1173485721 Q * meandtheshel1 Quit: Leaving. 1173487157 N * Bertl_zZ Bertl 1173487169 M * Bertl evening folks! 1173488295 Q * rob-84x^ Ping timeout: 480 seconds 1173488448 J * rob-84x^ ~rob@submarine.ath.cx 1173489251 J * shedi ~siggi@ftth-237-144.hive.is 1173490972 N * fosco_ fosco 1173491355 Q * rob-84x^ Ping timeout: 480 seconds 1173492831 Q * softi42 Ping timeout: 480 seconds 1173493857 J * softi42 ~softi@p549D674A.dip.t-dialin.net 1173494733 Q * infowolfe Ping timeout: 480 seconds 1173495008 J * b0c1 boci@pool-2893.adsl.interware.hu 1173495013 M * b0c1 hi 1173495033 M * b0c1 Ummm... how can I configure the default gw to the guest? 1173495575 Q * b0c1 Quit: Távozom 1173495666 Q * shedi Ping timeout: 480 seconds 1173496395 J * shedi ~siggi@ftth-237-144.hive.is 1173496604 J * infowolfe ~infowolfe@c-67-164-195-129.hsd1.ut.comcast.net 1173497780 M * Bertl okay, off to bed now ... have a good one everyone! 1173497787 N * Bertl Bertl_zZ 1173497925 J * Amsat ~Amsat@MW-ESR1-72-49-76-110.fuse.net 1173498037 P * Amsat 1173499032 Q * infowolfe Ping timeout: 480 seconds 1173499698 Q * ensc Ping timeout: 480 seconds 1173499885 J * ensc ~irc-ensc@p54B4E9C9.dip.t-dialin.net 1173501060 J * infowolfe ~infowolfe@c-67-164-195-129.hsd1.ut.comcast.net 1173503077 J * id23 ~id@203.177.252.98 1173503081 M * id23 hi 1173503799 J * DoberMann_ ~james@AToulouse-156-1-56-20.w90-16.abo.wanadoo.fr 1173503905 Q * DoberMann[ZZZzzz] Ping timeout: 480 seconds 1173508627 Q * id23 Ping timeout: 480 seconds 1173509298 J * id23 ~id@203.177.252.98 1173510495 Q * vasko Ping timeout: 480 seconds 1173510510 J * vasko ~vasko@unreal.rainside.sk 1173510510 A * vasko is gone. Gone since Wed Feb 28 16:04:00 2007 1173510821 Q * FireEgl Quit: ... 1173512922 J * dna ~naucki@p54BCE712.dip.t-dialin.net 1173518254 J * _dmax ~semaj@81.193.60.240 1173518276 Q * dmax Read error: Connection reset by peer 1173518281 N * _dmax dmax 1173518325 J * meandtheshell ~markus@85-124-233-229.work.xdsl-line.inode.at 1173519542 J * b0c1 boci@pool-2893.adsl.interware.hu 1173519544 M * b0c1 morning 1173519547 M * arachnist hi 1173519579 M * b0c1 How can I set up a default gw for my vserver? 1173519585 M * b0c1 I started and I can entered 1173519593 M * b0c1 but I can't ping external addresses 1173519605 M * b0c1 route say: Permission denied 1173519640 M * arachnist sysctl net.ipv4.conf..forwarding=1 1173519668 M * b0c1 in the host os 1173519671 M * b0c1 that's ok 1173519672 M * arachnist yup 1173519706 M * b0c1 My problem I can't set up the gw in the guest 1173519798 J * bonbons ~bonbons@83.222.38.146 1173520021 N * DoberMann_ DoberMann 1173520466 M * b0c1 so how can I set up guest default route gw? 1173521726 Q * id23 Quit: Leaving 1173522256 M * b0c1 Any idea? 1173522718 J * zLinux_ ~zLinux@88.213.12.25 1173522926 J * rob-84x^ ~rob@submarine.ath.cx 1173523033 Q * zLinux Ping timeout: 480 seconds 1173523143 M * Guy- b0c1: you need to set up routing on the host 1173523158 M * b0c1 hmm 1173523161 M * b0c1 really? 1173523163 M * Guy- yes 1173523166 M * b0c1 thnx 1173523171 M * b0c1 I try... 1173523175 M * b0c1 another question 1173523177 M * b0c1 in the quest 1173523184 M * b0c1 /etc/init.d 1173523193 M * b0c1 which script required? 1173523194 M * Guy- if the guest IP is in the same subnet as the host's IP, it can use the same default gw; otherwise, it gets slightly more complicated 1173523216 M * Guy- you can throw out all initscripts that are hardware related, for a start 1173523228 M * Guy- and mounting, and fsck, and network 1173523246 M * b0c1 Guy-: yepp the guest use 192.168.1.2 and I pull up in the host 192.168.1.1, the host have another ip and gw 1173523251 Q * m`m`h Ping timeout: 480 seconds 1173523261 Q * infowolfe Read error: No route to host 1173523261 M * b0c1 I create a routing for the subsystem 1173523326 M * b0c1 Guy-: ok, but what's the hw related script? alsa*, console*, hwclock, mount* 1173523327 M * b0c1 ? 1173523336 M * b0c1 (ok, mount is not ;) 1173523345 M * Guy- but you don't need mount either :) 1173523387 M * b0c1 yepp 1173523406 M * Guy- I suggest you throw everything out that you are not certain you need 1173523417 M * b0c1 ok but, boot* and other? 1173523426 M * b0c1 glibc.sh ? 1173523432 M * b0c1 makedev and others... 1173523459 M * Guy- have you looked at what glibc.sh does? it just prints a warning if your kernel is too old 1173523486 M * Guy- makedev won't work in a guest anyway (unless you give it appropriate capabilities, but normally you wouldn't) 1173523511 M * Guy- basically all you need is the initscripts that start your services, and perhaps cron, syslogd and such (but not klogd) 1173523518 M * b0c1 ok 1173523526 M * b0c1 sysklogd? 1173523526 M * b0c1 :) 1173523535 M * b0c1 hihi 1173523555 M * Guy- the script is called that, but I think it's possible to not start klogd (you can even uninstall it) 1173523585 J * ema ~ema@lart.galliera.it 1173523630 M * b0c1 Guy-: I have klogd and sysklogd :) 1173523635 M * Guy- btw, there is some vserver related stuff in Hungarian at https://wifi.tmit.bme.hu/unixlinux/index.php/Vserver 1173523659 M * Guy- you might want to look at it and perhaps add to it if you find something out that's missing 1173523674 M * Guy- (I wrote that) 1173523689 M * b0c1 500 Internal Server Error 1173523691 M * b0c1 :) 1173523713 J * m`m`h ~simba@deb30.mgts.by 1173523728 M * Guy- try again 1173523738 M * Guy- you got that because you didn't accept the certificate quickly enough :) 1173523764 M * b0c1 Guy-: :) 1173523794 M * b0c1 great guide 1173523801 M * Guy- I wouldn't go that far 1173523915 M * Guy- but you can help improve it :) 1173524016 M * b0c1 Guy-: a little question 1173524053 M * b0c1 Guy-: my host have an external ip (with own gw), and I pull up an 192.168.1.1 for alias eth0 1173524058 M * b0c1 my guest use 192.168.1.2 1173524064 M * b0c1 how can I do the route? 1173524102 M * daniel_hozac iptables -t nat -A POSTROUTING -s 192.168.1.2 -j SNAT --to 1173524105 M * b0c1 in the host route table I can't see the vserver ip etc... 1173524120 M * b0c1 daniel_hozac: that's ok, but I mean the guest not see the external network 1173524137 M * daniel_hozac what? 1173524138 M * Guy- b0c1: should the guest use the same gateway as the host or a different one? 1173524168 M * Guy- daniel_hozac: I think I can sort this out; we can drop back to Hungarian in msg if necessary :) 1173524185 M * daniel_hozac using different gateways doesn't make sense most of the time. 1173524215 M * Guy- well, if you want to pretend that the guest is a different computer, and you hook it up to a different physical network, then yes 1173524223 M * b0c1 http://rafb.net/p/FRZgMF47.html 1173524236 M * b0c1 the first two is the host, the last one is the guest 1173524268 M * b0c1 I want to see the internet from the guest 1173524279 M * Guy- b0c1: through the same gateway, yes? 1173524288 M * b0c1 now it's not work, I cant ping 217.20.131.2 example 1173524296 M * b0c1 Guy-: same gw? 1173524301 M * b0c1 Guy-: ohh 1173524303 M * b0c1 yess... 1173524313 M * b0c1 but the guest doesn't have external ip 1173524316 M * b0c1 only internal 1173524324 M * Guy- no problem 1173524328 M * b0c1 okie 1173524333 M * daniel_hozac that's why you NAT it. 1173524349 M * Guy- exactly 1173524370 M * b0c1 daniel_hozac: hmm... OHH! SNAT! :) 1173524372 M * Guy- b0c1: you need the SNAT rule daniel_hozac gave you or a similar one with MASQUERADE 1173524376 M * b0c1 okok 1173524379 M * b0c1 I use shorewall 1173524384 M * b0c1 I configure this :) 1173524387 M * b0c1 thnx a lot 1173524395 M * Guy- nincsmit :) 1173524412 M * b0c1 lol, koszi 1173524413 M * b0c1 :D 1173524421 M * daniel_hozac MASQUERADE might not work depending on your kernel though. 1173524497 Q * rob-84x^ Quit: That's it for today 1173524503 J * rob-84x^ ~rob@submarine.ath.cx 1173526041 Q * softi42 Ping timeout: 480 seconds 1173526132 J * softi42 ~softi@p549D674A.dip.t-dialin.net 1173526455 Q * rob-84x^ Ping timeout: 480 seconds 1173526994 A * waldi don't like whitespace changes. at least when they produce failed patches ... 1173527116 M * daniel_hozac it's getting removed for 2.2.0. 1173527128 M * daniel_hozac (generally always gets removed before a release) 1173527971 M * waldi 2.2.0-rc15 breaks on 2.6.20.2 because of such a change 1173528006 M * waldi boah, i think the power hypervisor have a bug. my linux reports 50% stolen cpu time 1173528468 Q * michal` Ping timeout: 480 seconds 1173528915 J * michal` ~michal@www.rsbac.org 1173529472 M * b0c1 Ehh 1173529478 M * b0c1 I never do SNAT in shorewall... :D 1173531516 M * b0c1 daniel_hozac: umm a little question 1173531529 M * b0c1 daniel_hozac: How can I do the name resolv? 1173531538 M * daniel_hozac hmm? 1173531547 M * b0c1 I need forward the port 53 to the internal? 1173531562 M * b0c1 but if I do that how can I resolve in the host? 1173531571 M * b0c1 (internal => guest) 1173531587 M * daniel_hozac you want to run a DNS server in the guest? 1173531592 M * b0c1 no 1173531603 M * b0c1 but if I want to ping index.hu host 1173531606 M * b0c1 in the guest 1173531613 M * b0c1 the package not find the way :) 1173531626 M * b0c1 so I can ping by ip 1173531631 M * b0c1 but I can't by hostname 1173531652 M * daniel_hozac have you created a /etc/resolv.conf file in the guest that points to your nameservers? 1173531700 M * b0c1 both guest and host resolv.conf have nameserver 212.40.68.21 1173531783 M * daniel_hozac and ping -I 192.168.1.2 index.hu on the host works as expected? 1173531842 M * b0c1 Grr 1173531845 M * b0c1 sorry my fault 1173531854 M * b0c1 the host contain nameserver 127.0.0.1 1173531863 M * b0c1 (run bind) 1173531869 M * b0c1 and it's copyed to the quest :D 1173531883 M * b0c1 bad cow, bad cow :) 1173532552 N * Bertl_zZ Bertl 1173532557 M * Bertl morning folks! 1173532921 Q * zLinux_ Ping timeout: 480 seconds 1173534268 Q * softi42 Read error: Connection timed out 1173535227 M * b0c1 Hmmm, where can I set my guest maximal disk space? 1173535243 M * b0c1 I installed util-vserver but it's not created all configuration files :| 1173535265 M * Bertl you need to have file tagging enabled to do so 1173535281 M * Bertl i.e. you need a partition for the guest(s) with tagxid enabled 1173535293 M * b0c1 hmm? 1173535310 M * b0c1 Bertl: ok, tell me how can I do, step-by-step pls... 1173535316 M * b0c1 I have an installed guest 1173535337 M * Bertl http://oldwiki.linux-vserver.org/Disk+Limits 1173535365 M * b0c1 thnx 1173535380 J * softi42 ~softi@p549D674A.dip.t-dialin.net 1173535380 M * Bertl you can basically leave out the last steps with recent util-vserver, as it does the space calculations 1173535407 M * Bertl i.e. you just set the limit for inodes and blocks according to the util-vserver documentation 1173535495 M * daniel_hozac which last steps are you referring to? 1173535521 M * daniel_hozac the notes at the bottom? 1173535533 M * Bertl yep 1173535559 M * daniel_hozac ok, yeah. i didn't feel comfortable just removing that when i updated it for 0.30.210. 1173535568 M * Bertl that's fine 1173535589 M * Bertl we'll rework that on the new wiki .... 1173535595 M * daniel_hozac yep... 1173535611 Q * shedi Quit: Leaving 1173535641 M * b0c1 Bertl: I not want to use separated partitions, but only disk limit :) 1173535663 M * Bertl yep, that's the way to do it ... 1173535684 M * daniel_hozac you're recommended to have a separate partition for your guests though. 1173535686 M * b0c1 but the /dev/whatever is use separated partition? 1173535687 M * Bertl the tagging is required to tell the files apart (i.e. which one belongs to what context) 1173535703 M * b0c1 hmm 1173535710 M * Bertl i.e. you can put all guests on a single partition 1173535713 M * daniel_hozac i.e. /vservers should not be on the same filesystems as /. 1173535720 M * Bertl but that one should be different from the host system 1173535735 M * b0c1 grat... now it's in the host partition 1173535749 M * b0c1 okok, I understand :| 1173535754 M * b0c1 I move it to another... 1173535820 M * Bertl the reason behind that is, that the tagging should not be enabled for the host filesystem, because it might cause 'unexpected' behaviour when host files get tagged 1173535858 M * Bertl daniel_hozac: btw, I did a few changes to the config defaults, will upload shortly for verification 1173535869 M * daniel_hozac okay. 1173536305 M * b0c1 tagxid work with xfs correctily? 1173536316 M * Bertl yes 1173536333 M * Bertl but disk limits don't IIRC :) 1173536366 M * b0c1 :) 1173536403 M * Bertl but that will be added soon I think, we now have enough folks interested in testing 1173536628 Q * Greek0 Remote host closed the connection 1173536723 J * Greek0 ~greek0@85.255.145.201 1173536742 M * Bertl daniel_hozac: okay, uploaded the patches I had here ... usual place 1173536814 M * Bertl I'd also appreciate comments to the following hunks: http://vserver.13thfloor.at/Stuff/OLPC/split-olpc-vs22x.02.2/99_under_investigation.diff 1173536879 M * daniel_hozac looks fine to me. 1173536898 M * daniel_hozac second hunk is from the memory split patches, IIRC. 1173536922 M * Bertl yes, but we removed them some time ago, no? 1173536929 M * daniel_hozac yep. 1173536939 M * Bertl what about the __iget? 1173536982 M * daniel_hozac nothing pops up about that, but grep shows it being used by vroot. 1173537037 M * Bertl ah, that'd explain it 1173537056 M * Bertl +#include in kernel/sys.c makes me wonder 1173537090 M * Bertl the init/Makefile part is definitely obsolete for 2.6.20+ 1173537101 M * daniel_hozac 2.6.19+ even, no? 1173537109 M * Bertl probably, haven't checked yet 1173537116 M * daniel_hozac (i assume that was obsoleted by the uts namespace) 1173537132 M * Bertl by the proper fixup for the banner in /proc 1173537245 M * daniel_hozac the kernel/sys.c hunk doesn't make much sense, IMHO. 1173537360 M * daniel_hozac it seems to have been around for a really long time too. 1173537374 M * daniel_hozac even 1.2 has it. 1173537489 M * daniel_hozac probably left overs from then when vs_reboot was in kernel/sys.c. 1173537545 M * Bertl okay, I reverted the sys, the init/makefile and the DYN_BASE for now 1173537560 M * Bertl what about the virt handler? 1173537783 M * Bertl I think we don't need that anymore, no? 1173537917 Q * Hollow Ping timeout: 480 seconds 1173538065 Q * phreak`` Ping timeout: 480 seconds 1173538293 M * daniel_hozac i guess not. 1173538338 M * Bertl what about he fake_table part? 1173538375 M * daniel_hozac if mainline doesn't have problems with it, i don't see why we would. 1173538400 M * Bertl IIRC, we added that when we were tracking down the virthandler issues back then 1173538417 M * daniel_hozac ah, so removing it should be fine then. 1173538445 M * Bertl not sure I remeber correctly though 1173538619 M * Bertl I don't think we modify proc_dostring nowadays 1173538644 M * Bertl so the hunk should be obsolete by now ... 1173538786 M * daniel_hozac if we do, i can't see it. 1173538838 M * Bertl so we drop that too 1173538863 M * Bertl what about the lock file checks? (vxd_assert) 1173538913 M * daniel_hozac are they wrong somehow? 1173538990 M * Bertl no, just the question remains: do we really need that check? do we benefit from having it? 1173539011 M * Bertl btw, the next hunk, the um new_utsname is unused by now 1173539016 M * daniel_hozac well, might help with debugging IMHO. 1173539018 M * daniel_hozac yeah, drop it. 1173539053 M * Bertl okay, we leave the assert in then 1173539075 M * daniel_hozac we might want to make it depend on some VXD_CBIT too though. 1173539079 M * Bertl the mm_init hunks somewhat confuse the hell out of me 1173539127 M * daniel_hozac :q 1173539132 M * daniel_hozac d'oh, heh. 1173539216 M * daniel_hozac we modify set_mm_counter to deal with the context accounting too, don't we? 1173539327 M * daniel_hozac i guess __set_mm_counter would be more appropriate... 1173539445 J * phreak`` ~phreak``@88.198.108.67 1173539936 M * Bertl I do not understand why we completely remove the set_mm_counter in the init 1173539976 J * Hollow ~hollow@85.10.237.60 1173540016 M * b0c1 hmm 1173540018 M * b0c1 another question 1173540033 M * b0c1 I don't have shadow file in vserver /etc dir... 1173540038 M * daniel_hozac i don't get it either. 1173540040 M * b0c1 so how can I authenticate? 1173540047 M * daniel_hozac /etc/passwd? 1173540058 M * b0c1 I want to tranfer my old users to the vserver 1173540062 M * b0c1 but it's used shadow file 1173540081 M * daniel_hozac might want to run pwconv though. 1173540104 M * b0c1 wow 1173540106 M * b0c1 thnx 1173540109 M * daniel_hozac Bertl: i guess it's because mm_alloc memset(0)s it first? 1173540131 M * daniel_hozac (that doesn't handle dup_mm though, so i still don't get it) 1173540260 Q * ema Quit: leaving 1173540272 M * daniel_hozac __set_mm_counter seems like the right thing to do, IMHO. 1173540880 M * Bertl instead of the 'removed' set_mm_counter, yes? 1173540977 M * daniel_hozac yeah. 1173541186 M * Bertl okay, we can do that, but we need to do some testing on that 1173541199 M * Bertl as for sure it isn't an obvious change :) 1173541234 M * daniel_hozac if anything, i think the current behaviour would be broken. 1173541278 M * daniel_hozac as dup_mm would get file_rss/anon_rss from the oldmm. 1173541411 M * daniel_hozac or hmm, no. 1173541435 M * daniel_hozac or yes, but only without CONFIG_MMU. 1173541469 M * Bertl so, on mmu less systems we now have broken behaviour? 1173541511 M * daniel_hozac IMHO, yep. 1173541565 M * Bertl any idea why we inlcude in alpha/mm/init? 1173541609 M * daniel_hozac none whatsoever. 1173541632 M * daniel_hozac we don't even patch linux/pagemap.h 1173541802 M * Bertl well, we could change something in init.c which requires that 1173541868 M * daniel_hozac that's the only hunk for arch/alpha/mm/init.c. 1173541879 M * daniel_hozac (in 2.6.20.1-vs2.2.0-rc15, at least) 1173541969 M * daniel_hozac i guess we'll have to ask Aiken if removing it breaks anything. 1173542016 M * daniel_hozac the fs/proc/generic.c hunk looks fine to me, what's it doing there? should it be vs_inode.h (assuming one existed)? 1173542177 M * Bertl yes, we don't have vs_inode yet 1173542348 M * Bertl okay, uploaded delta-vs220-clean01.diff (what we agreed upon :) 1173542365 M * Bertl doesn't include the __set_mm_counter() stuff yet 1173543168 J * fluor- ~fluor@tanneries.squat.net 1173543185 M * fluor- hi there 1173543192 M * daniel_hozac hello 1173543204 M * fluor- I'm currently designing a new server setup, 1173543225 M * fluor- and wondering where to go in between xen and vserver, with the extra option of vserver within xen 1173543271 M * fluor- I'd rather stick to a vserver only setup, but after having a look at xen's ressource limitation features, I'm nearly ready to share the doubt with my colleague 1173543292 M * daniel_hozac hmm? 1173543293 M * fluor- so my first question is: what kind of ressource control/limitation feature does vserver have, and what about future plans on that? 1173543311 M * daniel_hozac http://linux-vserver.org/Resource_Limits 1173543320 M * daniel_hozac http://linux-vserver.org/CPU_Scheduler 1173543320 M * fluor- oh thanks, didn't see this one 1173543353 Q * TrueBrain Ping timeout: 480 seconds 1173543365 M * daniel_hozac what limits are you interested in? 1173543408 M * fluor- CPU and memory 1173543434 M * fluor- for instance, if MySQL is going wild and eating all ressources (cpu/mem); 1173543441 M * fluor- it would be nice to be able to calm it down 1173543453 M * fluor- (mysql being a vserver of its own) 1173543462 M * daniel_hozac wouldn't be a problem at all. 1173543550 J * TrueBrain truelight@openttd.org 1173543567 M * fluor- daniel_hozac: how would I do that for a precise vserver context? 1173543732 M * daniel_hozac do what? 1173543742 M * daniel_hozac configure the resource limits/CPU scheduler? 1173543784 M * fluor- for instance 1173543826 M * daniel_hozac the resource limits has an example at the bottom of the page. 1173543889 M * daniel_hozac for the CPU scheduler, something like mkdir -p /etc/vservers//sched; echo 10 > /etc/vservers//sched/fill-rate; echo 50 > /etc/vservers//sched/interval would give the guest 20% CPU-time. 1173543998 M * daniel_hozac (the CPU scheduler needs support from the kernel though, so make sure it has HARDCPU enabled, and preferably idle time as well) 1173544053 M * fluor- it would be great to be able to configure that through a web frontend - I wonder if openvps or vsmon do that 1173544199 M * daniel_hozac i believe there is an #openvcp channel. 1173544218 M * fluor- daniel_hozac: indeed :) 1173544251 M * daniel_hozac (not sure why, but...) 1173544341 J * FireEgl Proteus@2001:5c0:84dc:1:211:9ff:feca:b042 1173544507 M * fluor- another question: it has been said to me that vservers were vulnetable to, say, fork-bombs, 1173544531 M * fluor- that it would alter other vservers and the host systems, since the kernel is shared among all virtual machines 1173544674 M * Bertl not really, unless you configure them to be vulnerable 1173544719 M * Bertl or to be fair, if you do not set proper limits, fork bombs might bite you :) 1173544730 M * Bertl (very similar to any linux system) 1173544824 M * arachnist fork bombs might bite you regardless of running xen/vserver/openvz/whatever if you don't set proper limits 1173544838 M * Bertl yes, that's what I tried to say :) 1173544850 M * fluor- :) 1173545371 M * sannes patch-2.6.20-vs2.2.0.ipv6-pre4.diff latest ipv6 patch? 1173545428 M * daniel_hozac yep. 1173545452 M * sannes hm, think I'm going to take it for a spin :) 1173545509 M * daniel_hozac should be fine for 2.6.20.1-vs2.2.0-rc15, if not, let us know. 1173545511 M * sannes what does -pre4 mean? (I know it is experimental code, but still) 1173545551 M * daniel_hozac bonbons' naming has always confused me a bit, but IIUC that means it's against 2.6.20-vs2.2.0-pre4. 1173545620 M * Bertl there will be an rc16 in a few minutes, so maybe check that out too 1173545707 M * sannes ah, changes between rc15-rc16 ? If it is only the 2.6.20.2 Makefile.rej and irq.c.rej I'm already ignored those (version and whitespace) rejects .. 1173545747 M * daniel_hozac lots of cleanups and a few fixes. 1173545852 M * sannes ah, so I'll just sit tight then :P 1173546361 Q * pusling Ping timeout: 480 seconds 1173546787 M * Bertl test compile should have finished in a few minutes 1173547976 M * daniel_hozac Bertl: oh, did you ever get around to trying that Ubuntu initscript fix? 1173548000 M * Bertl no, sorry, I left before I could give it a try 1173548009 M * daniel_hozac okay, np, i'll test it on monday. 1173548820 M * b0c1 Hmm 1173548824 M * b0c1 [42953104.230000] EXT3-fs: Unrecognized mount option "tagxid" or missing value 1173548839 M * daniel_hozac what kernel is that? 1173548858 M * b0c1 Ohh 1173548861 M * b0c1 holy shit... 1173548895 M * b0c1 but why? 1173548902 M * b0c1 I set fallback 1 2 1173548909 M * b0c1 and grub default to 0 1173549044 M * Bertl sannes: okay, rc16 is up 1173549075 M * sannes Bertl: perfect, just finnished a bugreport on Gentoos 6to4 initscript :P 1173549154 M * daniel_hozac Bertl: hmm, no update to 2.6.20.2? 1173549161 M * daniel_hozac or 2.6.19.7? 1173549174 M * Bertl I just checked a few minutes ago, then 2.6.20.2 wasn't there 1173549181 M * Bertl 2.6.19.7 is work in progress 1173549183 M * daniel_hozac where? it's been out for a few days, IIRC. 1173549192 M * daniel_hozac patch-2.6.19.7.bz2 03-Mar-2007 05:29 62K 1173549200 M * daniel_hozac patch-2.6.20.2.bz2 09-Mar-2007 19:08 37K 1173549215 M * daniel_hozac (okay, so just since last night) 1173549254 M * sannes on rc15 it was only version and a whitespace in irq.c .. so shouldn't be anything huge .. 1173549265 M * daniel_hozac whitespace should be fixed with rc16. 1173549286 M * Bertl anyway, not a big deal to update to 2.6.20.2 too 1173549310 M * sannes rc16 only version change in Makefile 1173549325 M * sannes daniel_hozac: ah, heh, yes :) 1173549342 M * daniel_hozac i was just wondering as you usually update to the new versions before i've even heard of them :) 1173549362 M * Bertl yeah, as I said, I checked, but seems I missed 2.6.20.2 1173549504 M * daniel_hozac well, i almost prefer this as interdiff will work ;) 1173549538 M * Bertl I decided to do all the mainline changes in a separate diff 1173549544 M * daniel_hozac ah, okay. 1173549558 M * Bertl i.e. the 2.6.19.5 was intentional 1173549573 M * Bertl 2.6.19.7 is now test compiling 1173549574 M * daniel_hozac 2.6.19.7 changes that much? 1173549581 M * Bertl (as is 2.6.20.2) 1173549588 M * Bertl no, not really that many changes 1173549706 M * daniel_hozac ok, i'll wait with updating the wiki then. 1173549762 M * b0c1 ok, now I mounted with tagxid 1173549809 M * b0c1 the /etc/vservers/guestname/context is required? 1173549828 M * Bertl yes 1173549857 M * Bertl otherwise the files change ownership on every guest startup 1173549860 Q * Johnsie Remote host closed the connection 1173549891 M * b0c1 ahhm 1173549918 M * b0c1 ok, so I create a context file with .... (?) a number or a tag id? 1173549982 M * b0c1 I think number :) 1173549996 M * b0c1 the value is unique by guest? 1173550024 M * daniel_hozac yes. 1173550029 N * DoberMann DoberMann[PullA] 1173550039 M * b0c1 and it's come from my mind? :) 1173550045 M * b0c1 ur /dev/urandom ? 1173550045 M * b0c1 :D 1173550051 M * b0c1 ur=or 1173550061 M * daniel_hozac whatever you like. 1173550070 M * b0c1 ok 1173550080 M * daniel_hozac personally i use the last octet of the IP address as the xid. 1173550086 M * b0c1 hmmm 1173550089 M * b0c1 good idea 1173550167 M * b0c1 Hmm 1173550170 M * b0c1 something bad 1173550172 M * b0c1 chxid -c higher.hu -R /vservers/higher.hu/ 1173550179 M * b0c1 /vservers/higher.hu/: Invalid argument 1173550237 M * daniel_hozac you sure you mounted it with tagxid? 1173550250 M * daniel_hozac what does lsxid /vservers/higher.hu say? 1173550263 M * daniel_hozac and even more importantly, what does grep tagxid /proc/mounts return? 1173550299 M * b0c1 daniel_hozac: yepp 1173550309 M * b0c1 !!ERR!! /vservers/higher.hu 1173550334 M * b0c1 I found the solution 1173550341 M * b0c1 remounted 1173550343 M * b0c1 now work 1173550352 M * daniel_hozac so, it wasn't mounted with tagxid. 1173550374 M * b0c1 possible, but it's in the fstab 1173550375 M * b0c1 :| 1173550386 M * b0c1 /dev/hdc1 /vservers/higher.hu ext3 defaults,tagxid 0 2 1173550389 M * b0c1 :) 1173550433 M * daniel_hozac hmm, i thought you wanted tagging for disk limits? 1173550861 M * b0c1 daniel_hozac: Yes I think in this... 1173551013 Q * michal` Ping timeout: 480 seconds 1173551407 J * michal` ~michal@www.rsbac.org 1173551889 J * onion jughead@virasto.com 1173551990 M * onion Hi, is it safe to patch 2.6.20.1-vs2.2.0-rc16.diff against 2.6.20.2. Only Makefile patch hunk failed... 1173552032 M * onion There were few offsets also... 1173552039 M * daniel_hozac 2.6.20.2-vs2.2.0-rc16 should be uploaded shortly. 1173552054 M * onion k, ;) thx 1173552089 M * b0c1 daniel_hozac: can I restict how many cpu resource use the guest? 1173552096 M * b0c1 daniel_hozac: example maximum 20% 1173552114 Q * almak Remote host closed the connection 1173552120 M * daniel_hozac 17:24 < daniel_hozac> for the CPU scheduler, something like mkdir -p /etc/vservers//sched; echo 10 > /etc/vservers//sched/fill-rate; echo 50 > /etc/vservers//sched/interval would give the guest 20% CPU-time. 1173552129 M * daniel_hozac http://linux-vserver.org/CPU_Scheduler 1173552149 M * b0c1 LOL 1173552150 M * b0c1 thnx 1173552152 M * b0c1 master 1173552152 M * b0c1 :D 1173552228 M * b0c1 daniel_hozac: and if I modify, I need restart the guest? 1173552285 M * daniel_hozac remember that it needs the hard CPU support in the kernel, as well as the sched_hard flag in the flags file. 1173552327 M * daniel_hozac well, restarting the guest would apply it, but so would vsched --xid --force --dir /etc/vservers//sched 1173552383 M * b0c1 ok 1173552396 M * b0c1 I use precompiled kernel :) 1173552407 M * daniel_hozac (if you haven't set the flag before, you'd need to do that as well with vattribute --xid --set --flag sched_hard) 1173552412 M * b0c1 so I dunno it's contain the hard flag, maybe yes... 1173552415 M * daniel_hozac Debian's? 1173552420 M * b0c1 ubuntu 1173552434 M * daniel_hozac hmm, the uniklu one? 1173552471 M * b0c1 dunno 1173552473 M * b0c1 I think not 1173552476 M * b0c1 http://dpkg.u32.net/ubuntu 1173552522 M * daniel_hozac first time i've ever heard of that... 1173552588 N * phedny__ phedny 1173552596 M * b0c1 :) 1173552600 M * b0c1 http://dpkg.u32.net/ubuntu/dists/edgy/vserver/ 1173552683 M * daniel_hozac well, no idea then. look in /boot/config-`uname -r` for VSERVER, maybe. 1173552689 M * daniel_hozac (or /proc/version.gz) 1173552741 M * b0c1 # CONFIG_VSERVER_HARDCPU is not set 1173552752 M * daniel_hozac so, there you go. 1173552764 M * b0c1 so I can't scale the guest processor usage 1173552804 M * daniel_hozac nope. 1173552816 M * daniel_hozac best you can do is limit it to certain processors. 1173553153 M * phedny hmm, my multicast problem doesn't seem to be vserver-related 1173553173 M * phedny eg. when running this particular app in the root context, it doens't work either 1173553197 M * phedny but doing a "ifconfig eth0 allmulti", it suddenly works, also within a vserver guest :) 1173553213 M * phedny regarding linux-vserver, it seems multicast is working :) 1173553224 M * daniel_hozac yay! ;) 1173553225 M * daniel_hozac thanks. 1173553232 M * daniel_hozac did you have to assign the addresses to the context? 1173553297 M * phedny yes, I have 239.255.255.250 in 1/ip en I have a 1/nodev 1173553323 M * phedny besides this entry in the routing table, which is always needed for multicast 1173553324 M * phedny 224.0.0.0 0.0.0.0 240.0.0.0 U 0 0 0 eth0 1173553345 M * daniel_hozac ok, cool. 1173553352 M * daniel_hozac might want to write something on the wiki. 1173553386 M * phedny indeed 1173553405 M * phedny for now I think I'll just add it to the ProblematicPrograms page 1173553446 M * daniel_hozac hmm, what's the problematic program? 1173553462 M * phedny TwonkyVision UPnP AV Media Server 1173553470 M * phedny but it's not ereally the program 1173553484 M * daniel_hozac well, a /Multicast page would be better then, IMHO. 1173553487 M * phedny more the multicast might would be the catch 1173553513 M * phedny well, it seems to be very easy, so a small entry in the FAQ could work too? 1173553545 M * daniel_hozac well, i guess. 1173553555 M * daniel_hozac i'm not really opposed to pages with pretty much no content though. 1173553568 M * daniel_hozac see e.g. http://linux-vserver.org/IPv6 :) 1173553815 Q * sid3windr Ping timeout: 480 seconds 1173553928 Q * onion Quit: leaving 1173554330 J * sid3windr luser@bastard-operator.from-hell.be 1173554476 M * phedny well, then I've created a http://linux-vserver.org/Multicast 1173554484 M * phedny my first submission to any wiki :) 1173554499 M * Bertl congrats! 1173555952 M * sannes The memory split patch, is that part of the kernel.org kernel now, or do I still need a patch to achive a 2G/2G split ? 1173556048 M * sannes a found it, hm. .heh, wonder how I missed it .. 1173556072 M * Bertl part of the vanilla kernel now 1173556119 M * Bertl okay, uploaded patches for 2.6.19.7 and 2.6.20.2 1173556141 M * Bertl next will be an update to 2.3.x 1173556245 M * sannes Bertl: I /searched in make menuconfig and it tells me it should be in Processor type and features and then memory split as expected, but I just can't see it .. 1173556269 M * Bertl it probably depends on some other settings 1173556280 M * Bertl you are using menuconfig? 1173556328 M * sannes yes 1173556350 M * Bertl then the search should also tell you about dependancies 1173556369 M * Bertl (i.e. try again and read/check what's listed there) 1173556407 M * sannes well, it didn't or rather it said it depends on .. heh 1173556413 M * daniel_hozac EMBEDDED. 1173556418 M * sannes but, in the Kconfig I found that.. yes, embedded :) 1173556453 M * daniel_hozac (i remove that dep in my kernels) 1173556535 J * ema ~ema@lart.galliera.it 1173556544 M * sannes hope the defaults in config_embedded isn't dangerous.. 1173556580 Q * b0c1 Quit: Távozom 1173556770 Q * softi42 Ping timeout: 480 seconds 1173556895 J * softi42 ~softi@p549D674A.dip.t-dialin.net 1173557113 J * shedi ~siggi@ftth-237-144.hive.is 1173559041 J * yarihm ~yarihm@84-74-16-109.dclient.hispeed.ch 1173559917 Q * ema Quit: leaving 1173559940 Q * Hunger Ping timeout: 480 seconds 1173560617 J * Hunger Hunger.hu@Hunger.hu 1173561653 Q * shedi Quit: Leaving 1173563404 J * ibu ~ibu@dslb-084-056-083-064.pools.arcor-ip.net 1173563766 M * Bertl welcome ibu! 1173564477 J * zLinux ~zLinux@88.213.58.16 1173565275 Q * fluor- Quit: . 1173565343 N * DoberMann[PullA] DoberMann 1173566174 Q * bonbons Quit: Leaving 1173567044 Q * doener Remote host closed the connection 1173567055 J * doener_ ~doener@host.magicwars.de 1173568616 M * arachnist http://phpfi.com/214449 <| any ideas? (i do have user arachnist logged in on the host though) 1173568749 M * daniel_hozac could you do strace -fF -o userdel.trace userdel arachnist and upload that somewhere? 1173568826 M * arachnist first i'll have to install strace 1173568828 M * arachnist wait ;> 1173568854 N * DoberMann DoberMann[ZZZzzz] 1173568954 M * arachnist http://88.156.185.52/userdel.trace 1173568956 M * arachnist here 1173568971 M * daniel_hozac well, connection refused. 1173568983 M * arachnist hmm? 1173568989 M * arachnist hmm 1173569045 Q * meandtheshell Quit: Leaving. 1173569047 M * arachnist http://rafb.net/p/PMD99y18.html 1173569051 M * arachnist brb 1173569152 M * daniel_hozac what does w output in the guest? 1173569486 Q * dna Quit: Verlassend 1173570554 M * mjt sed -i '/^arachnist:/D' /etc/passwd /etc/shadow 1173570556 M * mjt ;) 1173570595 M * mjt and rm -rf ~arachnist before that