehteh if you run database servers then you prolly better place them on dedicated machines i would say ... but actually you can just try and figure out what suits you best 1157082860 M * blerk yea 1157082867 M * blerk the sql box has been replaced by a small cluster 1157082881 M * blerk it was nearly impossible to login to it at some point due to overloading 1157082882 M * blerk :P 1157083100 J * jesse_ ~wenchien@59-105-176-11.adsl.static.seed.net.tw 1157083201 N * jesse_ wenchien 1157083617 M * blerk time to fetch some lunch 1157085292 Q * blerk Quit: leaving 1157086956 Q * cehteh Quit: Terminated with extreme prejudice - dircproxy 1.0.5 1157088117 J * ntrs_ ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com 1157088117 Q * ntrs__ Read error: Connection reset by peer 1157088257 Q * balbir Remote host closed the connection 1157089683 Q * comfrey Ping timeout: 480 seconds 1157089931 J * gerrit ~gerrit@c-67-160-146-170.hsd1.or.comcast.net 1157091682 J * olilo hiddenserv@tor.noreply.org 1157092762 J * cehteh ~ct@cehteh.homeunix.org 1157093200 J * mire ~mire@171-167-222-85.COOL.ADSL.VLine.Verat.NET 1157093445 Q * cdrx Ping timeout: 480 seconds 1157094457 Q * Aiken Ping timeout: 480 seconds 1157094530 Q * cehteh Quit: Terminated with extreme prejudice - dircproxy 1.0.5 1157094769 J * meandtheshell ~markus@85-124-232-207.work.xdsl-line.inode.at 1157095701 J * cdrx ~legoater@242.32.96-84.rev.gaoland.net 1157095760 N * nokoyak nokoya 1157095942 J * cehteh ~ct@cehteh.homeunix.org 1157096127 J * dna ~naucki@188-204-dsl.kielnet.net 1157096340 Q * olilo Quit: leaving 1157096615 J * olilo hiddenserv@tor.noreply.org 1157096826 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl 1157097699 J * coocoon ~coocoon@p54A0650D.dip.t-dialin.net 1157099492 J * balbir ~balbir@59.145.136.1 1157100885 Q * shedi Quit: Leaving 1157101020 Q * coocoon Quit: KVIrc 3.2.0 'Realia' 1157102795 Q * cdrx Ping timeout: 480 seconds 1157103823 J * shedi ~siggi@dsl-149-109-85.hive.is 1157107215 Q * olilo Remote host closed the connection 1157107263 J * olilo hiddenserv@tor.noreply.org 1157108009 J * id23 ~id@p50812568.dip0.t-ipconnect.de 1157108026 M * id23 greetings #vserver 1157108537 M * eyck greetings %randomUser 1157108904 M * matti Hello everyone! 1157108905 M * matti ;] 1157109006 M * derjohn hy all! 1157109032 M * derjohn qny news about the "ntrs_ bug" ? 1157109067 M * matti derjohn: You mean, the mysql one? Or some kinda new? 1157109311 M * derjohn matti, the sql-one ... 1157109534 J * dev_null ~arno@dslb-084-056-089-067.pools.arcor-ip.net 1157109537 M * dev_null hi there 1157109571 M * matti Hi. 1157109592 M * dev_null i've got a question, how do I set the default gateway for a vserver? according to http://www.nongnu.org/util-vserver/doc/conf/configuration.html there is no configuration switch for that purpose? 1157109679 J * |DiEgO| ~XOOM@200.230.80.153 1157109800 M * |DiEgO| Hi, would like to know if it has as to limit the space in the virtual server....Somebody knows? 1157109841 M * dev_null ask your isp, else you should know if you're using quotas .. 1157109845 M * derjohn |DiEgO|, _in_ the guest? or the space _for_ the guest ? 1157109865 M * derjohn _in_ the guest: per-guest quota, _for_ the guest disk_limits (easy) 1157109946 M * |DiEgO| yes, QUOTA, more vserver does not supply to some another command? Example: I want to create to vserver with 2GB of HDD 1157109969 M * derjohn |DiEgO|, that's what we call disklimits 1157109983 M * ex dev_null, vservers use gateway from main host. if you need different one, use source routing / routing tables in main node [afaik] 1157109992 M * derjohn you can say guest blabla gets only 2 GB without the need of an extra partition 1157110010 M * dev_null that isn't that good ex 1157110036 M * ex why? (i'm using in such way) 1157110049 M * derjohn |DiEgO|, but you should use recent tools + patches, and one partition for the guests (not /), as you need to mount it with option tagxid and user_xattr 1157110077 M * |DiEgO| disklimits is an instruction that places in the command to vserver when goes to create new a virtual one? 1157110111 M * dev_null ex: I don't mean it doesn't work, I mean that isn't such a fine way 1157110180 M * ex maybe it's not perfect, but I dont know other one.. 1157110191 M * derjohn |DiEgO|, no you add it afterwards in /etc/vservers/guestbla/ .. I sent ypu p-msg with a 5-liner 1157110197 M * dev_null ok, thank you anyway. 1157110215 M * derjohn |DiEgO|, you will find info in the wiki , too 1157111200 M * |DiEgO| derjohn: hey, tkz very much! :D 1157111502 M * |DiEgO| derjohn: to limit the memory ram, as makes? 1157112028 M * derjohn as = virtual memory (more or less) 1157112036 M * derjohn rss = physical memory 1157112365 Q * balbir Quit: Leaving 1157112503 Q * Vudumen Ping timeout: 480 seconds 1157112818 J * Vudumen 4750fd0049@perverz.hu 1157114280 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net 1157114606 Q * kir Quit: Leaving 1157115118 J * kir ~kir@swsoft-mipt-nat.sw.ru 1157115530 N * Belu_zZz Belu 1157115662 J * lilalinux ~plasma@dslb-084-058-192-106.pools.arcor-ip.net 1157115770 J * Pazzo ~thomas@dialin-225136.rol.raiffeisen.net 1157117451 J * [Bluba] hiddenserv@tor.noreply.org 1157118405 Q * [Bluba] Remote host closed the connection 1157118543 Q * cdrx Quit: Leaving 1157119407 Q * FireEgl Ping timeout: 480 seconds 1157120197 J * FireEgl FireEgl@2001:5c0:84dc:1:4:: 1157120392 Q * |DiEgO| Ping timeout: 480 seconds 1157121877 M * dev_null what ccaps do I need for icmp? shouldn't RAW_ICMP be enough? 1157121951 J * lilalinux_ ~plasma@dslb-084-059-001-034.pools.arcor-ip.net 1157122072 Q * ||Cobra|| Remote host closed the connection 1157122302 J * yarihm ~yarihm@84-75-123-221.dclient.hispeed.ch 1157122383 Q * lilalinux Ping timeout: 480 seconds 1157124235 J * Viper0482 ~Viper0482@p5497776F.dip.t-dialin.net 1157125244 M * michal_ dev_null: should be 1157127069 J * bonbons ~bonbons@83.222.36.236 1157127734 Q * Pazzo Quit: Ex-Chat 1157128457 Q * id23 Quit: Leaving 1157129464 Q * shedi Quit: Leaving 1157129483 A * Belu is away (iŽll be back later...) 1157129483 N * Belu Belu_zZz 1157129602 J * gott ~arno@dslb-084-056-106-133.pools.arcor-ip.net 1157129994 Q * dev_null Ping timeout: 480 seconds 1157130116 N * gott dev_null 1157130511 J * zkbrsnie ~zkbrsnie@83-64-146-226.klosterneuburg.xdsl-line.inode.at 1157130543 Q * zkbrsnie 1157130873 J * zkbrsnie ~zkbrsnie@83-64-146-226.klosterneuburg.xdsl-line.inode.at 1157130908 Q * zkbrsnie 1157130922 J * zkbrsnie ~zkbrsnie@83-64-146-226.klosterneuburg.xdsl-line.inode.at 1157131691 N * nokoya nokoyak 1157131749 N * nokoyak nokoya 1157132317 M * doener Hollow: you asked for me? 1157135780 Q * Viper0482 Remote host closed the connection 1157136864 J * shedi ~siggi@inferno.lhi.is 1157136921 Q * Skram Remote host closed the connection 1157136929 J * Skram ~Mark@hermes.sentiensystems.com 1157137141 J * Viper0482 ~Viper0482@p5497776F.dip.t-dialin.net 1157139493 P * Viper0482 und weg 1157139651 M * daniel_hozac doener: i believe he meant to ask you to move your stuff to people.linux-vserver.org. 1157139665 M * daniel_hozac but i guess that'll be hard if you don't have an account yet. 1157139857 M * matti http://www.killernic.com/KillerNic/ 1157139858 M * matti ROTFL 1157140590 M * Wonka hrhr 1157140594 M * Wonka that's a fun part 1157140602 M * Wonka with linux running on it 1157140617 M * Johnnie LOL matti 1157140624 M * Johnnie Hilarious. 1157140907 M * matti ;] 1157140915 M * matti I want those! 1157140921 M * matti I'll -100 ms pings with it! 1157140921 M * matti ;]p 1157140936 M * matti Probably tachyon powered or something ;] 1157141494 J * Beuc ~yo@perso.beuc.net 1157141518 M * Beuc Hey. I saw a TikiWiki demo in the form of a VMWare machine, and I wanna do the same with VServer :) http://tikiwiki.org/tiki-read_article.php?articleId=130 1157141536 M * Beuc Namely, that demo manages to bring up a new virtual interface via DHCP, and I can't manage to do that on a virtual interface (eth0:0) with Linux. 1157141570 M * Beuc Do you know if there's a limitation or if there's some interface parameters to play with? :) 1157141653 M * daniel_hozac you might want to look at how jacques' tools do it, they supposedly support it. 1157141664 M * Beuc will do 1157141968 J * DreamerC_ ~dreamerc@61-217-226-33.dynamic.hinet.net 1157142332 Q * DreamerC Ping timeout: 480 seconds 1157143474 Q * Beuc Ping timeout: 480 seconds 1157143574 J * Beuc ~yo@perso.beuc.net 1157144220 Q * meandtheshell Quit: bye bye ... 1157144290 Q * FireEgl Quit: Bye... 1157144580 Q * lilalinux_ Remote host closed the connection 1157145780 M * Beuc daniel_hozac: apparently Jacques' scripts uses standard dhclient. He asks it to use a different dhcp identifier for each interface, but that doesn't seem to help dhcp (SIOCSIFFLAGS: Cannot assign requested address / Bind socket to interface: No such device) 1157145780 M * Beuc Do you know where one can bug a dhcp developer? ;) 1157147068 Q * bonbons Quit: Leaving 1157148682 N * Bertl_zZ Bertl 1157148686 M * Bertl evening folks! 1157148756 J * Homere ~homere@did75-9-82-229-149-96.fbx.proxad.net 1157148758 M * Homere hello 1157148769 M * Bertl welcome Homere! 1157148854 M * michal_ evening Bertl :) 1157148880 M * michal_ Bertl: may i disturb you evening a bit and ask some offtopic question?;p 1157148906 M * Homere is there a vserver patch for 2.6.18-rc5 ? 1157148924 M * Bertl not yet ... 1157148940 M * Bertl michal_: how OT? :) 1157148941 M * Homere My vserver host is also a FTP server and there is a bug in 2.6.16 TCP networking 1157148951 M * Homere 2.6.17 (not 16) 1157148960 M * Bertl shouldn't that be fixed in 2.6.17.11+ 1157148969 M * Homere not yet 1157148976 M * michal_ Bertl: raid0 releated ;) 1157148988 M * michal_ Bertl: from your experince (i hope you have some with raid 0) 1157148993 M * michal_ *experience 1157148994 M * Homere some routers doesn't respect a 14 year old RFC 1157149003 M * michal_ is there anything like 'too small chunk size'? 1157149031 M * michal_ i'm choosing optimal one all evening adn am wondering - 4/8/16/32 KB. bigger ones actualy degrades performance here. 1157149032 M * Bertl Homere: okay, wouldn't it be easier to add that patch to 2.6.17.11 for example? 1157149066 M * Bertl Homere: there will be a version for 2.6.18-rc*, but I'm not sure when, 2.6.18 is still a moving target 1157149076 M * michal_ 16 seems to be performing quite nice - on both big files (10 GB) and a lot of small files - like, an unpacked kernel tree 1157149224 M * Bertl have you tried striping? and how many disks do you have? 1157149259 M * michal_ yes, it is striping - raid 0 on two disks 1157149322 M * Homere does Lycos still contributes to Vserver ? 1157149364 M * Bertl to some extend, yes 1157149420 M * Homere In France lycos seems bad now, they lost developpers, managers and their vserver help support is now very minimal 1157149436 M * Bertl yes, they moved to germany and armenia, IIRC 1157149663 Q * dna Quit: Verlassend 1157149700 Q * yarihm Remote host closed the connection 1157150157 P * Beuc Leaving 1157150205 M * Bertl michal_: which kind of disks? 1157150218 M * michal_ sataII, 7200rpm, large cache 1157150221 M * michal_ ncq 1157150238 M * Bertl 3GB jumper removed? 1157150251 M * michal_ 3 GB jumper? 1157150267 M * Bertl most sata II ship with a jumper to limit them to 1.5Gbit 1157150279 M * michal_ interesting 1157150289 M * michal_ whre is that jumper?;p 1157150294 M * Bertl well, compatibility :) 1157150300 M * michal_ ic 1157150322 M * Bertl check the spec sheets for your disks 1157150332 M * michal_ i will 1157150341 M * michal_ hm, had no idea about it. thx :) 1157150401 M * michal_ looks like 16KB is a good choice here - every possible tst i'm making here is faster with it. some very interesting - like that 'make cscope' speedup from almost 2 minutes to 40s :) 1157150441 M * Bertl check with bonnie++ 1157150456 M * Bertl IMHO the only relevant check besides kernel compiling :) 1157150470 M * michal_ hm, good idea 1157150477 M * michal_ yes, kernel compiling included 1157150479 Q * dev_null Quit: Verlassend 1157150492 M * michal_ and all that stuff around it too - that's my workinstation anyway ;) 1157150498 M * michal_ *workstation 1157150526 M * michal_ i will have to design some good backup solution - it's a raid0, no chance if something fai....yy...i won't say it ;p 1157150583 M * Bertl well, maybe add another disk, user raid5, will give you a speed boost _and_ some safety 1157150587 M * Bertl *use 1157150600 M * michal_ thining about it too 1157150603 M * michal_ *thinking 1157150607 M * michal_ i cannot type today 1157150626 M * Bertl raid5 over 4 disks is quite nice 1157150886 M * michal_ MB*(N-1) of storage? not that bad... 1157150901 M * michal_ (in case all disks are equal) 1157150947 M * michal_ is current hardware able to handle partity check without troubles? 1157151324 M * michal_ and why 4 actualy?;p 1157151388 M * Bertl well, soft raid5 works pretty fine with 4 disks (sata) and it is a good tradeoff between speed and space 1157151612 M * michal_ hm, 4x250, minus other OS space (not that much ;) minus some xen system volumes 1157151622 M * michal_ minus raid5 overhead 1157151630 M * michal_ still a plenty of space 1157151665 M * michal_ you know - when i'm thinking about it i start to like it 1157151833 M * matti Bertl: Hi. 1157152025 M * Bertl hey matti! 1157152083 P * Homere Client exiting 1157152206 M * Bertl okay, off for now .. back later ... 1157152212 N * Bertl Bertl_oO 1157152369 J * yarihm ~yarihm@84-75-123-221.dclient.hispeed.ch 1157152381 M * yarihm re everyone 1157152517 M * michal_ hey you 1157152520 M * michal_ hey matti 1157152820 M * matti ASUS drives me crazy. 1157152839 M * yarihm what is the build-method one should use if one wants to bootstrap the vserver on its own? i.e. i want to create a more or less empty dir with the configuration in /etc/vservers 1157152849 M * matti Why they removed all informations about discontinued products. 1157152856 M * matti This is really not fair at all. 1157152908 M * michal_ that's just asus 1157152931 M * michal_ i had only one part from them - and old 36x cdrom, working all the time, very solid. 1157152999 N * Belu_zZz Belu 1157153032 M * yarihm ah, skeleton 1157153059 M * matti michal_: Well, this was not really "made by asus" - only some oem with their logo probably. But good to hear if you were satisfied. 1157153073 M * matti I have very odd motherboard. 1157153095 M * michal_ i was - but it was a long (few years, recall me where 36x cdroms used to be top products) 1157153117 M * michal_ you are probably having problems with their mb? 1157153119 M * matti ASUS A7V revision 1.04, and I cannot find anything on such particular m/b. 1157153718 M * derjohn matti, besdides the bios I have (old) A7v in prodiction here , too 1157153732 M * matti You also have the A7V? 1157153748 M * derjohn matti, they, this was old athlon, socket a ? 1157153771 M * matti My revision have all dip-switches. Which is quite odd... 1157153773 M * matti Yes. 1157153881 M * derjohn Any news about the possible rc31 threads bug ? 1157153949 M * matti No. 1157153953 M * matti At least I don't know. 1157154178 Q * DreamerC_ Quit: leaving 1157154196 J * DreamerC ~dreamerc@61-217-226-33.dynamic.hinet.net 1157154877 M * michal_ ok 1157154887 M * michal_ raid a bit reorganised 1157154892 M * michal_ temporary ;) 1157154919 M * michal_ i should not repartition disk and make/restore backups and change raid layout at 2am 1157154922 M * michal_ i know ;) 1157154949 M * michal_ matti: btw, could you send me that log we've been talking about - about jfs @ gentoo-hardened 1157154967 M * michal_ i'm dying to read it ;) 1157154984 M * michal_ jfs seems very interesting to me after some tests i've been making month ago...