1121648447 J * eXplasm2 explasm@p549F73B7.dip.t-dialin.net
1121648887 Q * eXplasm Ping timeout: 480 seconds
1121649233 J * Aiken_ ~james@tooax6-246.dialup.optusnet.com.au
1121649562 Q * Aiken Ping timeout: 480 seconds
1121649775 M * Bertl yay! who else noticed that there is a new util-vserver release, and didn't tell me?
1121650488 M * daniel_hozac sorry :)
1121650584 M * Bertl looks like excellent work ...
1121650595 A * Bertl is reading through the changelog now ...
1121654434 J * Doener_ ~doener@p54873D65.dip.t-dialin.net
1121654672 J * revenger ~revenger@p54B01572.dip0.t-ipconnect.de
1121654683 M * Bertl welcome revenger! :)
1121654688 M * revenger hallo :)
1121654738 M * Bertl basically linux-vserver is something like bsd jails (a very improved version of)
1121654763 M * Bertl it does similar as UML or XEN just without a guest kernel
1121654770 M * revenger yes I see..
1121654805 M * Rushmoom Did I somehow miss the question? ;-)
1121654827 M * Bertl Rushmoom: yes :) but don't worry ... I'll explain
1121654836 M * Rushmoom np :)
1121654851 M * revenger well I'm off anyway have to work tomorrow .. ergh today ;-)
1121654861 M * Bertl Rushmoom: revenger was interested in linux-vserver on another channel ...
1121654862 Q * Doener` Ping timeout: 480 seconds
1121654871 M * Bertl revenger: okay, good night then
1121654882 M * Bertl (and feel free to hang around)
1121654897 J * revenger_ ~joe@bulldog.infosys.de
1121654900 M * revenger_ well I will :-)
1121654902 P * revenger 
1121654934 N * revenger_ revenger
1121654945 M * Rushmoom Bertl: You're really doing a great job getting the userbase of vserver growing :)
1121654981 M * Bertl hmm ... well, and I'm not even advertizing yet :)
1121655059 M * Rushmoom Bertl: Yes, but making projects attractive for users as well as developers by maintaining them intensly is the best advertisment anyway :)
1121655161 M * Bertl in the beginning I was worried because the 'userbase' basically consisted of providers and .. hmm providers :)
1121655209 M * Bertl today, we have a lot of folks using it for their own projects (testing, development, security, ...)
1121655244 M * Rushmoom Bertl: Yep, they still make a big percentage I guess, but it's quite important that you attract a more broad group of users I think
1121655282 M * Bertl I think the simple isolation concept has many yet unknown 'features' to offer ... and after almost two years, it's still an interesting project for me ...
1121655348 M * Rushmoom Yes, and that's probably the point where it really can play out its advantages to UML & co
1121655395 M * Rushmoom But having ISPs amongst your users is also a good thing I guess, so you also get feedback from usage on highly productive systems...
1121655428 M * Bertl ah, don't get me wrong, I love ISP as they are really demand a 'working' product ...
1121655494 M * Rushmoom But they're probably not the ones being open to strongly experimental code, and involving themselves in development :)
1121655500 M * Bertl it was interesting to see how the freevps project (which is an early spinoff of linux-vserver) was classified by some folks ...
1121655563 M * Bertl they (ISPs) said: "nice features, but it isn't really stable ..." (that was some time ago, though)
1121655571 M * Medivh morning ;)
1121655577 M * Bertl morning Medivh!
1121655577 M * Rushmoom good morning :)
1121655640 M * Bertl and I'm trying to provide both, a stable and an experimental/devel version, so that those who want to test new features can, and those who want the 'stable' thing, can have it too ...
1121655664 M * Bertl so far that seems to work quite fine ...
1121655749 M * Medivh Bertl, thanks again for your help a few days ago btw... 2.0rc is running very well here now ;)
1121655765 M * Bertl rc-what actually?
1121655771 M * Medivh rc6 right now
1121655799 M * Bertl okay, just curious (I know about the cat :)
1121655857 M * Medivh is there a changelog or something for the 2.0 rc's?
1121655881 M * Bertl I'm on it .. i.e. there will be in an hour or so ...
1121655904 M * Medivh ah ;)
1121655928 M * Bertl somebody changed the url text to 'not-so up-to-date' :)
1121655940 M * Bertl and I just have to prove him wrong :)
1121655953 M * Medivh btw, about the stable/devel branch, that's indeed working great... can only speak from very pleasant experience over the last, well years it is already i guess ;)
1121655988 M * Medivh started back with good ol' ctx16 or so ;)
1121656010 M * Bertl hehe, yeah, the beginning was indeed fun ...
1121656033 M * Bertl we had all kind of races and oddities ...
1121656089 M * Medivh i remember when the solucorp server was always down when i wanted to download the patch... ;P
1121656131 M * Medivh talking about down, gets me to hosting... if there's anything like domains, hosting and the like needed for the project, feel free to ask me ;)
1121656150 M * Medivh (in case i didn't offer that already, don't remember really)
1121656217 M * Bertl not that I remember either ... but I wrote it down now :)
1121656243 M * Bertl thanks for the offer, currently there is no pressing need, but I'm sure I'll come back to that offer sooner or later
1121656264 M * Medivh any time... enough servers around here to cover most needs ;)
1121656307 M * Bertl it's also easy for me to rearrange, as all my stuff is contained in vserver guests :)
1121656316 M * Medivh so much for the users being ISPs btw ;P
1121656327 M * Medivh s/the/many/
1121656566 M * Medivh guess it's time for me to catch some sleep now for me though... got lost in php coding again not looking at the time ;)
1121656572 M * Medivh gn8 ;-)
1121656620 M * Bertl night Medivh!
1121656631 M * Medivh s/for me//... brr can't even type any more ;) cu!
1121657354 Q * _are_ Ping timeout: 480 seconds
1121661222 Q * ddlp Remote host closed the connection
1121662851 J * atsab ~as@lotes.vtu.lt
1121662858 M * Bertl welcome atsab!
1121663613 Q * comdata Read error: Connection reset by peer
1121663643 J * comdata ~mertins@gateway.scheller.de
1121666140 M * DaPhreak morning Bertl :)
1121666483 M * Rushmoom So, I'm off to bed...
1121666496 M * Rushmoom good day everyone :)
1121666517 P * Rushmoom 
1121667035 N * Aiken_ Aiken
1121667044 M * Bertl morning Aiken!
1121667051 M * Aiken hello
1121667162 A * Aiken mutters something about time zones
1121667201 M * Bertl yeah, bad idea ... guess I'm off to bed now :)
1121667216 M * Bertl so have a nice one everyone ... cya later ...
1121667221 N * Bertl Bertl_zZ
1121667222 M * Aiken bad timing
1121667224 M * Aiken good night
1121669424 J * Blogmeister ~Blogmeist@site.lycos.de
1121669459 M * DaPhreak morning Blogmeister 
1121669679 Q * comdata Quit: using sirc version 2.211+KSIRC/1.3.12
1121670846 P * Blogmeister 
1121672352 J * _are_ ~are@62.112.159.81
1121674957 J * Duckx ~duckx@195.75.27.158
1121675754 J * prae ~prae@ezoffice.mandriva.com
1121680664 Q * _ag_ Ping timeout: 480 seconds
1121681278 J * _ag_ ag@caladan.roxor.cx
1121681835 Q * _ag_ Read error: Connection reset by peer
1121681839 J * _ag_ ag@caladan.roxor.cx
1121682529 Q * _ag_ Ping timeout: 480 seconds
1121682716 M * eyck anyone tried initng with vserver?
1121684214 J * Rushmoom ~me@80-219-252-33.dclient.hispeed.ch
1121684232 M * Rushmoom morning folks!
1121684437 Q * Aiken Ping timeout: 480 seconds
1121684500 M * eyck morning
1121684568 M * matti Hi eyck.
1121686408 J * _ag_ ag@caladan.roxor.cx
1121687598 N * BobR_oO_ BobR
1121687843 M * Hollow morning
1121688524 J * jsambrook ~jsambrook@aelfric.plus.com
1121688782 N * BobR BobR_afk
1121688941 M * Rushmoom morning Hollow!
1121689017 J * brc bruce@201008088244.user.veloxzone.com.br
1121689201 Q * Hollow Quit: /dev/zero
1121689370 J * Hollow ~Hollow@home.xnull.de
1121690433 J * wurd ~kvlt@modemcable181.93-202-24.mc.videotron.ca
1121690660 M * wurd hi, i'd like to know what the "vserver build" command does exactly?
1121690713 M * wurd does it create the file architecture in /vservers/ ?
1121690759 M * wurd (for example /vservers/myvserver/ would contain etc,usr,bin,home,mnt..)
1121690812 P * jsambrook 
1121692348 M * Rushmoom wurd: Yes, it creates a very basic directory structure, bot with nothing in it
1121692401 M * Rushmoom wurd: It contains an empty /etc, empty /proc and a /dev with some stuff in it
1121692434 M * Rushmoom wurd: But when actually installing a distro on your vserver, you probably are going to overwrite these files anyway
1121692532 M * Rushmoom wurd: The main purpose of "vserver build" is to create the skeleton, including proper configuration of interface aliases / IP adresses, context id, etc..
1121692554 M * wurd ok, well i,ve downloaded a rootfs rpm
1121692563 M * wurd so does this mean i have to skip the "build" part ?
1121692588 M * wurd (rootfs rpm : http://www.acytec.com/files/rootfs/vservers/ )
1121692610 M * Rushmoom wurd: No
1121692654 M * Rushmoom wurd: Just create the skeleton with "vserver build", and after that, extract the rootfs tarball to /vserver/yourvserver
1121692722 M * wurd ok thanks.
1121692727 M * Rushmoom wurd: At least, that's what I would do. There are some way more competent people in this channel, but some of them aren't awake yet ;-)
1121692929 M * wurd do i absolutely have to use "apt-rpm" in my build ?
1121692944 M * wurd i've builded a vserver before, but it was on redhat (now i'm on mandrake)
1121692948 M * wurd and i was told to use apt-rpm
1121692955 M * wurd (in the build command)
1121692963 M * wurd i was wondering if there existed another way ?
1121692995 M * Rushmoom wurd: No idea, I never used redhat or mandrake as vserver guest
1121693005 N * Bertl_zZ Bert_oO
1121693013 M * wurd what did your build command look like?
1121693017 M * Rushmoom vserver your-vserver build -m skeleton --hostname your-vserver-hostname --initstyle plain --context 12345 --interface your-vserver=eth0:192.168.1.1/24
1121693068 M * Rushmoom Talking about the devil.. morning Bertl! :)
1121693109 M * Bert_oO morning, does anybody read the docu?
1121693121 M * Bert_oO wurd: http://linux-vserver.org/alpha+util-vserver
1121693131 M * Rushmoom Fair question...
1121693264 N * Bert_oO Bertl_oO
1121694127 M * wurd is this the latest version of the utils? :  util-vserver-0.30.208.tar.bz2 
1121694148 M * wurd (found it in http://www.13thfloor.at/~ensc/util-vserver/files/alpha/)
1121695578 N * pg`aw|cereal cereal
1121697175 Q * _are_ Ping timeout: 480 seconds
1121699888 M * wurd Rushmoom are you still here ?
1121699895 M * wurd i'm just wondering what the "--context" option means
1121700809 M * Hollow wurd: maybe the context id?
1121701031 Q * Duckx Quit: Leaving
1121701039 N * BobR_afk BobR
1121702280 J * ydupont ~dupont-y@tomintoul.cri.univ-nantes.fr
1121702810 Q * ydupont Quit: Leaving
1121703478 M * Rushmoom wurd: 1 hour late, but anyway: This sets the context id under which the vserver will be running. Means, you can choose it yourself, only has to be different from existing context ids
1121703507 M * Rushmoom wurd: And you obviously still haven't read the docs :(
1121704336 M * wurd you mean ALL docs ?
1121705139 M * Rushmoom How about the relevant ones?
1121705189 N * BobR BobR_oO
1121707961 J * erwan_ho ~erwan@konilope.dyndns.org
1121708855 J * Vudumen vudumen@perverz.hu
1121712409 J * jkl_ eric@c-67-165-222-86.hsd1.co.comcast.net
1121712931 Q * erwan_ho Remote host closed the connection
1121712990 Q * Zoiah Quit: simon says: rehashing
1121716812 Q * jkl_ Ping timeout: 480 seconds
1121717589 Q * Doener_ Quit: Leaving
1121717795 J * Doener ~doener@p54873D65.dip.t-dialin.net
1121717966 J * Borg-Number-One ~BorgNr1@p54B8A2C0.dip0.t-ipconnect.de
1121717987 M * Borg-Number-One Greetings
1121718364 Q * wurd Ping timeout: 480 seconds
1121719759 M * aba if "cat /proc/mounts" contains /tmp inside a vserver, "chcontext --xid $context cat /proc/mounts" should also contain it (of course, with a different path)?
1121719984 M * daniel_hozac vnamespace -e $context cat /proc/mounts should, i think.
1121720007 M * daniel_hozac or maybe you need both of them.
1121720041 M * aba daniel_hozac: thanks
1121720050 M * daniel_hozac hmm, neither works for me.
1121720085 M * aba vnamespace works for me
1121720812 J * ddlp ~daniel@205.234.152.200
1121722442 Q * Rushmoom Read error: Connection reset by peer
1121723077 J * yarihm ~yarihm@80-218-5-17.dclient.hispeed.ch
1121723137 Q * eXplasm2 Remote host closed the connection
1121723368 J * Aiken ~james@tooax6-208.dialup.optusnet.com.au
1121723455 P * Borg-Number-One 
1121726563 N * Bertl_oO Bertl
1121726578 M * Bertl evening folks!
1121728040 P * yarihm Leaving
1121729554 M * Greek0 Bertl: do you have some time/motivation for another vserver question?
1121729575 M * Bertl always :)
1121729580 M * Greek0 kernel patch question, that is
1121729601 M * Greek0 kernel/vserver/dlimit.c, why is DL_HASH_SIZE == 13?
1121729621 M * Bertl lucky number, no?
1121729675 M * Greek0 well, you do % DL_HASH_SIZE on every hash calculation, if it were a power of 2 it would be faster.. or is it just too unimportant to optimize here?
1121729694 M * Bertl most likely it would not turn out to be an optimization
1121729704 M * Bertl (but you're welcome to try :)
1121729724 A * Greek0 will keep it in his notes :)
1121729728 M * Bertl the thing with hashes is, that most hash functions are bad ...
1121729782 M * Bertl ((xid ^ (unsigned long)sb) % DL_HASH_SIZE)
1121729789 M * Bertl this is our 'hash function'
1121729809 M * Greek0 I'm not sure if I really understand how that % 13 maps the source to the destination
1121729824 M * Bertl the super block might be aligned at 32, 64, 128 bytes
1121729833 M * Bertl (depends on the arch)
1121729845 M * Greek0 but it's more the mathematical "I should think about this a bit" type of "i don't really get this"
1121729864 M * Bertl the xid will be somewhere in the range 2-65535
1121729872 M * Greek0 ok, so with 13 you get a better hashing function?
1121729897 M * Bertl yes, even perfectly aligned values (0, 32, 64, 128 ...)
1121729911 M * Greek0 mhm, makes sense
1121729911 M * Bertl will give a nice distribution over 0-12
1121729987 M * Bertl so while the %13 might be a cycle longer than a &15
1121730009 M * Bertl it will save a bunch of cycles because of the better distribution
1121730036 M * Greek0 yep
1121730074 M * Bertl so in this case, consider the %13 part of the hash function, not part of the actual hash limit (as it is fixed anyway)
1121730223 M * Greek0 ok, another thing I didn't really get last week..
1121730239 M * Greek0 vx_map_[ug]id
1121730301 M * Greek0 I understand the return statement, but not the if clause before it
1121730519 M * Bertl well, it is defined (somewhere) that 'unknown' uids have to be mapped to nobody
1121730550 M * Bertl nobody:x:65534:65534:Nobody:/home:/bin/sh
1121730580 M * Bertl (grep nobody /etc/passwd)
1121730580 M * Greek0 ah, ok
1121730583 M * Greek0 yep
1121730586 M * Greek0 makes some sense then
1121730644 M * Bertl (not sure if that mapping is really correct for 24/32 bit uids though ...
1121730697 M * Bertl maybe some 'standard' also defines nobody to be 65534
1121730796 M * Greek0 well, with CONFIG_INOXID_UGID24 you won't get 65534 with that function..
1121730923 M * Bertl yes, that's what I meant ...
1121730962 M * Bertl (but nevertheless it will be a hopefully 'unused' value ...)
1121731035 M * Greek0 mm
1121731061 M * Greek0 ok, that's pretty much it for now..
1121731074 M * Greek0 thanks for your patience
1121731168 M * Bertl it's always a pleasure ...