|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Xen-users] LVM and VMM
Hi, I am trying to create a DomU under SLES10.1 that uses LVM. I created 3 lv's /dev/vg1//lv-024swap /dev/vg1//lv-024boot /dev/vg1//lv-024root In Virtual Machine Manager (VMM) I try to assign these to a new machine. The best I can get is using each as a disk, not a partition. When I manually edit the config file and put the lines as disk=[ 'phy:vg1/lv-024swap,hda1,w', 'phy:vg1/lv-024boot,hdb1,w', 'phy:vg1/lv-024root,hdc1,w', 'phy:/dev/sr0,xvdd,r', ] What is supposed to be in my bootloader and bootargs lines? Any help, or pointing me to a document to read would be appreciated. Thomas King IT Specialist Information Technology Services Americas (ITSA) IBM Global Services Office: 514-964-2067 Tie line 314-2067 e-mail: tking@xxxxxxxxxx
Send Xen-users mailing list submissions to xen-users@xxxxxxxxxxxxxxxxxxx To subscribe or unsubscribe via the World Wide Web, visit http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users or, via email, send a message with subject or body 'help' to xen-users-request@xxxxxxxxxxxxxxxxxxx You can reach the person managing the list at xen-users-owner@xxxxxxxxxxxxxxxxxxx When replying, please edit your Subject line so it is more specific than "Re: Contents of Xen-users digest..." Today's Topics: 1. Re: RE: Does it legal to analysize XEN source code and write a book about it (Tao Shen) 2. Re: RE: Does it legal to analysize XEN source code and write a book about it (Tim Post) 3. Re: Boot Problem with XEN 3.1. (Paradox) 4. Re: RE: Does it legal to analysize XEN source code and write a book about it (Tao Shen) 5. Re: RE: Does it legal to analysize XEN source code and write a book about it (Tim Post) ---------------------------------------------------------------------- Message: 1 Date: Mon, 17 Sep 2007 06:11:14 -0600 From: Tao Shen <taoshen1983@xxxxxxxxx> Subject: Re: [Xen-users] RE: Does it legal to analysize XEN source code and write a book about it To: echo@xxxxxxxxxxxx Cc: xen-users@xxxxxxxxxxxxxxxxxxx, Simon Crosby <simon@xxxxxxxxxxxxx> Message-ID: <46EE6EE2.6000109@xxxxxxxxx> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Tim Post wrote: > >> 3. In other words, you have a non-GPL'd controller application wrapped >> around an opensourced Xen hypervisor running on a GPL'd but custom Linux >> distribution made by XenSource. It's crazy how many people on this list >> say "oh Xen is GPL, but XenEnterprise is not" As Tim Post almost made an >> impression that XenEnterprise is a complete beast compared to Xen. >> > > Xen is free software. XenExpress, XenEnt, XenServer are not free > software. At what point did I call it a beast? > Tim, I missed a word. "a completely different beast"...and it's my interpretation of what you said about XenSource maintaining two different source trees and etc. Tim, you also are too strict on the following three terms, free, GPL, open source as if you are an IP lawyer. Of course Xen is free, and XenEnterprise is not. The point I am making is that XenEnterprise is 85% based on the core Xen which is free, GPL'd, and source made available by XenSource, and is marketed as if it's a different beast. > > >> XenEnterprise's non-GPL portion is much smaller compared to the GPL'd >> portion. >> > > Define smaller? Lines of code or cost of development? :) Both must be > considered. Just looking at the screenshots of XenEnt (and the rest) I > can tell you (off the top of my head) your looking at development costs > in the millions. > My definition of smaller is based on perceived value and perceived function. The core free Xen Hypervisor has more perceived value and function than the XenEnterprise the wrapping GUI application. To most people that is. To prove my point, if the Xen Hypervisor is designed correctly, everything you can do in XenEnterprise in a GUI environment, you should be able to do in a command line environment. If you look at the development costs in the millions, it's actually cheap. Development cost really depends on who's developing. People in the US get paid 5 times as much as people in India for example. In the US, 1Mil gives you about 20 programmers for a year on average, maybe 10 good ones for a year. I am sure Xen the hypervisor which is free, had way more programmer-years than the non-GPL'd GUI. And some of the functions that's in the Xen Hypervisor is not enabled in the XenExpress as if XenExpress arbitrarily limit functionality of the core Xen Hypervisor to promote higher priced Server/Enterprise versions even though the Xen Hypervisor hasn't changed. > >> 4. The issues with closed source Windows Paravirtualization is this: >> whether or not it's merged to the Xen Server/Enterprise source tree(the >> controller and UI parts, which is not GPL'd) or it's interfaced to the >> Xen hypervisor source tree(which is GPL'd). From a practical point of >> view, I don't see how a paravirtualized driver can be non-dependent on >> the GPL'd Xen Hypervisor. >> > > Someone other than me has to explain to you how hypercalls and program > API's work. > Yes, that's the question you and I both need answers to. > >> If it's indeed interfaced into the >> hypervisor, then the windows paravirtualization driver should be open >> sourced, GPL'd. >> > > That's up to XenSource. They are under no obligation to make those > drivers free. GPL software is not 'open source', every time you say that > I have to ask you to stop. > Ok, Timster. You are correct on that "They are under no obligations to make it free" My statement uses the words "should be" as merely suggestive that perhaps a open standard, API for the windows paravirtualization driver is of more utility and beneficial to both Microsoft, XenSource, and the end user. Of course XenSource doesn't see it that way, and they see it as another way to make money. Legal for sure. Alienating potential customers for sure. Actually I see Xen having a common API for paravirtualization drivers. Then whatever OS you install in a Xen DomU, the OS writer then becomes responsible for writing the Xen DomU paravirtualization driver. For example, Microsoft can be making a Windows XP-Xen Edition with the Xen drivers...etc. But of course if you are XenSource, you would like to have a monopoly of the paravirtualization driver development and charge people for it. Ok Tim, I got to get 1 hour of sleep before I have to wake up again :) We shall continue the discussion in another time. ------------------------------ Message: 2 Date: Wed, 09 Jan 2008 20:51:15 +0800 From: Tim Post <echo@xxxxxxxxxxxx> Subject: Re: [Xen-users] RE: Does it legal to analysize XEN source code and write a book about it To: Nico Kadel-Garcia <nkadel@xxxxxxxxx> Cc: Tao Shen <taoshen1983@xxxxxxxxx>, xen-users@xxxxxxxxxxxxxxxxxxx, Simon Crosby <simon@xxxxxxxxxxxxx> Message-ID: <1199883075.11788.80.camel@xxxxxxxxxxxxxxxxxxxxx> Content-Type: text/plain On Mon, 2007-09-17 at 12:50 +0100, Nico Kadel-Garcia wrote: > Tao Shen wrote: > > > > > > 7. Some of the examples in the open source industry right now > > include: using PostgreSQL based code which is GPL'd, add a non-GPL'd > > replication suite to it, and call it enterpriseDB. Using PostgreSQL > > based code, tweak some variables, add some non-GPL'd code (interfaced > > to the GPL'd one) to do distributed join and call it "bizgres" and > > "greenplum". MySQL's Enterprise vs Community editions....the > > examples are all other the place. All of them push the GPL boundary > > but don't violate it. And what I call the "wrapper GPL" type > > products, and "dual licensing". No, it's all perfect legal. From an > > ethics perspective...it's arguable. I think you have to look at things in a relative way. Relative to most other commercial software companies, XenSource is a saint. I'm running GNU/XenLinux on my desktop, if I had this developed myself, I would have spent millions on it. XenSource pushed Xen, hard. XenSource is what ensured Xen held up to critics. XenSource ensured that Xen got its market share that it enjoys today. This puts money in my pocket in a few ways : 1 - I'm paid to build xen stuff for web hosts 2 - I'm paid to design networks that rely heavily on Xen 3 - I'm able to give non profits cutting edge technology for free (I don't charge them) 4 - I learned A LOT studying Xen, reading xen-devel and the Xen academic papers. I'm a much better programmer than I was after studying Xen. 5 - Microkernels are now being looked at in a new light, because of Xen (broadly). This means, hardware makers are finally waking up to the fact that they _CAN_ reach beyond x86. This means new technology and new opportunities. What does Microsoft give you? What does Cisco give you? What does NetGear give you? First, Xen was made. Then XenSource made some neat programs to go with it that cost money. In order to imply something sinister, you would first have to demonstrate that XenSource knew how successful Xen (as a whole) would be. I don't think that's possible, because I don't think that they realized it until it smacked them. Please speak in proper time and context. As Nico said, look at RHEL. Look at the Novell-Microsoft deal, look at so many other things in the news as current events, then go enjoy your multi million dollar free hypervisor and command line tools ;) If there is a time for paranoia, this is not that time. Kindly, --Tim ------------------------------ Message: 3 Date: Mon, 17 Sep 2007 14:26:03 +0200 From: Paradox <paradox@xxxxxxxxxxx> Subject: [Xen-users] Re: Boot Problem with XEN 3.1. To: xen-users@xxxxxxxxxxxxxxxxxxx Message-ID: <46EE725B.4040006@xxxxxxxxxxx> Content-Type: text/plain; charset=us-ascii; format=flowed TNX Tim! I try to create a initrd like this command: mkinitrd -v -f --with=aacraid --with=sd_mod --with=scsi_mod -o initrd-2.6.18-xen.img -m 2.6.18-xen Now seems to be XEN start but after memory message and little count down I received another error about "/etc/fstab file not found" and "no /sbin/init found on rootdev (or not mouted!)". I can't past precise error because xen reboot system...argh...I think that initrd is bad builded!! You have any idea? initrd bad sure....How I can create one correctly? You think that I can use a prebuilded kernel? TNX again for you help! Best Regards Tim Post wrote: >>On Sat, 2007-09-15 at 23:12 +0200, Paradox wrote: >> >>Do you need an initrd? It looks like your kernel is not able to access >>whatever type of disk /dev/hda5 is. >> >>I use the following modules on my machine when making the initrd, >>booting to IDE drives as well: >> >>ide-cd >>ide-disk >>ide-generic >>libata >>ext3 >>jfs >> >>This lets me access all of the partitions that I need to continue >>booting, at which point I pick up device mapper modules and LVM. Likely, >>you don't need jfs. > > > >>--Tim > ------------------------------ Message: 4 Date: Mon, 17 Sep 2007 06:41:21 -0600 From: Tao Shen <taoshen1983@xxxxxxxxx> Subject: Re: [Xen-users] RE: Does it legal to analysize XEN source code and write a book about it To: echo@xxxxxxxxxxxx Cc: xen-users@xxxxxxxxxxxxxxxxxxx Message-ID: <46EE75F1.9020408@xxxxxxxxx> Content-Type: text/plain; charset=ISO-8859-1; format=flowed I see that great values exist in Xen and virtualization. That's why I am here, on the list. From a practical perspective, I don't see that XenSource's triple segregation of their product into Express, Server, Enterprise makes any sense from a revenue maximization perspective. For serious users who depend on Xen(Amazon EC2) they will be paying the support contracts but still cheaper than VMware Infrastructure. Software license isn't even considered cost for them. For semi-serious business users, currently Xen is on par in cost compared to VMware. For enthusiasts, you are stuck to Xen + Virt-manger on a stock Linux distro. Quite frankly, Xen Express doesn't cut it. So in the end, the only competitive advantage for Xen branded product is centered on Xen Enterprise for the VPS hosts. XenSource is severely decapitating their potential revenue stream. gosh i just lost another 30 minutes....got to go...30 minute naps will hurt for the day :) later Tim Post wrote: > On Mon, 2007-09-17 at 12:50 +0100, Nico Kadel-Garcia wrote: > >> Tao Shen wrote: >> >>> 7. Some of the examples in the open source industry right now >>> include: using PostgreSQL based code which is GPL'd, add a non-GPL'd >>> replication suite to it, and call it enterpriseDB. Using PostgreSQL >>> based code, tweak some variables, add some non-GPL'd code (interfaced >>> to the GPL'd one) to do distributed join and call it "bizgres" and >>> "greenplum". MySQL's Enterprise vs Community editions....the >>> examples are all other the place. All of them push the GPL boundary >>> but don't violate it. And what I call the "wrapper GPL" type >>> products, and "dual licensing". No, it's all perfect legal. From an >>> ethics perspective...it's arguable. >>> > > I think you have to look at things in a relative way. Relative to most > other commercial software companies, XenSource is a saint. I'm running > GNU/XenLinux on my desktop, if I had this developed myself, I would have > spent millions on it. XenSource pushed Xen, hard. XenSource is what > ensured Xen held up to critics. XenSource ensured that Xen got its > market share that it enjoys today. > > This puts money in my pocket in a few ways : > > 1 - I'm paid to build xen stuff for web hosts > 2 - I'm paid to design networks that rely heavily on Xen > 3 - I'm able to give non profits cutting edge technology for free (I > don't charge them) > 4 - I learned A LOT studying Xen, reading xen-devel and the Xen academic > papers. I'm a much better programmer than I was after studying Xen. > 5 - Microkernels are now being looked at in a new light, because of Xen > (broadly). This means, hardware makers are finally waking up to the fact > that they _CAN_ reach beyond x86. This means new technology and new > opportunities. > > What does Microsoft give you? What does Cisco give you? What does > NetGear give you? > > First, Xen was made. Then XenSource made some neat programs to go with > it that cost money. In order to imply something sinister, you would > first have to demonstrate that XenSource knew how successful Xen (as a > whole) would be. I don't think that's possible, because I don't think > that they realized it until it smacked them. Please speak in proper time > and context. > > As Nico said, look at RHEL. Look at the Novell-Microsoft deal, look at > so many other things in the news as current events, then go enjoy your > multi million dollar free hypervisor and command line tools ;) > > If there is a time for paranoia, this is not that time. > > Kindly, > --Tim > > > > ------------------------------ Message: 5 Date: Wed, 09 Jan 2008 21:14:04 +0800 From: Tim Post <echo@xxxxxxxxxxxx> Subject: Re: [Xen-users] RE: Does it legal to analysize XEN source code and write a book about it To: Tao Shen <taoshen1983@xxxxxxxxx> Cc: xen-users@xxxxxxxxxxxxxxxxxxx, Simon Crosby <simon@xxxxxxxxxxxxx> Message-ID: <1199884444.11788.102.camel@xxxxxxxxxxxxxxxxxxxxx> Content-Type: text/plain On Mon, 2007-09-17 at 06:11 -0600, Tao Shen wrote: > Tim Post wrote: > > > >> 3. In other words, you have a non-GPL'd controller application wrapped > >> around an opensourced Xen hypervisor running on a GPL'd but custom Linux > >> distribution made by XenSource. It's crazy how many people on this list > >> say "oh Xen is GPL, but XenEnterprise is not" As Tim Post almost made an > >> impression that XenEnterprise is a complete beast compared to Xen. > >> > > > > Xen is free software. XenExpress, XenEnt, XenServer are not free > > software. At what point did I call it a beast? > > > Tim, I missed a word. "a completely different beast"...and it's my > interpretation of what you said about XenSource maintaining two > different source trees and etc. Ah, no worries. It is a different beast, its development is headed in a _completely_ different direction. Did you read the press releases? > Tim, you also are too strict on the following three terms, free, GPL, > open source as if you are an IP lawyer. Egads! I am NOT a lawyer. Intellectual property is a mirage, show me one person responsible for shaping their own intellect. I am a free software freak, I freely admit it. I was there when it started and I helped to shape it :) > Of course Xen is free, and > XenEnterprise is not. The point I am making is that XenEnterprise is > 85% based on the core Xen which is free, GPL'd, and source made > available by XenSource, and is marketed as if it's a different beast. Most of us see Xen in pieces Piece 1 - The Hypervisor Piece 2 - The controls Piece 3 - The host OS (usually, GNU/Linux) Whenever you talk about XenEnt, I'm only going to consider piece2. XenEnt is 100% allowed to charge you for a nice re-packaging of GNU/Linux. Stallman himself encourages that, so long as you get the four freedoms with your OS. So, their adaptation of GNU/Linux is , if anything, a gift. Anyone can use (most of) their efforts. The Hypervisor is slightly modified from what I understand. Source is available including those changes, no big deal. The controls cost them millions to make, _of_course_ they're going to charge for them. Some people really like those programs. I am _HAPPY_ to see XenSource profitable, solvent and solid. It guarantees a future for my free software which I base my living from :) Some people want a 'cadillac' control system and they are more than happy to pay for it. I want Xen to get the biggest market share that it can (free or not free) because Xen has fed me, my wife and my kid for a couple of years now :) > My definition of smaller is based on perceived value and perceived > function. The core free Xen Hypervisor has more perceived value and > function than the XenEnterprise the wrapping GUI application. I happen to agree with you, however, many who want something that 'just works' will say "hyper WHAT?", they just want point and click utopia, XenSource delivers it. Would you rather those people went to VMWare or Microsoft who give nothing at all to free software? > To most > people that is. To prove my point, if the Xen Hypervisor is designed > correctly, everything you can do in XenEnterprise in a GUI environment, > you should be able to do in a command line environment. Xen (GPL) is building blocks. Leggos. You snap the stuff together how you want and make your own controls. That's why it has such a big following amongst more experienced system integrators. System integrators are often programmers _and_ administrators, we like Xen, it lets us do whatever the hell we want :) > If you look at the development costs in the millions, it's actually > cheap. I did not say how many millions, Simon might :) > Development cost really depends on who's developing. People in > the US get paid 5 times as much as people in India for example. In the > US, 1Mil gives you about 20 programmers for a year on average, maybe 10 > good ones for a year. Err, no. Most _GOOD_ programmers are about $80 hourly. 50 hours a week, 52 weeks a year. Do the math ;) 1 Mil gets your 4 programmers and a slightly lesser paid project manager to track and oversee productivity. > I am sure Xen the hypervisor which is free, had > way more programmer-years than the non-GPL'd GUI. And some of the > functions that's in the Xen Hypervisor is not enabled in the XenExpress > as if XenExpress arbitrarily limit functionality of the core Xen > Hypervisor to promote higher priced Server/Enterprise versions even > though the Xen Hypervisor hasn't changed. Actually, I believe, the HV did change. Maybe someone can post a diff from the source packages included with XenExp or XenEnt? > > > >> 4. The issues with closed source Windows Paravirtualization is this: > >> whether or not it's merged to the Xen Server/Enterprise source tree(the > >> controller and UI parts, which is not GPL'd) or it's interfaced to the > >> Xen hypervisor source tree(which is GPL'd). "Interfaced" is where your losing it. You need to better understand hypercalls and API's. Xen has 'hooks' for any program that speaks its language to talk to it. That means, those programs use a method, not HV code to do their work. > Ok, Timster. You are correct on that "They are under no obligations to > make it free" > My statement uses the words "should be" as merely suggestive that > perhaps a open standard, API for the windows paravirtualization driver > is of more utility and beneficial to both Microsoft, XenSource, and the > end user. I agree, but we're coming back to, they have given you a mile and you continue to ask for more. They _NEED_ to make money too. Windows is not free software, nobody in their right mind expects free drivers from Microsoft, why would you expect them from XenSource? > Of course XenSource doesn't see it that way, and they see it > as another way to make money. Legal for sure. One of the _FEW_ ways that they make money, you mean. XenSource had to get very creative to make the offerings that they have. If they give up any more, they'll find themselves in the hosting business as a last resort. That would be very unfortunate. > Alienating potential > customers for sure. Well, I'm not a good one to guess on that. I am a FSF fundie, I have not used software that was not GPL since X-Windows was stable. Prior to that, I got e-mail with PINE. I can not stand Microsoft products, I have been annoyed with them ever since they refused to give me the source code to EDLIN. Man I hated that editor and I was STUCK with it because there was no alternate for DOS at the time. Once I got something that could build and use emacs, I never went back. I've also never had to deal with a single piece of spyware :) > Ok Tim, I got to get 1 hour of sleep before I have to wake up again :) > We shall continue the discussion in another time. If you stop saying words like 'trickery', I'm happy to discuss it. Every time you say those words I have to ask you to stop. Kindly, --Tim > > ------------------------------ _______________________________________________ Xen-users mailing list Xen-users@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-users End of Xen-users Digest, Vol 31, Issue 57 ***************************************** _______________________________________________ Xen-users mailing list Xen-users@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-users
|
 |
Lists.xenproject.org is hosted with RackSpace, monitoring our |