1132100254 J * statutoryRape ~BelovedxC@pcp0010149132pcs.midltn01.nj.comcast.net 1132100301 P * statutoryRape 1132100671 Q * dddd44 Ping timeout: 480 seconds 1132101880 N * Bertl_oO Bertl 1132101887 M * Bertl evening folks! 1132102131 Q * michal_ Ping timeout: 480 seconds 1132102161 Q * Marchildon Ping timeout: 480 seconds 1132102182 J * Marchildon ~nicolas@dsl.speedline207.106.electronicbox.net 1132102277 J * michal_ ~michal@mprivacy-update.de 1132103177 M * Bertl wb Marchildon, michal_! 1132104443 M * ^Cist moin bertl 1132104529 M * ^Cist the 2.0 version doesn't seem to suffer from the lockups i experienced with 2.1.0rc5.. runs stable so far 1132104864 J * ryker ~ryker@c-67-176-240-74.hsd1.in.comcast.net 1132105240 M * Bertl ^Cist: good, which versions did you compare? 1132105249 M * Bertl welcome ryker! 1132105362 M * ^Cist 2.6.14-vs2.1.0rc5 - lockup... 2.6.14.2-vs2.0.1rc2 works so far (1.5days now) 1132105392 M * Bertl okay, if you can, please check with 2.1.0-rc7 too 1132105423 M * Bertl we fixed a nasty issue in -rc6 IIRC 1132105430 M * ^Cist after you told me about changes related to locking between these series, i think it a locking problem 1132105467 M * Bertl okay, please describe your findings once again ... 1132105510 M * ^Cist you remember, smp box, dual opteron, domino within a vserver, around 120-150 users 1132105535 M * Bertl okay, lockup means? 1132105540 M * ^Cist so there are lots of domino processes ~300 1132105567 M * ^Cist with many users the domino doesn't answer the clients anymore but seems to be still running 1132105583 M * Bertl okay, 'seems' means? 1132105595 M * ^Cist which means: deals with mails, database replication to other domino cluster members and stuff 1132105608 M * Bertl but doesn't answer 1132105638 M * ^Cist but i see packets in the tcp received queues from the connected clients that just stay there / fill up the queues 1132105662 M * Bertl hmm ... 1132105674 M * ^Cist so it seems like the threads/processes responsible for the user comunication are somehow blocked 1132105707 M * Bertl do you use the Hard CPU scheduler? 1132105712 M * ^Cist no 1132105718 M * Bertl okay 1132105728 M * ^Cist nothing of the new fancy stuff 1132105840 M * ^Cist i got snapshots of the proc../limit file and a netstat -t when this happened, but i doubt this is of any help 1132105867 M * Bertl hmm, please upload it anyways ... 1132105886 M * ^Cist (first thought was that i need to tune some kernel limits like i had to do with the P4s) 1132105929 Q * anonymousc Quit: adios 1132106036 M * ryker hi guys 1132106093 M * ryker i've been doing some reading about the gentoo vserver 1132106108 M * Bertl good! 1132106116 M * ryker and I'm about to buy a new server and use vserver on it 1132106140 M * ryker i wanted to ask a couple of questions b4 i waste some money though :) 1132106168 M * ryker i'm planning on getting a new dell 2850 with dual 3.4 Xeon 1132106187 M * ryker 4 GB ram, and 3 73G ultra 320 scsi drives 1132106197 M * ryker dual gigabit copper nics 1132106215 M * Bertl sounds good 1132106217 M * ryker will the gentoo vserver kernel run good on this 1132106231 M * ryker is there anything i should avoid (hardware wise)? 1132106279 M * ^Cist if you do plenty of memory IO and IO in general, i'd avoid the intel and go for opteron 1132106294 M * ryker is it relatively painless to install on this? 1132106296 M * ryker well 1132106303 M * ryker we have to use dell 1132106310 M * ryker i don't think they offer opteron 1132106318 M * ryker i would prefer opteron also 1132106327 M * ryker but, we have a dell contract 1132106338 M * ^Cist that sucks 1132106339 M * ryker and have standardized on dell rack servers 1132106366 M * ryker we get pretty good deals though 1132106375 M * ryker this box will cost me about $5700 1132106390 M * ryker they are 15k rpm drives 1132106412 M * ^Cist i wouldn't consider this such a good deal 1132106456 M * ryker where do you buy from? 1132106489 M * ryker it will also have redundant 700w power supplys 1132106494 M * ryker and a dell RAC card 1132106497 M * ^Cist well... regarding this kind of servers i'm a dealer myself :) 1132106527 M * ryker you have a website? 1132106557 M * Bertl the problem is often the liability and replacement policy 1132106588 M * ryker in our experience, dell is great with warranty work 1132106605 M * ryker they quite often have a part to us within 4 hours 1132106624 M * ^Cist dual opteron (2GHz), 4GB RAM, 240GB SATA RAID-10 (4*120GB), hotplug, 19", 650W redundant powersupply, hotplug, dual gigabit ethernet ~3800EUR 1132106657 M * ryker 19" what? 1132106672 M * ^Cist 19" 3U case 1132106679 M * ryker ahhh 1132106692 M * ryker haven't tried sata in a server yet 1132106714 M * ryker i don't think the throughput can even compare to ultra 320 scsi with 15k rpm drives though 1132106715 M * ^Cist 3ware raid controller 1132106752 M * Bertl ryker: no, they are much cheaper :) 1132106760 M * ryker yes they are 1132106771 M * ryker i think that's a big part of the price difference 1132106776 M * ^Cist hotplugable, and fine throughput 1132106784 M * ryker also, the dell RAC cards are really nice 1132106811 M * ryker you can watch the box boot into the bios screen remotely 1132106835 M * ryker does your box have something similiar? 1132106859 M * Bertl don't you talk about the dell RACs to me , they are crap! 1132106871 M * ryker lol 1132106875 M * ryker really? 1132106885 M * Bertl yes, I have to use one for a customer ... 1132106886 M * ryker i only have a couple, but they work fine 1132106889 M * ^Cist so you get those ultra fast disk drives in your box but you can't get the data through the congested northbridge on the board... sounds like a cool hardwaredesign to me 1132106906 M * ryker i agree that opteron is MUCH better 1132106907 M * Bertl ryker: ever tried over a slow line? or without java? 1132106925 M * ryker Bertl, no, only on intranet 1132106928 M * Bertl (or to be precise, without the latest and greates java) 1132106950 M * ryker Bertl, i always have my java updated, so i never noticed 1132106962 M * Bertl ryker: well, personally I prefer a simple serial remote console a 100 times over this java crap 1132106990 M * Bertl ryker: *sorry* spent too much time with it :) 1132106994 M * ryker ^Cist, i wish i could use opteron, but like i said, we pretty much have to go with dell 1132106995 M * ^Cist it's not that i often need full remote console access anyway 1132107016 M * ryker ^Cist, agreed 1132107040 M * ryker so, do you think i would run into any immediate problems with the setup i mentioned? 1132107051 M * Bertl no, should be fine ... 1132107053 M * ^Cist so some external solution you just plug into the box is sufficient 1132107056 M * ryker besides that i should choose opteron :) 1132107081 M * ryker ^Cist, not sure what you mean? 1132107125 M * ryker i was thinking about going with vmware esx server 1132107135 M * ^Cist there are KVM solutions that you just plug into keyboard/mouse/monitor connectors 1132107138 M * ryker finding gentoo vserver will save lots of money 1132107161 M * Bertl yup, definitely 1132107163 M * ryker ^Cist, oh , ya 1132107167 M * ^Cist and cpu cycles if you just need linux 1132107181 M * ryker eys 1132107183 M * ryker *yes 1132107216 M * ryker i just pray that someone doesn't come to me absolutely needing a windoze server for some reason 1132107275 M * ryker i was hoping that if i really absolutely had to have 1 windows server running on it, i might be able to run vmware in 1 vserver 1132107288 M * ryker have either of you tried that? 1132107436 M * ^Cist not me, it's been some time since i used vmware 1132107468 M * ^Cist do they have kernel modules to accelerate things or is it plain userland? 1132107514 M * ryker they have modules to accelerate 1132107530 M * Bertl I always used QEMU for system emulations 1132107531 M * ryker runs pretty good on my box at home 1132107540 M * ^Cist that could be problematic 1132107547 M * ryker never tried QEMU 1132107552 M * ryker it's pretty slow isn't it? 1132107565 M * Bertl well, should be as fast as vmware 1132107579 M * Bertl (so yes, pretty slow :) 1132107591 M * ryker vmware properly configured feels just as fast as a regular install 1132107597 M * ^Cist vmware wasn't that slow last time i used it 1132107605 M * ^Cist 2-3 years ago 1132107605 M * ryker vmware is very fast 1132107611 M * ryker they really have a great product 1132107623 M * ^Cist it's just way too expensive 1132107627 M * ryker it has become a lot better over the past couple of years 1132107639 M * ryker they just release vmware player 1132107641 M * ryker have you tried that 1132107644 M * ryker it's free 1132107653 M * Bertl what should I play with that? 1132107663 M * ^Cist play windows! 1132107664 M * ryker you can't create a new vm, but you can run one 1132107679 M * ryker or linux 1132107689 M * Bertl what would be the gain for me? 1132107694 M * ryker you can use qemu tools to create a new vm 1132107715 M * ryker Bertl, depends on what you use it for 1132107725 M * ryker i just use vmware at home in linux 1132107728 M * ryker to run windows 1132107729 M * Bertl so, what would I use it for? 1132107748 M * ryker for the apps that wine doesn't work properly for 1132107750 M * ^Cist does it have this nice commit/rollback (don't know how it's called) feature for disk writes? 1132107766 M * ryker ^Cist, no it doesn't unfortunately 1132107779 M * ^Cist this was a really usefull feature 1132107795 M * ryker well, vmware player is still nice for free though 1132107809 M * Bertl ryker: well, I have not much use for wine either ... 1132107812 M * ryker i work for a university 1132107823 M * ryker we'll use vmware player next semester 1132107827 M * ryker in some labs 1132107848 M * ryker to run linux on all those damn windows boxes 1132107880 M * Bertl hmm, wouldn't installing linux there be a much better solution? 1132107898 M * ryker i work at a backwards university 1132107904 M * ryker all of our machines are windows 1132107912 M * ryker i'm one of the few that are pushing to introduce linux 1132108116 J * dddd44 ~dhb55@218.111.178.108 1132108765 M * Bertl dddd44: ping! 1132110186 Q * sebi Ping timeout: 480 seconds 1132110718 J * sebi ~sebi@Fccac.f.strato-dslnet.de 1132110903 J * stefani ~stefani@c-24-19-46-211.hsd1.wa.comcast.net 1132110915 M * Bertl wb sebi! welcome stefani! 1132110925 M * stefani hola how are you ? 1132110930 J * Cru ~mindwarp@bastardrouterfromhell.e.de.wahlich.com 1132110935 M * Cru hi 1132110999 M * Bertl welcome Cru! 1132111010 M * Bertl stefani: fine thanks! and you? 1132111030 M * Cru hi Herbert 1132111049 M * stefani surviving. 1132111134 M * Bertl what's up Cru? 1132111254 M * Cru hard to say ;) 1132111296 Q * tchan Remote host closed the connection 1132111299 M * Cru I am sitting aside my first vserver 2.0 host and its driving me crazy because of 7x 10kU/min fans... 1132111324 J * dos000 ~dos000@CPE00119572fd49-CM00137186e53a.cpe.net.cable.rogers.com 1132111341 M * Bertl Cru: lol 1132111355 M * Cru heh 1132111360 M * ^Cist just unplug them.. fans are overrated anyway ;-) 1132111398 M * Cru already thought about that, but the real noise is caused by the 2 PSU fans ;) 1132111418 M * ^Cist stick a pencil into it 1132111459 M * Cru do I need to tell you what that server does if one of the fans stop rotating? ;) 1132111460 M * ^Cist it's probably new enough so there's warranty if it starts burning ;) 1132111504 M * ^Cist well.. stick the pencil into the speaker/buzzer once before you use it to stop the psu fans 1132111507 J * tchan ~tchan@c-67-174-18-204.hsd1.il.comcast.net 1132111521 M * Cru every of these fan modules has its own piezo beeper, about 100dB... ;) 1132111596 M * Cru but they do fancy blinking and flashing - if I put the pencils in in the right order, I could have some kind of color organ... ;) 1132111598 M * ^Cist if the fans weren't that noisy, they wouldn't need that loud beepers 1132111649 M * Cru well, thats true ;) 1132111798 M * Cru hmm, is there no online copy of configuration.html (from util-vserver) on linux-vserver.org? 1132111874 M * Bertl http://www.nongnu.org/util-vserver/doc/conf/configuration.html 1132111902 M * Cru eeh, nice css... 1132111906 M * Cru ;) 1132111911 M * Bertl whcih one? 1132111916 M * dos000 why would i get iptables -nL working inside a guest ??? this has me puzled 1132111938 M * Cru the default css on that uri ;) 1132111984 M * Bertl dos000: # chcontext --xid 100 --secure -- iptables -nL 1132111984 M * Bertl New security context is 100 1132111984 M * Bertl iptables v1.2.9: can't initialize iptables table `filter': Permission denied (you must be root) 1132111987 M * Bertl Perhaps iptables or your kernel needs to be upgraded. 1132112021 M * dos000 i am using 2.6.14-2.0.1rc1 .. iam getting "Can't open dependencies file" in one guest on a host while the other is working. i wondering where is the diffrence because i did nothing special 1132112045 M * dos000 i have two separate hosts btw 1132112097 M * Cru oh, multiple css for choice ;) 1132112150 M * Cru at least one without any hemp leafs 1132112206 M * ^Cist i smoked them away to get a better view %-) 1132112235 M * Cru heh 1132112430 M * dos000 #$!#$ i am going to reinstall now ! 1132112445 M * Cru have fun 1132112525 M * Cru just for it might prevent lots of in-depth testing, do I have to expect any problems when moving my vs 1.2x vservers to the vs 2.0 machine and running them in legacy mode? 1132112573 M * Bertl not really, you might observe one or two minor changes, but nothing problematic 1132112594 M * Cru okay, good to know 1132112726 M * dos000 worst is reboot in guest reboots the host ! 1132112746 M * Cru dos000: eh, do you still have /etc/initctl? 1132112753 M * Cru /dev/initctl 1132112755 M * dos000 in guest ? 1132112759 M * Cru in guest 1132112826 M * dos000 tow ! .. in fact the machine just did a halt after i did the reboot in guest .. restarting now 1132112859 M * dos000 Cru, problem is i did the exact documented steps i did in both machines 1132112872 M * dos000 one is behaving weird now 1132112876 M * Bertl interesting documentation :) 1132112952 M * Cru well, when it is true what I believe to be your problem, the guest is definitely not done as documentation advises you ;) 1132113091 M * Cru btw, I did not announce it on the ML again, but I reworked my RHEL4/CentOS4/FC4 packages so they can be installed thru apt-rpm 1132113109 M * dos000 Cru, /dev/initctl is there 1132113117 M * Cru dos000: remove it and try again ;) 1132113149 M * Cru dos000: please do a ls -1 /dev|wc -l for us :) 1132113167 M * dos000 dump here or pastebin? 1132113176 M * Cru ? 1132113191 M * Cru should just print one line 1132113246 M * dos000 1587 1132113280 M * Cru okay, that confirms what I worried about ;) 1132113301 M * Cru you did this in the guest, right? 1132113313 M * dos000 yes 1132113318 M * Cru your /dev is not as it should be for a vserver 1132113338 M * Bertl roughly 1581 devices too many 1132113352 M * dos000 tow! 1132113355 M * Cru your guest will have direct hardware access atm 1132113386 M * Cru i.e. if you harddist is /dev/sda and the guest has a /dev/sda device node, it will be able to format your harddist 1132113387 M * Cru k 1132113392 M * dos000 i remember i went over the step by step for this 1132113435 M * Cru what distro is that vserver? 1132113446 M * dos000 sarge 1132113450 M * Bertl well, don't know which one, but guests created by util-vserver will not contain those devices 1132113477 M * dos000 ia m using vserver vhost build -m debootstrap --hostname vhost --interface \ 1132113477 M * dos000 eth0:192.168.33.3/24 -- -d sarge -- --exclude=$REMOVE_PACKAGES 1132113500 M * Bertl and this installs all those devices? 1132113501 M * Cru dos000: okay, then my rpms wont help - only if unpacked using alien.. 1132113580 M * dos000 as i said both hosts are sarge and i used the exact kernel/vserver/utils for them 1132113616 M * dos000 in fact the kernel was built on a third machine which is behaving good. 1132113616 M * Bertl well, debian is broken in many ways ... 1132113649 M * ^Cist make sure you have capabilities in your kernel.. don't know if the tools check this nowadays 1132113688 M * Bertl recent kernels auto select the capability system 1132113689 M * ^Cist without i had guests reboot my host too 1132113699 M * ^Cist oh, ok :) 1132113733 M * dos000 testme.sh just passed 1132113733 M * ^Cist reported this quite some time ago on the list 1132113772 M * dos000 when you remove a vhost do you just rm -rf it ? 1132113780 M * dos000 after stopping of course 1132113800 M * dos000 i am going to make a separate guest 1132113871 M * Cru dos000: 1. stop your vserver guest, 2. go to your vserver guest's / and rename dev, 3. create a new dev, 4. this script _inside_ your vserver's new /dev : http://ircnet.de/paste/21 1132113898 M * Cru s/4\. this/4\. run this/ 1132114001 M * dos000 ok 1132114074 J * Aiken_ ~james@tooax8-052.dialup.optusnet.com.au 1132114316 M * dos000 Cru, i just did a new vhost and there not seem to be that many dev there 1132114338 M * dos000 only 9 of them 1132114355 M * Cru then copy them over to your other guest 1132114416 M * dos000 ok. 1132114425 Q * Aiken Ping timeout: 480 seconds 1132114450 M * dos000 now do you have any idea how did this came about so i can avoid .. maybe i did something stupid ! 1132114494 M * Cru did you do an apt-get dist-upgrade from woody to sarge? 1132114510 M * dos000 no .. i always install sarge and stick to it 1132114536 M * Bertl nothing _inside_ the guest will be able to create those devices 1132114549 M * Bertl so they have been created on the host ... 1132114568 M * Bertl dos000: what kernel/tools do you use? 1132114577 M * Cru well, I could only explain this by either a b0rken util-vserver package or the host ran some upgrade package or script that re-installed those devnodes 1132114613 M * dos000 i am using 2.6.14-2.0.1rc1 .. 1132114621 M * dos000 2.09 utils 1132114649 M * Bertl okay, you should upgrade to rc2 (fixes a nasty bug) but kernel and tools are fine (regarding your issue) 1132114663 M * dos000 if the devs can be created from within the guest then .. its bad 1132114665 M * Bertl dos000: could it be that the guest was created by some other script? 1132114686 M * dos000 no . i am sure of that 1132114710 M * Bertl device creation inside a guest will fail, one of the testme.sh checks verifies that 1132115095 M * Cru is it possible to mount --bind external directories into the vserver using the /etc/vservers/*/fstab file? 1132115132 M * Bertl yes 1132115202 M * Cru I never did sth like this using fstab - how does this need to look like? /dir/on/host /dir/in/guest bind defaults 0 0 ? 1132115325 M * Bertl /var/tmp /tmp none bind 0 0 1132115350 M * Bertl basically 'bind' is an option, the fs is 'none' 1132115357 M * Cru thanks :) 1132115386 M * dos000 darn .. all the dev are back ! 1132115390 M * Cru works great 1132115403 M * Bertl dos000: so when did they come back? 1132115436 M * Bertl (i.e. what did you do since you created the new /dev?) 1132115449 M * dos000 i did an base-config inside the guest. to set timezones and what not. 1132115456 M * Bertl how? 1132115526 M * Cru dos000: did you not configure the vserver's fstab to use devfs? 1132115543 M * Cru s/not// 1132115585 M * dos000 Cru, i just followed what was on the step by step guide. and what you mention was not part of it 1132115625 M * Cru well, I am running a couple of sarge guests, too, but I never discovered problems like this... 1132115645 M * dos000 this is @$#!$#@ confusing 1132115647 M * Cru well, they are on vs 1.2 host, but I cant imagine a situation where this could happen... 1132115658 M * Cru how did you enter the vserver? 1132115673 M * Cru hopefully not chroot ;) 1132115695 M * Bertl that's why I simply asked 'how' :) 1132115709 M * dos000 noo .. i just ssh into it after i had the root passwd setup via base-config 1132115730 M * Bertl dos000: again, 'how' did you do the base-config? 1132115733 M * Cru but you had to start base-config 1132115767 M * dos000 can you wait a bit .. i will redo what i did ? 1132115780 M * Bertl okay, take your time :) 1132115784 M * Cru heh 1132115815 M * dos000 i will again do a fresh deboostrap install 1132115827 A * Cru strews a circle of Chipsfrisch Pepperoni around dos000 1132115842 M * Bertl good, just copy the lines you enter here ... 1132115923 M * dos000 REMOVE_PACKAGES="sparc-utils,dhcp-client,lilo,makedev,pcmcia-cs,ppp,\ 1132115923 M * dos000 pppconfig,pppoe,pppoeconf,setserial,syslinux,fdutils,libpcap0,\ 1132115923 M * dos000 iptables,pciutils" 1132115945 M * dos000 vserver test000 build -m debootstrap --hostname test000 --interface \ 1132115945 M * dos000 eth0:192.168.33.5/24 -- -d sarge -- --exclude=$REMOVE_PACKAGES 1132116057 M * dos000 I: Base system installed successfully. 1132116118 M * dos000 ls -l /home/vservers/test000/dev/ | wc -l 1132116120 M * dos000 9 1132116147 M * dos000 vserver test000 start 1132116163 M * dos000 Starting system log daemon: syslogd. 1132116163 M * dos000 Starting kernel log daemon: klogd. 1132116163 M * dos000 Starting MTA: exim4. 1132116163 M * dos000 Starting internet superserver: inetd. 1132116163 M * dos000 Starting deferred execution scheduler: atd. 1132116165 M * dos000 Starting periodic command scheduler: cron. 1132116179 M * dos000 vserver test000 enter 1132116197 M * dos000 base-config 1132116223 M * Bertl okay, now, please do the following for me: 1132116232 M * dos000 i just did ls -l /dev inside the guest and i still get 9 1132116234 M * Bertl (while still inside the guest, do) 1132116248 M * dos000 i have not done case-config yet 1132116250 M * Bertl grep Cap /proc/self/status 1132116281 M * dos000 CapInh: 0000000000000000 1132116281 M * dos000 CapPrm: 00000000344c04ff 1132116281 M * dos000 CapEff: 00000000344c04ff 1132116297 M * Bertl okay, looks good, now let's try: 1132116329 M * Bertl mknod /tmp/test c 0 0 1132116384 M * dos000 mknod: `/tmp/test': Operation not permitted 1132116393 M * Bertl okay, good .. please continue 1132116558 M * Cru intangible tension ;) 1132116569 M * dos000 configure-timezone 1132116577 M * dos000 no gmt 1132116599 M * dos000 select canada/east 1132116605 M * dos000 setup passw 1132116624 M * dos000 no normal account 1132116665 M * dos000 finish ! 1132116670 M * dos000 still 9 devices ! 1132116674 M * Bertl okay, good :) 1132116690 M * dos000 login ssh now 1132116711 M * Bertl once you're in, please stop 1132116728 M * Bertl and verify that you did not 'accidentially' logon to your host :) 1132116742 M * Cru heh 1132116755 M * Bertl (e.g. by trying to list the vserver's dir) 1132116773 M * dos000 ok ... looks like ssh is loging into the host ! 1132116780 M * Bertl aha! 1132116783 M * Cru do netstat -lntp on the host 1132116790 M * dos000 but the ip is the guest ! 1132116793 M * Bertl so you 'forgot' to restrict your sshd on the host 1132116793 M * Cru is there sth like 0.0.0.0:22? 1132116807 M * dos000 f* ! 1132116810 M * Bertl dos000: and that's the difference to the other machine :) 1132116837 M * Cru heh, I hope you did not delete all device nodes on your host now... ;) 1132116854 M * dos000 noo .. i apreciate you guys help ! 1132116870 M * Bertl well, probably udev there by now ... so it should be easy to recreate any missing device nodes 1132117308 M * Cru okay, morning has come, work is to be done, Cru needs a shower 1132117318 Q * lilo Remote host closed the connection 1132117318 M * Bertl go for it! 1132117324 M * Cru have a nice day, I will return later on ;) 1132117329 M * Bertl k, cya! 1132117347 Q * Cru Quit: use Unices; $live->free() || die; 1132117619 J * lilo ~lilo@lilo.usercloak.oftc.net 1132117916 M * Hollow morning 1132117923 M * Bertl morning hollow! 1132117929 M * Hollow evening bertl :P 1132118144 M * Bertl I had an idea which might come handy for your plans 1132118202 M * Bertl (regarding tools, setup and such) 1132118305 M * Hollow Bertl: yup? 1132118324 M * Bertl well, and as it is a simple change, I just implemented it ;( 1132118327 M * Bertl :) 1132118347 M * Hollow heh, what does it do? 1132118347 M * Bertl (for testing of course) 1132118363 M * Bertl defines a new flag called VXF_PERSISTANT 1132118367 M * Johnnie I have an odd question...sort of. 1132118386 M * Bertl Hollow: which basically gets a ref/claim on the context 1132118395 M * Bertl Johnnie: go ahead ... 1132118403 M * Hollow Bertl: i.e.? 1132118416 M * Johnnie How does one bind a port to an IP for a piece of software which doesn't necessarily have support for that built in? 1132118423 M * Bertl Hollow: this allows to keep the context even when the last task did exit (until you unset the flag) 1132118438 M * Hollow oh, how lovely :) 1132118457 M * Bertl Johnnie: examples` 1132118463 M * Bertl s/`/? 1132118484 M * Johnnie Promise not to laugh? 1132118494 M * Johnnie Webmin 1132118500 M * Johnnie And nevermind, I found it. LOL 1132118503 M * Bertl hmm, cool app 1132118515 M * Johnnie I thought it didn't. 1132118519 M * Johnnie Oh, that reminds me... 1132118526 M * Johnnie I found a kick-ass theme for Webmin. 1132118538 M * Bertl is it somehow liquid? 1132118547 M * Johnnie http://www.stress-free.co.nz/webmin-theme/ 1132118562 M * Johnnie It's ... Apple/Tiger-ish 1132118582 M * Bertl hah! I knew it! :) 1132118585 M * Johnnie haha 1132118591 M * Johnnie It 1132118592 M * Johnnie err 1132118596 M * Johnnie It's pretty. 1132118606 M * Hollow Bertl: how about namespaces? they need to be kept alive as well 1132118620 M * Johnnie Anyway, sorry about my question. 1132118626 M * Bertl Hollow: they will, as they are copied in the context struct 1132118631 M * Johnnie I couldn't, for the life of me, find where it binds. 1132118650 M * Hollow ah ok 1132118663 M * Bertl I'll upload a patch in a few minutes for testing 1132118670 M * Bertl (so you can test when you like) 1132118705 M * Hollow still namespaces: how do i set the namespace? with persistant flag one could do this outside the context as well, but you'd need sth like the namesapce id to set it in the context struct, no? 1132118735 M * Bertl you need a 'process' to copy the namespace from 1132118798 M * Johnnie Hmmm... 1132118830 M * Johnnie Hollow: Remember when we discussed using htop with 'vcontext --migrate --xid 1 -- htop'? 1132118839 M * Hollow ok, so i created a context wchih is persistant, cloned my current namespace.. how do i set the namespace in the context now? vc_set_namespace does not have an xid argument 1132118840 M * Johnnie I just noticed that when I run it, it acts like it has a memory leak. 1132118844 M * Hollow Johnnie: yup 1132118864 M * Johnnie It'll slowly eat up memory... haha 1132118884 M * Bertl what kind of memory does it prefer? 1132118901 M * Johnnie 2 GB of the real stuff. 1132118902 M * Johnnie LOL 1132118907 M * Bertl Hollow: http://vserver.13thfloor.at/Experimental/delta-persist-feat01.diff 1132118911 M * Johnnie I never let it eat swap. 1132118929 M * Bertl lowmem, or himem, or maybe dma mem? 1132118940 M * Johnnie Oh, I dunno, I'm an idiot, Bertl. 1132118941 M * Johnnie LOL 1132118950 M * Johnnie You're talking to a peon in the Linux world. 1132118959 M * Johnnie I just know enough to be very dangerous. 1132118960 M * Johnnie LOL 1132118967 M * Bertl hey, you found webmin, no? 1132118968 M * Hollow heh 1132118994 M * Johnnie Don't get smart. 1132118995 M * Johnnie LOL 1132119001 M * Bertl okay, I'm off for tonight ... 1132119010 M * Johnnie Rest well, Bertl. 1132119010 M * Hollow Bertl: maybe answer my namespace question? 1132119012 M * Hollow :) 1132119028 M * Bertl have a nice whatever everyone ... especially you two ... 1132119035 M * Johnnie As Bertl's free agent, that'll cost you $0.10 per word. 1132119037 M * Johnnie Just kidding. 1132119039 M * Johnnie haha 1132119043 M * Bertl Hollow: we'll talk about that (and possible extensions later) 1132119048 M * Hollow oki 1132119060 M * Bertl night then :) 1132119065 N * Bertl Bertl_zZ 1132119068 M * matti Bertl_zZ: :) 1132119073 M * matti Bertl_zZ: HDD works? 1132119082 M * Johnnie Hollow: How'd your Asterisk stuff turn out? 1132119096 M * Hollow emerge -C asterisk 1132119097 M * Hollow :P 1132119100 M * Johnnie I got side tracked last night and had some administrative duties to tend to. 1132119104 M * Johnnie LOL 1132119107 M * Johnnie I see you got far. 1132119118 M * Johnnie Do you have any SIP devices? 1132119129 M * Hollow except for a possible softphone: no 1132119133 M * Johnnie i.e.: ATAs or softphones 1132119137 M * Johnnie Okay. 1132119143 M * Johnnie That's a start. 1132119170 M * Hollow yeah, i was just too lazy yesterday, and - tbh - i don't not unmerge it, i just stopped hacking it ;) 1132119177 M * Hollow didn't 1132119179 M * Johnnie haha 1132119183 M * Johnnie That reminds me... 1132119194 M * Johnnie I can't get it to run within a vserver. 1132119200 M * Johnnie It wants the standard baselayout. 1132119212 M * Johnnie Is there any way to fool the package into using it? 1132119216 M * Johnnie i.e. stubbing ? 1132119216 M * Hollow too bad, another package not knowing of virtual/baselayout 1132119227 M * Hollow please file a bug 1132119241 M * Hollow they seriously should depend on virtual/baselayout 1132119244 M * Johnnie With the Asterisk guys or on Gentoo or what? :) 1132119246 M * Hollow instead of sys-apps/baselayout 1132119274 J * lilo_ ~lilo@lilo.usercloak.oftc.net 1132119284 M * Hollow the gentoo asterisk guys.. but well.. i just looked.. it does not depend on baselayout 1132119303 Q * lilo Read error: Connection reset by peer 1132119359 M * Johnnie Eh? 1132119395 M * Johnnie Maybe someone sent a bug report already. 1132119400 M * Hollow maybe some other package does 1132119430 M * Hollow use -t with emerge and it'll show you the dependency tree 1132119431 P * stefani parting (is such sweet sorrow) 1132119570 M * dos000 i happen to be using asterisk with vservers :-) 1132121202 Q * virtuoso Ping timeout: 480 seconds 1132122308 Q * ryker Quit: Leaving 1132122329 J * _mcp ~hightower@wolk-project.de 1132122357 Q * mcp Read error: Connection reset by peer 1132122367 N * _mcp mcp 1132123355 M * dlippolt any smart multicast guys on? 1132123599 J * virtuoso ~s0t0na@shisha.spb.ru 1132123732 M * dos000 i did multicast stuff in a previous life and i am definitely not smart :-) 1132124136 J * domtec sp@83.215.237.10 1132124144 J * _mcp ~hightower@wolk-project.de 1132124150 Q * mcp Read error: Connection reset by peer 1132124159 N * _mcp mcp 1132124489 Q * domtec Quit: get satisfied! • :: ««« (Gamers.IRC) »»» www.gamersirc.net :: 1132125199 M * dlippolt dos000: are you following the vserver list? 1132125243 M * dos000 not really 1132125271 M * dos000 i am new on this 1132125577 N * lilo_ lilo 1132125789 Q * alexx Quit: Bye 1132127698 Q * Johnnie Quit: G'bye! 1132127758 M * dlippolt cool, no worries. i probably just need to talk to Herbert 1132128016 J * alpha ~Miranda@83.215.237.5 1132128071 N * alpha Guest209 1132128147 J * Johnnie ~john@acs-24-154-53-217.zoominternet.net 1132128992 J * alexx ~alexx@proxy.ikse.net 1132128999 Q * ^Cist Ping timeout: 480 seconds 1132130217 Q * shedi Quit: Leaving 1132130988 J * erwan_taf ~erwan@81.80.43.67 1132131168 Q * erwan_taf Remote host closed the connection 1132131435 J * erwan_taf ~erwan@81.80.43.67 1132132376 J * arok ~root@p5495A1D8.dip0.t-ipconnect.de 1132132487 M * arok help!^^ 1132132499 M * arok bertl where are you ^^ 1132132505 Q * Aiken_ Ping timeout: 480 seconds 1132132657 M * arok maybe someone different can help me too. i have ~hide_netif in my cflag-file. now the routing table is on every vserver the same but i need different default gateways... is that possible? 1132132962 J * shedi ~siggi@tolvudeild-201.lhi.is 1132133573 Q * alexx Quit: Parti 1132133601 Q * Hunger Ping timeout: 480 seconds 1132133745 J * alexx ~alexx@proxy.ikse.net 1132134062 J * lilo_ ~lilo@lilo.usercloak.oftc.net 1132134156 Q * lilo Ping timeout: 480 seconds 1132134207 Q * arok Quit: leaving 1132134345 J * arok ~thorsten@p5495A1D8.dip0.t-ipconnect.de 1132135431 P * erwan_taf Leaving 1132135784 M * sizo moin 1132136219 J * click_ click@ti511110a080-4155.bb.online.no 1132136391 Q * click Ping timeout: 480 seconds 1132137135 M * BWare arok: Kernel routes are always the same 1132137175 M * BWare There has been some discussion on the mailing list to use different routing tables per vserver using iproute2 1132137290 M * arok mh oO 1132137300 M * arok but without ~hide_netif i had different routing oO 1132137307 J * Hunger Hunger.hu@Hunger.hu 1132138129 J * shed| ~siggi@tolvudeild-202.lhi.is 1132138306 Q * shedi Ping timeout: 480 seconds 1132138550 Q * dddd44 Read error: Connection reset by peer 1132138943 J * dddd44 dhb55@218.111.178.108 1132139216 J * erwan_taf ~erwan@81.80.43.67 1132140013 J * DavidS ~david@chello062178045213.16.11.tuwien.teleweb.at 1132140228 P * erwan_taf Leaving 1132140325 M * DavidS hi, i've read vserver-build(8) but it seems that this can only build a fresh vserver or copy from the running system... but I have the system i want to vserverize currently in /root/www/ (rsynced from remote host) ... any hints? 1132140345 M * DavidS or is it easier to vserver build debootstrap a new system and then replace that? 1132140413 A * DavidS is blind 1132140421 M * DavidS I just found the skeleton method ... 1132140482 M * FaUl DavidS: should be enough to copy the /etc/vserver/ part from another vserver and adapt it 1132140511 M * DavidS FaUl: it's my first :) but "skeleton" just does the right thing ... 1132140583 M * FaUl DavidS: excellent ;-) 1132141395 N * click_ click 1132141678 P * click [IRSSI] 1132142114 J * click click@ti511110a080-4155.bb.online.no 1132142946 M * daniel_hozac DavidS: just remember to use the skeleton's /dev ;) 1132143077 M * DavidS daniel_hozac: thanks for the hint, but i was just copying the data (+binaries) anyways ... 1132143462 M * arok 11:31 < BWare> There has been some discussion on the mailing list to use different routing tables per vserver using iproute2 <-- do you mean that: http://archives.linux-vserver.org/200311/0470.html ? 1132144257 M * DavidS woot. it runs :) 1132144312 M * DavidS there are still some minor things, ike mysql wanting some localhost thingies .. but that should be no problem 1132144391 Q * dddd44 Read error: Connection reset by peer 1132144421 M * DavidS scary though, seeing a server boot in the bash :) 1132145360 M * DavidS ah: mysql din't start, because i screwed the permissions on /var/run/mysqld while rsyncing ... 1132145387 M * arok grrr. i try uml now... 1132145455 A * DavidS just barely missed a cardiac arrest: "Rebooting..." 1132145468 M * DavidS (in the vserver of course ;) 1132145475 M * daniel_hozac you should disable that script. 1132145485 M * daniel_hozac it does nothing valuable in a vserver. 1132145513 J * dddd44 dhb55@218.111.178.108 1132145585 M * DavidS daniel_hozac: yeah deleted it (and friends) already, but re-rsyncing from the (yet) live system restored them too 1132145691 M * DavidS now i just have to take down the services on the running server, rsync one last time and start the vserver ... *sigh* 1132146358 Q * virtuoso Read error: Connection reset by peer 1132146378 J * virtuoso ~s0t0na@80.253.205.251 1132148576 M * DavidS yay :) thumbs up: everything is running again on the new hardware, within VServer. Thanks for everything! 1132148887 N * Bertl_zZ Bertl 1132148892 M * Bertl morning folks! 1132149065 M * Hollow morning Bertl! 1132149102 M * Bertl arok: UML with or without Linux-VServer? 1132149120 M * Hollow Bertl: i found out what is left for the persistant thing... i need to pass the xid to vc_set_namespace, then it should work 1132149150 M * Hollow or probably i'd need it even without persitatn flag ;) 1132149152 M * Hollow who knows 1132149235 M * Bertl sec 1132149300 M * Bertl yep, that should be trivial, we just make a 'new' version to pass the context id (the id field is unused anyway) 1132149309 M * Hollow great 1132149316 M * DavidS uhm .. what is the difference between ulimits and rlimits? 1132149345 M * Bertl DavidS: simple, you know what ulimits are? 1132149361 M * Bertl DavidS: welcome here btw :) 1132149393 M * DavidS Bertl: yeah, i can set them (e.g. in the shell) and it keeps me from doing stupid things like :(){:|:} ... 1132149403 M * BWare arok: that is one of them... there should be a more recent discussion on the matter 1132149408 M * DavidS (fork bombs that is) 1132149413 M * Bertl DavidS: okay, good, now imagine similar limits for a whole context 1132149426 M * Bertl (i.e. not for a process/user) 1132149453 M * DavidS Bertl: and then i have rlimits? so i don't need the nproc flag when i "echo 200 > rlimits/nproc" ? 1132149461 M * Bertl http://linux-vserver.org/Resource+Limits 1132149466 M * DavidS Bertl: I rescued a webserver from a dying disk into a vserver today :) 1132149471 M * Bertl DavidS: no, nproc is obsolete since more than a year 1132149479 M * DavidS Bertl: d'oh :) 1132149526 M * DavidS http://linux-vserver.org/linux-vserver_administrators_guide still advertises what i now come to recognise as the "legacy" configuration method ... 1132149541 M * Hollow apropos limits... the data struct for disk limits contains uint32_t members, whereas CDLIM_* defines are unsigned long long.. 1132149577 M * BWare arok: This is the message I refered to: Subject: Re: [Vserver] [SOLVED]need a point in the right direction in certain kernel networking settings 1132149577 M * BWare Date: Tue, 27 Sep 2005 09:31:30 +0200 1132149672 M * Bertl DavidS: please, if you get around, correct and/or mark the page as 'legacy' 1132149698 M * DavidS Bertl: can do 1132149752 M * Bertl Hollow: does that cause gcc 4.x to complain? 1132149764 M * Hollow yeah, though it's a warning only 1132149786 M * Hollow ../../src/vdlimit.c:105: warning: large integer implicitly truncated to unsigned type 1132149799 M * Bertl hmm ... 1132149871 M * Bertl maybe we should define separate ones for userspace 1132149884 M * Bertl kernel does explicitly cast them as required 1132149911 M * Bertl if (dli->dl_space_total == (uint64_t)CDLIM_INFINITY) 1132149911 M * Bertl *space_total = (uint32_t)CDLIM_INFINITY; 1132149913 M * Hollow hm, ok, so should i simply cast it manually? 1132149927 M * Bertl no, let's clean that up properly 1132149936 M * Hollow ok 1132149938 M * Bertl we have two sets for the limits 1132149958 M * Bertl so it is better to keep that somewhat in sync 1132150148 M * Bertl so, we'll probably make those defines DLIM_* ones 1132150172 M * Bertl (internal) 1132150230 M * Hollow and CDLIM for userspace? 1132150249 M * Bertl yep, with an explicit cast, or maybe just UL 1132150255 M * Hollow ok 1132150255 M * arok 14:50 < Bertl> arok: UML with or without Linux-VServer? <-- without. what sense makes vserver with uml? oO 1132150287 M * Bertl arok: well, you can have different kernel versions running 1132150310 M * Bertl similar setup as with xen + linux-vserver (not untypical) 1132150336 M * arok i don't need that ^^ i just need different routings 1132150349 M * Bertl different routing? 1132150363 M * arok each vserver needs a different default gw 1132150371 M * Bertl well, that is simple 1132150388 M * arok maybe you remember me. i asked for a solution with building tun's 1132150395 M * arok is it? ^^ 1132150411 M * Bertl yes, different routing is quite simple, as long as you config it on the host 1132150445 M * Bertl http://archives.linux-vserver.org/200311/0470.html (from the Documentation section on linux-vserver.org) 1132150485 M * arok would be better if it worked on the vservers too... but not essential 1132150517 M * arok ah. i already tried that 1132150519 M * Bertl will need some time, until ngnet is available, then it will work inside the guests too (for some overhead) 1132150566 M * arok i didn't got this solution working oo 1132150572 M * arok okay, i give a second try... 1132150580 M * Hollow this hole limit stuff confuses me.. VLIM, RLIM, CRLIM, DLIM, CDLIM.. 1132150602 M * Hollow please enlighten me ;) 1132150636 M * Bertl well, its something which results from the fact that hardware archs are quite different 1132150672 M * Bertl some have 32bit, others 64bit, some have rlimits with 2^32-1 others with 2^31-1 (max) 1132150691 M * Bertl (the 64bit ones with 2^64-1 or 2^63-1 :) 1132150732 M * Bertl so we encounter many many different 'values' there, and we have to 'define' and 'compare' them properly 1132150764 M * Bertl this is supposed to work transparently for userspace, which should only see one set of those defines 1132150779 M * Hollow ok 1132150791 M * Bertl (unless you want to do it like glibc and have a different implementation for each arch :) 1132150797 M * Hollow :P 1132150852 M * Bertl so ... the CRLIM_* is the userspace variant 1132150856 J * Cru ~mindwarp@bastardrouterfromhell.e.de.wahlich.com 1132150862 M * Hollow C for combined? 1132150865 M * Bertl (maybe we should explicitly cast that to uint64_t 1132150876 M * Cru hi 1132150882 M * Bertl welcome Cru! 1132150895 M * Bertl Hollow: IIRC, the C was for Context 1132150902 M * Hollow ok ;) 1132150914 M * Hollow so, userspace has CRLIM and CDLIM, right? 1132150922 M * Bertl yes, that's the idea ... 1132150942 M * arok hm oO Bertl... help me please ^^ i'm not so familiar with the ip-command 1132150946 M * arok tracer26:/tmp# ip route add default dev tun0 table 100 1132150946 M * arok RTNETLINK answers: File exists 1132150956 M * Hollow and diff between VLIM and RLIM is just the vanilla ones, and the vserver specific ones, right? 1132150964 M * DavidS arok: ip route ls ? 1132150967 M * Bertl arok: ip route ls table 100 1132151049 M * arok ah. "ip route ls table 100" shows an entry "ip route ls" doesn't show... 1132151073 M * arok ah 1132151086 M * arok ^^ 1132151103 M * Bertl for each guest, make a separate table 1132151121 Q * shed| Ping timeout: 480 seconds 1132151131 M * Bertl redirect the routing decisions to the proper tables (usually based on IPs) 1132151138 J * ^Cist ~x@p54A55254.dip.t-dialin.net 1132151147 M * Bertl welcome ^Cist! 1132151157 M * ^Cist moin Bertl :) 1132151278 M * arok can i use "ip rule add iif tun0 table 100" instead of "ip rule add from ..." ? 1132151294 M * Bertl Hollow: I guess the uint32_t should work for most cases (allows up to 4TB per guest) 1132151326 M * Bertl arok: man ip or info ip? 1132151368 M * Hollow Bertl: ok, i just implement what is needed ;) 1132151400 M * Bertl Hollow: so I will move the CDLIM defines into the _cmd ... 1132151411 M * arok i know "iif" from the man-page ^^ i'm just not sure if it will work... well... i try ^^ 1132151413 M * DavidS Bertl: iproute2 documentation was awful the last time i checked .. (which I have to admit was not recently) 1132151414 M * Bertl arok: will work as long as the traffix uses that interface 1132151417 M * Hollow ok, thx 1132151437 M * Bertl DavidS: right, let me point to the advanced routing howto then :) 1132151496 M * Bertl e.g. http://www.ibiblio.org/pub/Linux/docs/HOWTO/other-formats/html_single/Adv-Routing-HOWTO.html 1132151546 M * DavidS Bertl: well, it has at least some text between the syntax ;) 1132151608 M * arok doesn't work. -_- 1132151703 M * arok hm... 1132151717 A * DavidS notices, that "cp -av /var/lib /var/lib/vservers/mail/" is no good idea ... 1132151734 M * ^Cist wow, there are even tc manpages these days. with text! 1132151766 J * shedi ~siggi@tolvudeild-201.lhi.is 1132151880 J * sepp ~|Rogue@port-212-202-71-63.dynamic.qsc.de 1132151885 M * sepp hi 1132151930 M * sepp is it normal that a user in a vserver can sniff traffic from the host system ? 1132151938 M * Bertl welcome sepp! 1132151972 M * Bertl sepp: no, unless you have a bad config which gives CAP_NET_RAW or similar (debian tended to create such broken configs) 1132151983 M * sepp ok ich check that 1132152037 M * sepp your are right 1132152080 Q * Hollow Remote host closed the connection 1132152084 M * sepp Bertl: could you tell me how i can limit cpu usage to 50% for one vserver 1132152092 M * Bertl Hollow: could you check if UL without cast works on 'some' archs (at least i386 and x86_64 gcc)? 1132152109 M * Bertl sepp: which kernel and tool version do you use? 1132152171 M * arok oh man so ein rotz, ich checks net. 1132152179 M * arok sorry ^^ 1132152184 M * Bertl lol 1132152215 M * Bertl *sorry* too :) 1132152230 M * Bertl arok: I'm kind of leaving right now ... but I'll be back later (or at least tomorrow) 1132152251 M * Bertl arok: but I'm pretty sure folks around here will be able to help you too 1132152271 M * arok i don't think so... i know how it should work... 1132152288 M * arok and i have a very special situation here with my tunnels etc ;) 1132152294 M * ^Cist what are you trying to do arok? 1132152311 M * Bertl okay, off for now ... back later ... have fun! 1132152317 N * Bertl Bertl_oO 1132152320 M * ^Cist cu Bertl 1132152321 M * arok Bertl: cya Bertl 1132152323 M * sepp Bertl: 2.4.31 with 0.30 utils 1132152336 M * sepp ... cu bertl 1132152341 M * arok ^Cist: atm different defualt gw's for each vserver 1132152342 M * Bertl_oO sepp: you'll need a 2.6/vs2.0 kernel for that 1132152357 A * Bertl_oO *gone* 1132152371 M * sepp ok i wanted to upgrade anyway 1132152814 J * Hollow ~hollow@home.xnull.de 1132153085 A * DavidS is getting the hang of the vserver stuff :) 1132153222 J * yungyuc ~yungyuc@220-135-53-220.HINET-IP.hinet.net 1132154179 Q * yungyuc Quit: leaving 1132154748 M * DavidS I recon the next thing I'll have to do is to create another vserver so the name server and the mail server don't run in the same context? 1132157277 Q * shedi Quit: Leaving 1132157329 Q * Guest209 Read error: Connection reset by peer 1132158194 Q * DavidS Quit: Download Gaim: http://gaim.sourceforge.net/ 1132158849 J * Hollow_mobile ~hollow@p5497BB49.dip0.t-ipconnect.de 1132158878 Q * arok Quit: blubbi 1132158878 J * Vinz ~rmoquin@61-59-182-155.adsl.static.seed.net.tw 1132158926 M * Vinz Hi guys! short question: what is the recommended way to nfs mount directories in a vserver? it doesn't have to be user mounted from the vserver itself ... i can do it from the root host 1132159029 M * Cru I did not do this with vserver 2.0 til now, but it should be as easy as vnamespace -e mount :/ /vservers// 1132159043 M * Cru -t nfs 1132159058 M * Hollow_mobile Vinz: /etc/vserver/name/fstab.remote 1132159077 M * Vinz oh that's even better 1132159101 M * Hollow_mobile it's executed in the network context of the vserver 1132159175 M * Cru but that also means that the vserver is given direct access to the NFS server 1132159187 M * Vinz is that bad? 1132159206 M * Cru well, depends on why you are using vservers ;) 1132159207 M * Hollow_mobile it shouldn't be different from a normal mount 1132159264 M * Cru if you use vservers for security enhancement, you would firewall access from the vservers IP context to other services than those whitelisted 1132159295 M * Cru but if mounting using the vserver's IP context, NFS needs to be "open" for the vserver 1132159306 M * Vinz i don't mind that 1132159327 M * Hollow_mobile well, you can mount it via normal fstab too, so it's executed purely in the host env 1132159337 M * Vinz as long as it doesn't allow someone to get out of the chroot context 1132160262 J * Doener doener@i5387D7D2.versanet.de 1132160267 M * Cru bbl 1132160283 M * Hollow_mobile hey Doener! 1132160295 M * Cru lecker, Doener 1132160300 M * Hollow_mobile :P 1132160321 Q * Cru Quit: use Unices; $live->free() || die; 1132160339 M * Hollow_mobile i got the whole namespace/secure chdir thingy working now.. thanks again for your explanation.. 1132160874 J * stefani ~stefani@superquan.apl.washington.edu 1132161012 Q * Vinz Quit: 1132161361 Q * bragon Ping timeout: 480 seconds 1132161634 J * bragon ~bragon@god.geeknode.org 1132162639 J * Viper0482 ~Viper0482@p54976080.dip.t-dialin.net 1132163042 J * shedi ~siggi@inferno.lhi.is 1132164170 M * Viper0482 hi 1132164332 Q * Hollow_mobile Quit: This computer has gone to sleep 1132164864 P * stefani I'm Parting (the water) 1132164914 J * ryker ~ryker@clo-122-046.calumet.purdue.edu 1132165581 M * Doener Hollow: you're welcome! 1132168210 J * Xav montagut@scipc-xm.unilim.fr 1132169106 J * prae ~benjamin@sherpadown.net 1132169785 J * NikDaPhreak ~NikDaPhre@193.193.162.69 1132169789 M * NikDaPhreak hi all 1132171141 Q * NikDaPhreak Quit: My damn controlling terminal disappeared! 1132171756 Q * gregster Remote host closed the connection 1132172142 Q * Hollow Remote host closed the connection 1132172953 J * Hollow ~hollow@home.xnull.de 1132173830 J * Cru ~mindwarp@instructor.e.de.wahlich.com 1132173833 M * Cru re 1132174906 Q * Viper0482 Remote host closed the connection 1132175506 Q * prae Quit: Pwet 1132175676 J * Aiken ~james@tooax8-222.dialup.optusnet.com.au 1132175864 J * yarihm ~yarihm@84-73-119-83.dclient.hispeed.ch 1132176233 J * prae ~benjamin@sherpadown.net 1132176638 N * Bertl_oO Bertl 1132176641 M * Bertl evening folks! 1132176669 M * Cru re bertl 1132176712 M * dlippolt Bertl, are you Herbert Poetzl? 1132176731 M * Bertl dlippolt: let me check ... 1132176736 M * dlippolt haha 1132176742 M * Cru heh 1132176762 M * Bertl dlippolt: hmm, yes, /whois says so 1132176765 M * Bertl :) 1132176805 M * dlippolt its true i'm totally irc illiterate. only know what i learned playing warcraft :) 1132176826 M * dlippolt we had a multicast issue in the mailing list started as a thread 1132176828 M * Hollow evening Bertl! 1132176833 M * Bertl evening Hollow! 1132176839 M * dlippolt subject: support for multicast? 1132176843 M * Bertl dlippolt: yes, I remeber ... 1132176848 M * Hollow seems like ther persitent thing is not working yet 1132176850 M * dlippolt you said ping u on irc... :) 1132176859 M * dlippolt PING 1132176877 M * Bertl good, give me a few minutes to read my emails ... 1132177068 M * Doener Bertl: i can confirm the nr_running/loadavg issues mentioned on the ml, but had no time looking into it, yet 1132177304 M * Bertl yup, already observed negative nr_running in a few cases 1132177322 M * Bertl I think I can even reproduce them 1132177355 M * Bertl I suspect we decrement twice under certain conditions 1132177577 M * Bertl dlippolt: okay, so let's hear what works and what not 1132177596 M * Bertl Hollow: what issues do you see? 1132177621 M * Hollow well, the context does not stay alive, it deallocs immediatly 1132177696 M * Bertl how do you 'create' it? 1132177715 M * Bertl (the problem is, the flag is not honored on creation yet) 1132177730 M * Bertl Hollow: but that should be easy to fix 1132177732 M * Hollow ok, that seems like the problem then 1132177786 M * Bertl Hollow: say, what are your plans regarding kernel/user communication? (do we want to consider netlink or similar?) 1132177846 M * Hollow ah, the old question :) yeah, well.. the question is what do we do with it (except state changes probably).. 1132177968 M * Bertl hmm, IIRC, there was some kind of request regarding namespaces and mounts and such ... (which actually reminded me of this) 1132177982 M * Hollow request to do what? 1132178036 M * Bertl well, to list them, basically ... 1132178072 M * Bertl dlippolt: still here? 1132178089 M * Hollow ok, so we have state changes and information grabbing 1132178121 M * Bertl yes, especially information retrieval, like for graphing or vtop kind of stuff and such ... 1132178162 M * dlippolt Bertl: what works... i can broadcast multicast 1132178169 M * dlippolt Bertl: i cant receive 1132178184 M * dlippolt Bertl: interestingly, i can ping the multicast address using the ping utility and get a response 1132178194 M * dlippolt Bertl: right now i'm running 3 tomcat instances in the same vserver 1132178197 M * Hollow ok, i'm off for about an hour, could you look at the persitant thing? i'd like to test it before going to bed :9 1132178199 M * Hollow :) 1132178219 M * dlippolt Bertl: the goal, is to separate them out into separate vservers, one on each real host per multicast ip 1132178235 M * dlippolt Bertl: so i'm comfortable binding a particular multicast ip to a particular vserver 1132178290 M * Bertl Hollow: will do ... 1132178310 M * Hollow oki, cu later then 1132178356 M * Bertl dlippolt: okay, soo .. you have some tools to check multicast? 1132178378 Q * PerlOffice Read error: Operation timed out 1132178387 M * Bertl i.e. some kind of 'server' and 'client' 1132178387 M * dlippolt Bertl: also interesstingly, and somewhat off topic, when i run tomcat in exactly the same config on osx, the second tomcat can never start due to a Bind Exception port already in use... on multicast 1132178409 M * dlippolt Bertl: absolutely. filip's code from cvs.apache.org 1132178413 M * dlippolt Bertl: let me get u a link 1132178455 M * Bertl okay, now, multicasting works on special IPs around 224 or so, right? 1132178667 M * dlippolt Berl: yes, i'm using: 1132178669 M * dlippolt mcastAddr="228.0.0.4" 1132178669 M * dlippolt mcastBindAddr="192.168.1.237" 1132178669 M * dlippolt mcastPort="45564" 1132178709 M * dlippolt i'm using this to test 1132178711 M * dlippolt http://mail-archives.apache.org/mod_mbox/jakarta-tomcat-user/200404.mbox/%3C01af01c42d75$f4bff1a0$dd01dc0a@Corp.LaQuinta.com%3E 1132178722 M * dlippolt in addition to the tomcat code itself 1132178810 M * dlippolt using that tool would show output something like the following 1132178811 M * dlippolt moose:/home/sitongia/tmp# java -cp tomcat-replication.jar MCaster 1132178811 M * dlippolt 228.0.0.4 45564 Terminal1 1132178811 M * dlippolt Usage MCaster [address port message] 1132178811 M * dlippolt BEGIN TO RECEIVE 1132178811 M * dlippolt SENT:Terminal11 1132178813 M * dlippolt RECEIVED:Terminal11 FROM /128.117.224.85:45564 1132178815 M * dlippolt BEGIN TO RECEIVE 1132178817 M * dlippolt SENT:Terminal12 1132178819 M * dlippolt RECEIVED:Terminal12 FROM /128.117.224.85:45564 1132178821 M * dlippolt BEGIN TO RECEIVE 1132178825 M * Bertl okay, so, could you start the 'receiving' test tool inside your guest and 'strace -fF -o mc.strace ...' that? 1132178855 M * dlippolt when i run that test, i never get receives, even to itself 1132178861 M * Bertl btw, I'm not sure that java apps are the best test tools 1132178865 M * dlippolt here is what i get when i run it 1132178866 M * dlippolt i'm sure 1132178885 M * Bertl well, let's see the strace ... then we talk about that :) 1132178963 M * Bertl note: for now, I'm not interested in _any_ reception, just the 'binding' 1132178970 M * dlippolt when i run that tool, i get the following 1132178971 M * dlippolt tomcat0@v237:/opt$ java -cp tomcat-replication.jar MCaster 224.0.0.4 45564 Terminal1 1132178971 M * dlippolt Usage MCaster [address port message] 1132178971 M * dlippolt BEGIN TO RECEIVE 1132178971 M * dlippolt SENT:Terminal11 1132178971 M * dlippolt SENT:Terminal12 1132178973 M * dlippolt SENT:Terminal13 1132178980 M * dlippolt strace forthcoming 1132179000 M * Bertl yep, please upload the output somewhere (e.g. pastebin.com) 1132179090 M * dlippolt doing 1132179222 M * dlippolt http://www.moverotech.com/mc.strace 1132179278 M * Bertl see, that's what I meant with "I'm not sure that java is the best way to test it" ... 1132179296 J * PerlOffice ~stuart@220-253-85-202.QLD.netspace.net.au 1132179303 M * dlippolt i read strace otuput about as slow as i read assembly 1132179305 M * Bertl (for comparison, check out 'strace -fF ping -c 1 127.0.0.1 ) 1132179327 J * jkl eric@c-67-173-248-142.hsd1.co.comcast.net 1132179335 M * Bertl wb PerlOffice! 1132179338 M * Bertl welcome jkl! 1132179346 J * cede ~cemil@dslb-084-056-076-090.pools.arcor-ip.net 1132179349 M * cede hi there :) 1132179354 M * Bertl hey cede! 1132179377 M * jkl hey bertl 1132179380 M * jkl how goes it today 1132179382 M * cede i have a problem... i need some more possible connections for my httpd in my vserver - with 2.4 it was no problem to set ulimits - where can i do it with 2.6? 1132179389 M * cede hallo Bertl :) 1132179443 M * Bertl dlippolt: I do not see any multicast connect here (grep for 'socket' and 'connect') 1132179481 M * Bertl cede: 'some more possible connections'? 1132179492 M * Bertl jkl: fine, thanks! and for you? 1132179503 M * cede i need an bigger fs.file-max as example, bert :) 1132179520 M * cede i cant handle more then 512 clients with the debian package - so i want to increase to 2048 1132179533 M * Bertl hmm, sysfs/sysctl should allow that on 2.6 too, no? 1132179541 M * dlippolt Bert: interestingly, when i run strace -fF ping -c 1 228.0.0.4 while the tool is running 1132179548 M * dlippolt i get a socket not permitted error 1132179555 M * cede Bertl: is this enough? 1132179563 Q * prae Quit: Pwet 1132179563 M * dlippolt but when i just run ping -c 1 228.0.0.4 i get a good reply 1132179564 M * cede doesnt i need to control my vserver even? 1132179586 M * Bertl cede: I'd assume you have not limited your guest yet, so probably the host limits apply 1132179606 M * Bertl cede: http://linux-vserver.org/Resource+Limits 1132179618 M * cede merci :) 1132179630 M * cede hah, it work! 1132179659 M * Bertl dlippolt: okay, so my assumption (in this, your testcase) is that the tool (after doing a ton of useless java stuff) tries to create an ipv6 socket, and fails. period. 1132179703 M * Bertl (at lest that's all I see from the strace) 1132179709 M * dlippolt Bertl: interesting 1132179720 M * dlippolt Bertl: composing something for u to look at 1132179729 M * Bertl socket(PF_INET6, SOCK_STREAM, IPPROTO_IP) = -1 EAFNOSUPPORT 1132179768 J * FireEgl Atlantica@2001:5c0:84dc:: 1132179792 M * Cru Bed is waiting 1132179795 M * Cru bbl 1132179804 Q * Cru Quit: use Unices; $live->free() || die; 1132179988 M * dlippolt Bertl: the interesting thing is whenrunning the tool: 1132179988 M * dlippolt tomcat1@v237:~$ ping 228.0.0.4 1132179988 M * dlippolt PING 228.0.0.4 (228.0.0.4) 56(84) bytes of data. 1132179988 M * dlippolt 64 bytes from 192.168.1.99: icmp_seq=1 ttl=64 time=0.018 ms 1132179988 M * dlippolt 64 bytes from 192.168.1.99: icmp_seq=2 ttl=64 time=0.008 ms 1132179989 M * dlippolt tomcat@v237:~$ strace -fF ping -c 1 228.0.0.4 1132179991 M * dlippolt 1132179993 M * dlippolt fstat64(3, {st_mode=S_IFCHR|0620, st_rdev=makedev(136, 8), ...}) = 0 1132179995 M * dlippolt old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40018000 1132179997 M * dlippolt _llseek(3, 0, 0xbffff6d0, SEEK_CUR) = -1 ESPIPE (Illegal seek) 1132179999 M * dlippolt write(3, "ping: icmp open socket: Operatio"..., 48ping: icmp open socket: Operation not permitted 1132180001 M * dlippolt ) = 48 1132180003 M * dlippolt close(3) 1132180005 M * dlippolt 1132180022 M * dlippolt 192.168.1.99 is the ip of the host 1132180049 M * dlippolt strange that icmp succeeds out of strace, but fails inside of strace 1132180079 M * Bertl well, that's at least interesting, what distro do you use? 1132180087 M * dlippolt debian sarge 1132180092 M * dlippolt on top of redhat enterprise linux 1132180110 M * Bertl let's see if other distros do that too :) 1132180122 M * dlippolt base or guest 1132180127 M * Bertl do I need to configure anything for that ping to work? 1132180140 M * dlippolt no, just download the jar file and run 1132180194 M * dlippolt u have to have java running 1132180201 M * dlippolt *installed 1132180252 M * Bertl hmm, url of the jar? 1132180273 J * Funzo Funzo@CPE0013460bf939-CM001225009118.cpe.net.cable.rogers.com 1132180289 M * Bertl dlippolt: I'm currently looking at the MLA 1132180294 M * Bertl welcome Funzo! 1132180310 M * Funzo thank you 1132180334 M * dlippolt www.moverotech.com/tomcat-replication.jar 1132180341 M * Bertl tx 1132180341 M * dlippolt as dl'ed from the cvs.apache.rog 1132180371 M * Bertl and can I run that on a different machine in the local lan too? 1132180378 M * dlippolt yes, ideally 1132180389 M * dlippolt thats how you test functioning multicast on the network 1132180396 M * dlippolt since often hosts are free, but the net devices have it blocked 1132180413 M * Bertl okay, any advised java version? maybe with url too? 1132180414 M * dlippolt just use same multicast ip and port on both ends 1132180428 M * dlippolt i'm using sun jdk 1.5 repackaged into deb 1132180456 M * dlippolt let me give u url, violates sun's distribution rules, so delete after, ok :) 1132180469 M * Bertl give me a moment ... 1132180490 M * sizo n8 1132180500 M * Bertl night sizo! 1132180508 M * dlippolt www.moverotech.com/jdk-1_5_0_5-linux-i586.bin 1132180528 M * sizo night Bertl 1132180530 M * dlippolt ack, should be jdk-1_5_0_05-linux-i586.bin 1132180544 M * Bertl 1.4.1 will work too? 1132180559 M * dlippolt yes, works fine 1132180566 M * dlippolt i hve 1.4.2 installed, lemme se if its broken, too 1132180573 M * dlippolt *also have 1132180757 M * dlippolt Bertl: interestingly... same behavior for 1.4.2, but not the following 1132180765 M * dlippolt Bertl: when i run as roon in the vserver, both pings work 1132180784 M * dlippolt Bertl: when i run as tomcat0 in the vserver, strace ping doesn't work, but regular ping dos 1132180787 M * dlippolt *does 1132180829 M * Bertl hmm, okay, maybe your ping is suid root? 1132180848 M * dlippolt it is 1132180854 M * Bertl (it is not allowed to strace suid apps as user) 1132180863 M * dlippolt yup 1132180867 M * dlippolt mystery solved 1132180874 M * Bertl good :) 1132180882 M * dlippolt still remaining question as to why tool isn'tworking 1132180882 M * dlippolt :) 1132180883 M * Bertl java is building here :) 1132180899 M * Bertl no, I'm not on gentoo *G* 1132180936 M * dlippolt i was gonna say... :) 1132180946 M * dlippolt we could be here all night 1132181057 Q * bragon Quit: leaving 1132181236 M * Bertl okay, compile done 1132181498 M * Bertl hmm, and the MCaster thingy is supposed to print those increasing Terminal messages? 1132181541 M * dlippolt yes, but intermingled are supposed to be "Receives" 1132181548 M * dlippolt since its both a client and a server 1132181561 M * dlippolt so if on one box you run the command with the Termianl1 argument 1132181570 M * dlippolt and on the other box u run the command with the Terminal2 argument 1132181581 M * dlippolt each should receive Terminal1's and Terminal2';s 1132181593 M * dlippolt inside vserver, it doesn't even receive its own 1132181597 M * dlippolt much less a peer's 1132181605 M * Bertl okay, first for the ping, what do I need there? 1132181618 M * dlippolt i dont understand the question 1132181629 M * Bertl you did a ping to 224.0.0.8, right? 1132181641 M * dlippolt yes 1132181643 M * dlippolt just a simple 1132181653 M * dlippolt actually, 228 1132181655 M * Bertl hmm, 228.0.0.4 1132181673 M * Bertl okay, what do I need for that to work? 1132181679 Q * ryker Quit: Leaving 1132181690 M * Bertl (basically minimum setup) 1132181692 M * dlippolt just be running the tool 1132181699 M * Bertl on the same host? 1132181701 M * dlippolt while running you should be able to simply 1132181705 M * dlippolt ping 228.0.0.4 1132181707 M * dlippolt on either 1132181714 M * Bertl okay, what about network? 1132181725 M * Bertl it sure requires some networking to be configured 1132181726 M * dlippolt well, on my setup, i'm not crossing boxes 1132181739 M * dlippolt i'm still trying to let the same vserver send and receive to itself 1132181741 M * dlippolt which isn't working 1132181743 M * SNy I can tell you that, depending on what you bind to, you cannot join the same mulitcast group multiple times on the same host 1132181776 M * dlippolt plz talk about that dependency (i.e. 'what you bind to') 1132181811 M * SNy as a matter of fact, I don't know about it exactly 1132181813 M * dlippolt tomcat clustering presumes that you can, since all nodes in the cluster operate onthe same multicast ip/port combo, and it supports runningmultiple nodes per box 1132181829 M * Bertl okay, folks, I know a little about multicasting (theory) but I never set up anything which actually uses that ... so ... 1132181845 M * Bertl consider the following setup: 1132181854 M * SNy FWIW, I have ACE under windows and linux 1132181862 M * Bertl - host A (currently running nothing) 1132181865 M * SNy and there must be subtle differences 1132181874 M * Bertl - host B configured like this: 1132181880 M * Bertl # ifconfig lo 127.0.0.1 1132181880 M * Bertl # ifconfig eth0 10.0.0.2 1132181884 M * SNy because I can join multiple times under windows but not under linux 1132181891 M * SNy I haven't looked into it 1132181896 M * Bertl - host A is similar config with 10.0.0.1 1132181900 M * SNy I just know that there's some catch to it 1132181906 M * Bertl # ping -c 1 10.0.0.1 1132181906 M * Bertl PING 10.0.0.1 (10.0.0.1) from 10.0.0.2 : 56(84) bytes of data. 1132181906 M * Bertl 64 bytes from 10.0.0.1: icmp_seq=0 ttl=64 time=21.812 msec 1132181923 M * Bertl works ... now what do I need to setup for a multicast ping? 1132181938 M * dlippolt you need soemthing to bind to a multicast ip 1132181946 M * dlippolt i.e. the tomcat-replication.jar command above 1132181954 M * dlippolt while that is running, ping the multicast ip 1132181957 M * Bertl okay, I start that on host A (how?) 1132181968 M * dlippolt its above inthis session, lemme find 1132181975 M * Bertl okay, I have that 1132181989 M * Bertl java -cp tomcat-replication.jar MCaster 224.0.0.4 45564 Terminal1 1132181998 M * dlippolt java -cp tomcat-replication.jar MCaster 228.0.0.4 45564 Terminal1 1132181998 M * dlippolt java -cp tomcat-replication.jar MCaster 228.0.0.4 45564 Terminal1 1132181998 M * dlippolt java -cp tomcat-replication.jar MCaster 228.0.0.4 45564 Terminal1 1132182002 M * dlippolt ya 1132182009 M * Bertl hmm okay 228 again ... 1132182016 M * dlippolt thats like the "server" which you can ping 1132182030 M * dlippolt but it acts like a publisher and a consumer 1132182037 M * dlippolt it should send and receive Terminal1 messages 1132182039 M * SNy try a more simple approach 1132182044 M * Bertl okay, it alternates now begin, sent received 1132182053 M * SNy there are basic examples for multicast on the net 1132182066 M * Bertl SNy: url? and tx btw! 1132182069 M * SNy I don't have my little test program at hand, unfortunately 1132182083 M * Bertl # ping -c 1 228.0.0.4 1132182083 M * Bertl connect: Network is unreachable 1132182103 M * dlippolt are your devices setup for multicast? 1132182106 M * dlippolt when you run ifconfig 1132182111 M * dlippolt do you see MULTICASYT 1132182114 M * Bertl no, that's what I asked you before :) 1132182124 M * SNy URL: google "multicast example program"; first hit 1132182129 M * dlippolt ifconfig eth0 multicast 1132182143 M * Bertl hmm, well, okay, eth0 has MULTICAST on 1132182159 M * dlippolt then u should be set 1132182168 M * Bertl nope, same result 1132182169 M * dlippolt is the java thing running before you try to piong? 1132182177 M * dlippolt *ping 1132182185 M * Bertl yes, it is repeating the whole stuff 1132182190 M * Bertl BEGIN TO RECEIVE 1132182190 M * Bertl SENT:Terminal156 1132182190 M * Bertl RECEIVED:Terminal156 FROM /216.114.78.253:45564 1132182195 M * SNy I don't think ping is the right tool for the job 1132182200 M * dlippolt i agree 1132182213 M * Bertl well, if even that doesn't work ... 1132182230 M * dlippolt which is why i'm using the tomcat-replication.jar as my multicasta test program 1132182244 M * dlippolt i dont ever get RECEIVED messages 1132182251 M * Bertl dlippolt: yes, but we already figured that this fails on the IPV6 socket 1132182281 M * dlippolt urs doesn't seem to fail :) 1132182283 M * Bertl ipv6 is disable inside linux-vserver guests, so that will not succeed anytime soon 1132182286 M * SNy quote: 1132182289 M * SNy /* use setsockopt() to request that the kernel join a multicast group */ 1132182289 M * SNy mreq.imr_multiaddr.s_addr=inet_addr(HELLO_GROUP); 1132182289 M * SNy mreq.imr_interface.s_addr=htonl(INADDR_ANY); 1132182289 M * SNy if (setsockopt(fd,IPPROTO_IP,IP_ADD_MEMBERSHIP,&mreq,sizeof(mreq)) < 0) { 1132182291 M * SNy perror("setsockopt"); 1132182294 M * SNy exit(1); 1132182300 M * SNy } 1132182313 M * SNy this is for the receiver 1132182321 M * SNy you need to explicitly join the group 1132182328 M * SNy the sender just sends to the group 1132182338 M * Bertl dlippolt: okay, I suggest you compile such test tools and try to make them really simple 1132182342 M * SNy so I revoke my statement about ping not being the right tool 1132182371 M * SNy I do have a simple tool called multirec 1132182389 M * Bertl once this is working between two hosts, and can be easily traced with strace (no linux-vserver involved) 1132182395 M * SNy that just outputs what it receives on a given multicast address 1132182414 M * Bertl we can continue to look at the guest situation ... 1132182414 M * SNy I do not have access to it ATM, though 1132183233 Q * cede Quit: leaving 1132183331 Q * shedi Quit: Leaving 1132183758 M * Hollow back 1132183778 M * harry to the future 1132183788 M * Hollow indeed 1132183789 M * Hollow :) 1132183838 M * Hollow Bertl: did you take a look at the persistant flag? 1132184399 M * Bertl yep, should work now ... let me upload a patch :) 1132184783 M * Bertl http://vserver.13thfloor.at/Experimental/delta-persist-feat02.diff 1132184795 M * Bertl http://vserver.13thfloor.at/Experimental/delta-namespace-feat01.diff 1132184814 M * Bertl http://vserver.13thfloor.at/Experimental/delta-dlimit-fix01.diff 1132184845 M * Bertl Hollow: anything else? 1132184899 M * Bertl harry: do you have your flux compensator ready? 1132185057 M * Hollow Bertl: looks good! i'll test it tomorrow, too tired already, good night everyone! 1132185119 M * Bertl k, good night!