TheGrebs.com:IRC Logs:#xen:2006:07:06

---	Log	opened Thu Jul 06 00:00:53 2006
00:08	\|-\|	ns [~nivedita@c-67-171-181-157.hsd1.or.comcast.net] has quit [Quit: Leaving]
00:08	\|-\|	ns [~nivedita@c-67-171-181-157.hsd1.or.comcast.net] has joined #xen
00:10	\|-\|	ns [~nivedita@c-67-171-181-157.hsd1.or.comcast.net] has quit [Quit: ]
00:10	\|-\|	ns [~nivedita@c-67-171-181-157.hsd1.or.comcast.net] has joined #xen
00:12	\|-\|	ns [~nivedita@c-67-171-181-157.hsd1.or.comcast.net] has quit [Quit: ]
00:27	\|-\|	muli [~muli@87.69.40.180.cable.012.net.il] has quit [Quit: BitchX-1.0c19 © 1996-2002 Colten Edwards]
00:29	\|-\|	Basic_py [~Basic@astound-64-85-227-99.ca.astound.net] has quit [Ping timeout: 480 seconds]
00:30	\|-\|	mpo [mpolla@jt10-28.tky.hut.fi] has quit [Quit: leaving]
00:39	\|-\|	segher [~segher@dslb-084-056-141-255.pools.arcor-ip.net] has joined #xen
00:40	\|-\|	Basic_py [~Basic@warden.real-time.com] has joined #xen
00:43	\|-\|	gerrit_ [~gerrit@67.160.146.170] has joined #xen
00:44	\|-\|	jforbes_ changed nick to jforbes
00:51	\|-\|	sdague [~sdague@user-12lc6iq.cable.mindspring.com] has quit [Remote host closed the connection]
00:53	\|-\|	sdague [~sdague@user-12lc6iq.cable.mindspring.com] has joined #xen
01:09	\|-\|	sdague_ [~sdague@user-12lc6iq.cable.mindspring.com] has joined #xen
01:11	\|-\|	sdague [~sdague@user-12lc6iq.cable.mindspring.com] has quit [Read error: Connection reset by peer]
02:06	\|-\|	Quentusrex [~chatzilla@c-24-22-212-181.hsd1.mn.comcast.net] has joined #xen
02:18	\|-\|	sdog [~sdog@62.58.98.210] has joined #xen
02:27	\|-\|	Sir_Ahzz [~ahzz@24.0.241.237] has quit [Remote host closed the connection]
02:28	<Quentusrex>	I currently have a network setup of 5 servers that include dns/mysql/mail/apache/voip/etc servers. I would like to be able to migrate them all onto the xen system to allow better uptime and allow me to maintain the server hardware without turning off that service for the extent of the work. After reading through some of the manuals. It seems that I want to configure Xen with LVM. are there...
02:28	<Quentusrex>	...any other things that I need to read up on before I start the migration? I'd want to be able to dynamically change the size of the 'drives' or partitions assigned to each system and change the RAM assigned as well.
02:38	\|-\|	Basic_py [~Basic@warden.real-time.com] has quit [Quit: Leaving]
02:55	\|-\|	parabelboi [~parabelbo@blueice1n1.de.ibm.com] has joined #xen
03:41	\|-\|	tschwinge [cpts2523@l2.cip.ei.uni-stuttgart.de] has joined #xen
04:01	\|-\|	pvanhoof [~pvanhoof@62.58.98.194] has joined #xen
04:05	\|-\|	fabien_ [~fabien@nat5.emn.fr] has quit [Quit: Quitte]
04:12	\|-\|	athomas [~athomas@host86-136-84-147.range86-136.btcentralplus.com] has joined #xen
04:29	\|-\|	womble [~mpalmer@220.233.135.106] has quit [Quit: What's behind the round window...]
04:33	\|-\|	ivan [~ikelly@213.228.247.71] has joined #xen
04:58	\|-\|	Marco [~marco@249-174.surfsnel.dsl.internl.net] has quit [Quit: Client exiting]
05:19	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has joined #xen
05:22	\|-\|	horms [~horms@210.128.90.14] has quit [Quit: Leaving]
05:42	\|-\|	ivan [~ikelly@213.228.247.71] has quit [Remote host closed the connection]
05:45	\|-\|	ivan [~ikelly@213.228.247.71] has joined #xen
05:47	\|-\|	mooli [~muli@nesher3.haifa.il.ibm.com] has quit [Quit: My damn controlling terminal disappeared!]
05:49	\|-\|	muli [~muli@nesher3.haifa.il.ibm.com] has joined #xen
06:01	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has quit [Quit: mdday__]
06:26	\|-\|	freitag [~f@84.137.128.3] has quit [Quit: I quit]
06:29	\|-\|	lilo2_ [~0710AAD4@tor-irc.dnsbl.oftc.net] has joined #xen
06:30	\|-\|	lilo2 [~0710AAD4@tor-irc.dnsbl.oftc.net] has quit [Ping timeout: 480 seconds]
06:41	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has joined #xen
06:58	\|-\|	bernarde [~bernarde@32.104.18.240] has joined #xen
07:28	\|-\|	SLot [~SLot___@gustavo.gruposim.com.br] has quit [Read error: Connection reset by peer]
07:29	\|-\|	SLot [~SLot___@gustavo.gruposim.com.br] has joined #xen
07:40	\|-\|	jimix [~jimix@ip139.152.45.216.suscom.net] has joined #xen
07:58	\|-\|	endre_ [~endre@blueice4n1.de.ibm.com] has joined #xen
07:59	<endre_>	hi all! I'd like to use Xen, do I need any extra kernel configuration for it?
08:00	<endre_>	point 5.1 in the FAQ is not clear to me
08:03	<endre_>	e.g. I have no arch/xen in the kernel source tree...
08:06	<ivan>	endre_: you using stock kernel? xen isn't in mainline yet
08:07	\|-\|	sdague_ changed nick to sdague
08:08	<endre_>	ivan: this can be the problem :) I use 2.6.16-gentoo-r7
08:08	<ivan>	and therein lies your problem
08:08	<ivan>	you understand how xen handles kernels?
08:08	\|-\|	msinhore [~msinhore@correio.samurai.com.br] has joined #xen
08:09	<endre_>	in what level? I am not experienced in kernel development....
08:09	<ivan>	not development, jsut how xen uses them
08:10	<endre_>	the host kernel or the guest kernel?
08:10	<ivan>	really should put this on a wiki page
08:10	<ivan>	but anyhow,
08:10	<ivan>	you have the xen kernel which runs on top of hardware
08:10	\|-\|	sdog [~sdog@62.58.98.210] has quit [Quit: Everything is a Fscking DNS Problem]
08:10	<ivan>	on top of this runs your linux-xen kernels
08:11	<ivan>	there is one special kernel called the linux-xen0 or dom0 kernel
08:11	<ivan>	this has full access to hardware
08:11	<endre_>	ok, it should be the kernel of the host machine i guess
08:11	<ivan>	then you have linux-xenU or domU kernels which run beside the dom0 on top of the xen kernel
08:11	<ivan>	the xen kernel is it's own kernel, unrelated to linux
08:12	<ivan>	usually your current linux install will end up being the dom0
08:12	<endre_>	ok
08:12	<ivan>	filesystem wise
08:12	<ivan>	but using the linux-xen0 kernel
08:12	<endre_>	so every linux kernel will see a virtual hadrware?
08:13	<ivan>	yes and no
08:13	<ivan>	the linux-xen0 will see the real hardware, but internally xen will handle communication
08:13	<ivan>	like interrupts and stuff, but you dont need to know that
08:14	<ivan>	other linux-xenU kernels will see virtual hardware using a frontend driver
08:14	<ivan>	the dom0 will have a backend driver to handle requests to this virtual hardware
08:16	\|-\|	Sir_Ahzz [~ahzz@24.0.241.237] has joined #xen
08:16	\|-\|	mode/#xen [+o Sir_Ahzz] by ChanServ
08:17	<endre_>	can I use mainstream kernels as linux-xenU
08:17	<endre_>	?
08:17	<ivan>	no, they need to have patch applied
08:18	<ivan>	http://svn.skynet.ie/~ikelly/MinixOnXen/minixxen.html <- there an animation on this i used to explain it before
08:18	<endre_>	ok, thx
08:19	\|-\|	pvanhoof [~pvanhoof@62.58.98.194] has quit [Ping timeout: 480 seconds]
08:22	<endre_>	and how can I install a distro as a domU?
08:23	<endre_>	I think they not contain xen0 kernels....
08:23	<ivan>	yup
08:23	<endre_>	sorry, I mean xenU kernels
08:24	<ivan>	no, but the domU kernel is usually stored in dom0
08:24	<ivan>	when a domU is started (from dom0) the kernel is loaded into memory my the dom0 and it is told where its filesystem is
08:25	<ivan>	which could be a LVM volume, or a flat file or an ordinary partition
08:25	<endre_>	ok
08:25	<ivan>	not sure how to install some distros, but with debian youd use debootstrap from dom0
08:26	<endre_>	unfortunately I have to install fedora....
08:27	<ivan>	there are many ways, i just dont know them :)
08:28	<endre_>	ok, I think fedora kernels....
08:28	<endre_>	so they are highly wizarded
08:28	<endre_>	I mean I have no idea how to configure that
08:28	<endre_>	in order to make the whole distro working well
08:29	<endre_>	I think for this purpose I'll use VMWare....
08:29	<endre_>	although I think Xen is really a cool stuff :)
08:30	<ivan>	VMWare does have a lot smaller learning curve
08:31	<endre_>	yes
08:32	<endre_>	I think now it will fit better for a Windows like distro
08:33	<endre_>	but for hacking, xen is better, I think :)
08:33	<ivan>	vmware and xen do different things
08:33	<ivan>	but if you want a quick solution vmware will probably be easier
08:34	<endre_>	sure :)
08:34	<endre_>	ok for debian or gentoo I'd use Xen
08:34	<endre_>	because they are not overwizarded
08:34	<endre_>	and I have experiences with them
08:34	<ivan>	you can even use xen inside of vmware
08:34	<ivan>	for testing/devel
08:34	<ivan>	anyhow, im going for food
08:34	<endre_>	ok
08:44	\|-\|	Sir_Ahzz [~ahzz@24.0.241.237] has quit [Quit: Client exiting]
08:46	<robster>	hhhm. anyone else having trouble building from xen-3.0-testing ?
09:03	\|-\|	ns [~nivedita@67.171.181.157] has joined #xen
09:09	\|-\|	Quentusrex [~chatzilla@c-24-22-212-181.hsd1.mn.comcast.net] has quit [Quit: Chatzilla 0.9.74 [Firefox 1.5.0.4/2006050817]]
09:20	\|-\|	bernarde [~bernarde@32.104.18.240] has quit [Quit: Konversation terminated!]
09:22	\|-\|	bernarde [~bernarde@32.104.18.240] has joined #xen
09:26	\|-\|	muli_ [~muli@87.69.40.180.cable.012.net.il] has joined #xen
09:29	\|-\|	harry_ [~harry@blueice4n2.uk.ibm.com] has joined #xen
09:32	\|-\|	rharper [~rharper@pixpat.austin.ibm.com] has joined #xen
09:43	\|-\|	SLot [~SLot___@gustavo.gruposim.com.br] has quit [Ping timeout: 480 seconds]
09:48	\|-\|	SLot [~SLot___@gustavo.gruposim.com.br] has joined #xen
10:06	\|-\|	parabelboi [~parabelbo@blueice1n1.de.ibm.com] has quit [Read error: Connection reset by peer]
10:14	\|-\|	aliguori [~anthony@70.116.9.243] has quit [Remote host closed the connection]
10:25	\|-\|	aliguori [~anthony@70.116.9.243] has joined #xen
10:30	<rharper>	hey aliguori
10:33	\|-\|	CosmicRay [~jgoerzen@69.44.136.67] has joined #xen
10:33	<CosmicRay>	I'm not quite understanding the memory balooning in xen.
10:34	<CosmicRay>	when I set the memory parameter for a domU, is that setting the minimum, maximum, or both?
10:34	<CosmicRay>	and how can I set the other (minimum or maximum) value in the config file?
10:34	<aliguori>	rharper, howdy
10:34	<rharper>	CosmicRay: memory sets the amount, max_mem sets the upper bound, so xm create domU1 memory=128 max_mem=512, would allow the domain to have up to 512, though linux won't see any more memory than the initial 128
10:35	<aliguori>	s/max_mem/maxmem/g
10:35	<aliguori>	:-)
10:35	<aliguori>	last time i checked at least
10:35	<rharper>	aliguori: yeah
10:35	[~]	rharper double checks
10:35	<CosmicRay>	rharper: I understand setting it on the command-line, but I want to but this in a configfile under /etc/xen/auto so it comes up automatically with those settings
10:35	<CosmicRay>	and there is only documented a "memory" setting for the config file
10:35	<rharper>	you can set those in your domU config
10:35	<rharper>	aliguori: yeah, maxmem=
10:36	<aliguori>	rharper, i only know that because i was tracking down a bug recently :-)
10:36	<rharper>	CosmicRay: you can put both memory and maxmem in the config file
10:36	<rharper>	aliguori: heh
10:36	<rharper>	aliguori: was looking that up for the NUMA balloon overhead benchmark
10:36	<aliguori>	rharper, so did you ever get more clarification from ian on what exactly he meant?
10:37	<rharper>	overhead was what he was after
10:37	<rharper>	it turned out, without numa, balloon from 128m to 15G is 4ms, with numa is 5ms
10:37	<CosmicRay>	rharper: oh, ok.
10:37	<rharper>	he said he wants me to run that a few more times
10:37	<rharper>	CosmicRay: you can check that it gets it by xm create domUconf -n (which will pretty print the config without launching the domain)
10:37	<aliguori>	rharper, interesting.. so there's overhead but it's so fast does it really matter?
10:38	<rharper>	aliguori: yep
10:38	<CosmicRay>	rharper: so I can put in the config file anything I could put on the xm set command line?
10:38	<rharper>	CosmicRay: no
10:38	<aliguori>	xm set?
10:38	\|-\|	aw [~awilliam@ftcrel4.hp.com] has joined #xen
10:38	<rharper>	xm help create gives a list
10:39	<rharper>	not sure if that is exhaustive, or up-to-date, but it should give you a good starting point
10:39	<CosmicRay>	rharper: wow, most of these are not documented in the xen user docs pdf
10:40	<rharper>	CosmicRay: sounds like a good patch if you want to send that to xen-devel
10:40	<CosmicRay>	rharper: so let's say I set memory=128 and maxmem=2048, and I have swap enabled on a domU. How does the hypervisor decide when to baloon up the RAM? when the machine would have had to use swap?
10:41	<aliguori>	CosmicRay, it never decides. you have to manually balloon
10:41	<CosmicRay>	oh, well then what is the point of maxmem?
10:41	<rharper>	CosmicRay: a hard limit
10:41	<aliguori>	CosmicRay, b/c if you don't specify maxmem, you cannot balloon beyond the initial allocation
10:41	<CosmicRay>	I was sure I saw it balloon up automatically when there was no swap during the OOM driver
10:41	<CosmicRay>	aliguori: ah
10:41	<rharper>	memory=maxmem, then you can balloon down
10:42	<CosmicRay>	rharper: and the system will automatically balloon down then?
10:42	<aliguori>	CosmicRay, and, a guest can always balloon itself up to maxmem
10:42	<rharper>	aliguori: did anyone ever hook OOM into balloon driver ?
10:42	<aliguori>	i don't think so
10:42	<aliguori>	and i doubt it would actually work
10:42	<aliguori>	b/c by the time the oom killer is invoked, things are very bad
10:42	<rharper>	yeah
10:43	<aliguori>	rharper, riel's got a pretty interesting memory pressure algorithm he's presenting at ols
10:43	<aliguori>	that ought to be a good way of autoballooning
10:43	<rharper>	aliguori: yeah?, sounds interesting
10:43	<riel>	aliguori: feel free to pass the paper on to rharper :)
10:44	<rharper>	hey riel
10:44	<aliguori>	riel, you never sent me the paper :-)
10:44	<riel>	wait, I didn't?
10:44	<rharper>	hehe
10:44	<aliguori>	no, you just explained it in irc
10:44	<CosmicRay>	aliguori: under what circumstances would a guest balloon itself up?
10:44	<aliguori>	unless you sent it to my ibm account which i'm not currently checking :-)
10:48	<CosmicRay>	I guess I still don't get maxmem, since you can xm set it even.
10:48	<CosmicRay>	so if you needed to balloon past there, you could manually do so.
10:48	<CosmicRay>	but there is absolutely no automated process that does this ballooning?
10:48	<CosmicRay>	either up or down?
10:48	<riel>	CosmicRay: that'd need memory hotplug support in Linux
10:48	<aliguori>	CosmicRay, you cannot change maxmem once a domain has started
10:48	<riel>	CosmicRay: there is no automated process yet, but I know how to do it :)
10:49	<CosmicRay>	aliguori: the manpage for xm claims that you can xm set mem-max for it
10:49	<aliguori>	CosmicRay, I know, but it doesn't work
10:49	<riel>	CosmicRay: you can, but the Linux guest doesn't know how to take advantage of it
10:49	<CosmicRay>	urk.
10:49	\|-\|	hollisb [~hollisb@192.35.232.241] has joined #xen
10:49	<rharper>	hey hollisb
10:50	<hollisb>	hey
10:50	<CosmicRay>	so memory config in xen is essentially static, barring manual intervention? no automatic balloon either up or down?
10:51	<aliguori>	CosmicRay, yes
10:53	<riel>	you could implement automatic balloon in some system management tool on top of Xen
10:53	<riel>	but I don't think anybody has done that yet
10:53	\|-\|	sonorous [~sonorous@itanic.ipv6.attenuate.org] has joined #xen
10:54	<aliguori>	ballooning is too dangerous
10:56	<riel>	oh?
10:56	<aliguori>	well, if you balloon down, you run the risk of invoking the OOM killer
10:56	<aliguori>	and taking down your system
10:56	[~]	rharper did that on the NUMA experiement
10:57	<aliguori>	riel, if you've got sufficient swap, it may be okay
10:57	[~]	aliguori hasn't used swap in a system for a long time
10:58	<riel>	aliguori: well, that's why you have Committed_AS in /proc/meminfo
10:59	<aliguori>	riel, that's still just a guess though right?
10:59	<rharper>	AS being address space ?
10:59	<riel>	no, that's accurate
10:59	<riel>	rharper: yeah
11:00	<CosmicRay>	is dom0 ballooning any different?
11:00	<riel>	it's the maximum amount of RAM+swap that may be needed to support the current virtual memory allocations
11:00	<CosmicRay>	does it automatically balloon up or down?
11:00	<CosmicRay>	and what's the point of the minimum memory size for it?
11:00	<riel>	CosmicRay: it may automatically balloon down to create space for guests
11:00	<CosmicRay>	ah!
11:00	<CosmicRay>	so dom0 can automatically balloon down, but not any domU?
11:00	<rharper>	the balloon down is done by the xen tools when you create a domain
11:01	<rharper>	not a Linux thing
11:01	<riel>	well, the dom0 ballooning is done through xend, same as manual ballooning
11:02	<CosmicRay>	why does it balloon down dom0 in that situation, but not a domU?
11:02	<riel>	because dom0 gets all the system's memory at startup
11:02	<riel>	also, which domU(s)? and by how much?
11:03	\|-\|	athomas [~athomas@host86-136-84-147.range86-136.btcentralplus.com] has quit [Quit: Leaving]
11:04	<CosmicRay>	riel: ahh. so it was not really necessary for me to limit the initial dom0 RAM with kernel parameters.
11:05	<CosmicRay>	s/kernel/xen/
11:15	\|-\|	mastermind [~mastermin@mastermind.kaltenbrunner.cc] has joined #xen
11:16	\|-\|	lazyb0y [~henning@85.176.113.244] has joined #xen
11:16	\|-\|	Q3aiml_ [~Q3aiml@c-67-190-251-183.hsd1.mn.comcast.net] has joined #xen
11:19	\|-\|	Q3aiml [~Q3aiml@67.190.251.183] has quit [Ping timeout: 480 seconds]
11:28	<riel>	(XEN) (file=traps.c, line=1143) Domain attempted WRMSR 0000000000000199 from 00000000:00000f26 to 00000000:00000c20.
11:28	[~]	riel grumbles
11:29	<riel>	why does the hypervisor never tell me _which_ domain committed the error?
11:30	\|-\|	xai [~pasta@user-0vvdfs3.cable.mindspring.com] has joined #xen
11:31	\|-\|	Q3aiml [~Q3aiml@67.190.251.183] has joined #xen
11:33	\|-\|	Q3aiml_ [~Q3aiml@c-67-190-251-183.hsd1.mn.comcast.net] has quit [Ping timeout: 480 seconds]
11:33	<mastermind>	riel: because that would be useful to know :-(
11:40	\|-\|	endre_ [~endre@blueice4n1.de.ibm.com] has quit [Quit: leaving]
12:01	\|-\|	bernarde [~bernarde@32.104.18.240] has quit [Read error: Connection reset by peer]
12:08	\|-\|	Netsplit charon.oftc.net <-> arion.oftc.net quits: VS_ChanLog
12:09	\|-\|	Netsplit over, joins: VS_ChanLog
12:15	\|-\|	lilo2_ [~0710AAD4@tor-irc.dnsbl.oftc.net] has quit [Remote host closed the connection]
12:16	\|-\|	lilo2 [~0710AAD4@tor-irc.dnsbl.oftc.net] has joined #xen
12:24	\|-\|	Q3aiml [~Q3aiml@67.190.251.183] has quit [Ping timeout: 480 seconds]
12:24	\|-\|	Q3aiml [~Q3aiml@c-67-190-251-183.hsd1.mn.comcast.net] has joined #xen
12:54	\|-\|	harry_ [~harry@blueice4n2.uk.ibm.com] has quit [Quit: Leaving]
12:54	\|-\|	lilo2 [~0710AAD4@tor-irc.dnsbl.oftc.net] has quit [Read error: Connection reset by peer]
12:56	\|-\|	lilo2 [~0710AAD4@tor-irc.dnsbl.oftc.net] has joined #xen
13:14	\|-\|	mastermind [~mastermin@mastermind.kaltenbrunner.cc] has quit [Ping timeout: 480 seconds]
13:15	\|-\|	mastermind [~mastermin@mastermind.kaltenbrunner.cc] has joined #xen
13:15	\|-\|	pvanhoof [~pvanhoof@84.192.226.126] has joined #xen
13:16	\|-\|	bernarde [~bernarde@201.44.214.48] has joined #xen
13:17	\|-\|	tobi [~tobi@81.6.49.116] has joined #xen
13:43	<tessier__>	Can someone remind me of the command to use mercurial to get the latest xen 3.0?
13:43	<tessier__>	It is conveniently hidden in an obscure page on the web somewhere which I can never find.
13:44	<hollisb>	http://xenbits.xensource.com/
13:44	<hollisb>	not that obscure
13:44	<hollisb>	hg clone http://xenbits.xensource.com/xen-unstable.hg
13:44	\|-\|	Tv [~tv@38.98.1.19] has joined #xen
13:46	<tessier__>	hg clone <mercurla url> is the obscure bit. :)
13:46	<tessier__>	mercurial, rather
13:47	<hollisb>	well, not if you've ever used mercurial :)
13:47	<tessier__>	You assume everyone knows and uses mercurial. Given just a .hg file I never remember what to do because xen is the only project that uses it. :)
13:48	\|-\|	bernarde [~bernarde@201.44.214.48] has quit [Ping timeout: 480 seconds]
13:49	<riel>	tessier__: git and bk also use "clone", in exactly the same way
13:49	<tessier__>	Two more tools which I never have occasion to use, not being a kernel developer. :)
13:50	[~]	tessier__ is about to network boot this diskless box into xen and mount disk with AoE
13:50	<tessier__>	And it's gonna be sweet.
14:02	\|-\|	Basic_py [~Basic@warden.real-time.com] has joined #xen
14:06	\|-\|	msinhore [~msinhore@correio.samurai.com.br] has quit [Remote host closed the connection]
14:10	\|-\|	bernarde [~bernarde@201.44.214.48] has joined #xen
14:21	\|-\|	visik7 [~visi@82.60.8.39] has joined #xen
14:32	\|-\|	msinhore [~msinhore@correio.samurai.com.br] has joined #xen
14:34	\|-\|	msinhore [~msinhore@correio.samurai.com.br] has quit [Quit: ]
14:34	\|-\|	msinhore [~msinhore@correio.samurai.com.br] has joined #xen
14:38	<murb>	/win 24
14:38	<murb>	sorry
14:53	<@anticw>	/win 95
14:55	<cfreak>	:)
14:55	<cfreak>	bluescreen
14:57	\|-\|	bernarde [~bernarde@201.44.214.48] has quit [Remote host closed the connection]
14:59	\|-\|	visik7 [~visi@82.60.8.39] has quit [Ping timeout: 480 seconds]
15:14	\|-\|	visik7 [~visi@82.54.169.15] has joined #xen
15:25	\|-\|	tobi [~tobi@81.6.49.116] has quit [Quit: tobi]
15:49	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has quit [Read error: Connection reset by peer]
15:50	\|-\|	lazyb0y_ [~henning@85.176.124.188] has joined #xen
15:51	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has joined #xen
15:57	\|-\|	jcm_ [~jcm@salud.ucsd.edu] has joined #xen
15:57	\|-\|	lazyb0y [~henning@85.176.113.244] has quit [Ping timeout: 480 seconds]
16:21	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has left #xen []
16:21	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has joined #xen
16:25	<jcm_>	/whois jcm
16:25	\|-\|	jcm_ changed nick to jcmdev0
16:46	\|-\|	zwane_wrk [~zwane@24.85.144.101] has quit [Ping timeout: 480 seconds]
16:54	\|-\|	felix [~felix@chtwpe01bba-142177239041.pei.aliant.net] has joined #xen
16:59	\|-\|	CosmicRay [~jgoerzen@69.44.136.67] has quit [Quit: Client exiting]
17:03	\|-\|	hollisb [~hollisb@192.35.232.241] has quit [Quit: leaving]
17:09	\|-\|	anticw [~anticw@c-71-202-63-228.hsd1.ca.comcast.net] has quit [Quit: Reconnecting]
17:09	\|-\|	anticw [~anticw@c-71-202-63-228.hsd1.ca.comcast.net] has joined #xen
17:28	\|-\|	felix [~felix@chtwpe01bba-142177239041.pei.aliant.net] has quit [Quit: Client exiting]
17:30	\|-\|	mdday__ [~mdday@cpe-024-163-120-222.nc.res.rr.com] has quit [Quit: mdday__]
18:00	\|-\|	rharper [~rharper@pixpat.austin.ibm.com] has quit [Quit: Leaving]
18:09	\|-\|	segher [~segher@dslb-084-056-141-255.pools.arcor-ip.net] has quit [Ping timeout: 480 seconds]
18:13	\|-\|	segher [~segher@dslb-084-056-142-175.pools.arcor-ip.net] has joined #xen
18:13	\|-\|	rusty [~rusty@bh02i525f01.au.ibm.com] has joined #xen
18:13	\|-\|	aw [~awilliam@ftcrel4.hp.com] has quit [Quit: Leaving]
18:17	<tessier__>	hmm
18:21	<tessier__>	So PXELINUX loads gets mboot32 then xen then the kernel then the initrd then it says something about booting then the whole machine reboots
18:30	<jcmdev0>	is it panicking on the rootfs?
18:33	\|-\|	segher [~segher@dslb-084-056-142-175.pools.arcor-ip.net] has quit [Ping timeout: 480 seconds]
18:34	\|-\|	segher [~segher@dslb-084-056-160-215.pools.arcor-ip.net] has joined #xen
18:35	<tessier__>	No, it isn't even getting to the rootfs.
18:35	<tessier__>	I found a posting to the syslinux mailing list describing these symptoms and apparently it is a bug
18:36	<tessier__>	I just downloaded the latest syslinux (syslinux-3.20-pre9) and I am compiling it now...
18:37	<tessier__>	root is going to be an AoE device which will be another adventure. :)
18:38	<tessier__>	I'm just trying to get to the point where xen loads, dom0 loads, and it panics on trying to get the rootfs.
18:39	<jcmdev0>	good times.
18:40	<tessier__>	YES!
18:40	<tessier__>	That was it. It now boots xen, dom0 kernel, and panics when it can't find a rootfs. :)
18:40	[~]	tessier__ dances the dance of joy
18:42	<tessier__>	I wish there was an AoE mailing list or something but there does not appear to be
18:45	<jcmdev0>	AoE is voodoo magic as far as I can tell.
18:45	<anticw>	AoE works for me
18:45	<anticw>	im not sure i would call it voodoo magic
18:45	<anticw>	a hack maybe though
18:45	<tessier__>	anticw: How are you using AoE?
18:45	<tessier__>	AoE is dead simple. 8 pages for the whole RFC.
18:46	<anticw>	tessier__: i wrote a small AoE server as an experiment
18:46	<anticw>	tessier__: i used that against the linux client
18:46	<tessier__>	anticw: Are you doing root on AoE? Does it still require a kernel patch?
18:46	<anticw>	tessier__: not root on AoE, but you could
18:46	<anticw>	a kernel patch would be gross and undesirable given how it works, klibc would be much cleaner IMO
18:47	<tessier__>	http://www.coraid.com/support/linux/contrib/mcmullan/aoeroot-2.6.15.diff
18:48	<tessier__>	Then what is that?
18:48	<anticw>	a patch i guess
18:48	<anticw>	the point is it's not needed
18:49	<tessier__>	You should just be able to load the aoe module from an initrd, discover aoe, then switchroot to it right?
18:50	<anticw>	yes
18:50	<anticw>	though i would just compile it in
18:50	<anticw>	modules are a pain the the ass
18:50	<tessier__>	Cool. I hate applying more kernel patches.
18:50	\|-\|	visik7 [~visi@82.54.169.15] has quit [Remote host closed the connection]
18:50	<tessier__>	Actually I did compile AoE in.
18:50	<tessier__>	So I guess I just need to mess with the init script in the initrd now.
18:58	[~]	tessier__ compiles vblade
19:25	\|-\|	ivan [~ikelly@213.228.247.71] has quit [Quit: leaving]
19:30	\|-\|	pvanhoof [~pvanhoof@84.192.226.126] has quit [Remote host closed the connection]
19:34	<um>	anyone know where I can get some info on getting vserver working under a domU ?
19:35	\|-\|	horms [~horms@210.128.90.14] has joined #xen
19:37	\|-\|	zwane_wrk [~zwane@24.85.144.101] has joined #xen
19:39	\|-\|	minemaz [~mine@ppps0271.kitakyushu01.bbiq.jp] has quit [Read error: Connection reset by peer]
19:40	\|-\|	minemaz [~mine@ppps0271.kitakyushu01.bbiq.jp] has joined #xen
19:44	<tessier__>	anticw: The kernel should automatically create any needed AoE devices right? If I want a specific AoE device to be the root fs how do I specify it?
19:49	\|-\|	womble [~mpalmer@220.233.135.106] has joined #xen
20:23	\|-\|	Tv [~tv@38.98.1.19] has quit [Quit: foo]
21:05	\|-\|	minemaz [~mine@ppps0271.kitakyushu01.bbiq.jp] has quit [Read error: Connection reset by peer]
21:05	\|-\|	minemaz [~mine@203.135.192.143] has joined #xen
21:39	\|-\|	jimix [~jimix@ip139.152.45.216.suscom.net] has quit [Read error: Connection reset by peer]
21:41	\|-\|	jimix [~jimix@ip139.152.45.216.suscom.net] has joined #xen
21:49	\|-\|	jforbes_ [~jmforbes@c-71-199-195-167.hsd1.co.comcast.net] has joined #xen
21:49	\|-\|	jforbes [~jmforbes@c-71-199-195-167.hsd1.co.comcast.net] has quit [Read error: Connection reset by peer]
21:59	\|-\|	minemaz [~mine@203.135.192.143] has quit [Read error: Connection reset by peer]
22:00	\|-\|	minemaz [~mine@ppps0271.kitakyushu01.bbiq.jp] has joined #xen
22:59	\|-\|	VS_ChanLog [~stats@ns.theshore.net] has left #xen [Rotating Logs]
22:59	\|-\|	VS_ChanLog [~stats@ns.theshore.net] has joined #xen
23:27	<anticw>	tessier__: the kernel has a state machine, it will discover AoE devices as they appear
23:27	<anticw>	make sure you have aoeutils or whatever to help poke the discover, etc.
23:28	<tessier>	Don't I have to echo 1 into /dev/aoe/discover or something like that?
23:29	<anticw>	something like that
23:29	<anticw>	i forget now
23:29	<tessier>	I hate building initrd's. I spent most of the afternoon messing with it and still don't have a working initrd.
23:30	<tessier>	CentOS/RHEL initrd comes with something called nash to be the shell.
23:31	<anticw>	use busybox or something less crappy
23:31	<anticw>	who cares if it's like 10MB even
23:33	<am88b>	busybox can easilly be less than 1M if you do it right
23:33	<tessier>	I did switch to busybox.
23:33	<tessier>	The kernel complains it can't execute /init and then reboots.
23:33	<tessier>	Not sure why. I rm'd nash and switched the init script to ash.
23:33	<am88b>	but anyway, why would you use initrd anyway? I wouldn't bother at all
23:34	<am88b>	well just make ln -s /bin/busybox /init ?
23:34	<tessier>	am88b: I am PXE booting the diskless xen cpu node and doing root (and everything else) over AoE. That's why. :)
23:34	<tessier>	am88b: /init is the init script which prepares everything else. It is a shell script.
23:35	<am88b>	are you sure kernel is able to execute init that is shellscript?
23:35	<tessier>	Yes. That is how the standard redhat initrd works.
23:35	<am88b>	Ok. I was pretty sure you can't do it (at least I wasn't)
23:36	<tessier>	Maybe I misunderstood how the original initrd was put together.
23:36	<tessier>	Actually, I just double checked. mkinitrd creates an initrd with a shell script for init
23:36	<am88b>	ok
23:40	\|-\|	xai [~pasta@user-0vvdfs3.cable.mindspring.com] has left #xen [Zneet Znatter Zneet]
23:43	\|-\|	tobi [~tobi@zux006-049-116.adsl.green.ch] has joined #xen
---	Log	closed Fri Jul 07 00:00:50 2006