From: <ge...@co...> - 2007-09-03 23:49:46
|
Hi Folks, I've been spending some time playing around with KVM and I am starting to become interested in making a contribution to the project. I'm considering the idea of writing a library interface to libqemu in a KVM context. I am aware of libvirt, but that isn't what I'm thinking here. This lib would be very specific to KVM. I have been poking around in the source a bit and it seems like it would be possible. Before I go further, I thought I would ask a few questions. I would also appreciate any and all opinions on the subject. 1. Is anyone already thinking about , or actually doing this? 2. Does this sound useful? 3. Is the libqemu API stable enough at this point to write code against? 4. Can anyone see any immediate road blocks to doing this? This would probably also include a cross platform GUI interface at some point. This interface would be specific to KVM and the subtle nuances that make it unique. I've been looking for a good excuse to use FLTK on a project anyway ;-) Thanks for your consideration, -G |
From: Dor L. <dor...@qu...> - 2007-09-04 07:38:23
|
Hi Folks,=20 I've been spending some time playing around with KVM and I am starting to become interested in making a contribution to the project. I'm considering the idea of writing a library interface to libqemu in a KVM context. I am aware of libvirt, but that isn't what I'm thinking here. This lib would be very specific to KVM. I have been poking around in the source a bit and it seems like it would be possible. Before I go further, I thought I would ask a few questions. I would also appreciate any and all opinions on the subject.=20 1. Is anyone already thinking about , or actually doing this?=20 That's a real interesting idea. Actually this is the first time I discovered people use libqemu apart from qemu itslef. We didn't thought about this yet. 2. Does this sound useful?=20 It does create some interesting VM-process merges although a regular VM has monitor interface or dbus interface so many actions can be done by tcp to qemu. Actually there is work in progress to add dbus interface to qemu. I might not fit every thing but can help for some cases. 3. Is the libqemu API stable enough at this point to write code against? I believe KVM triggers some changes to it. 4. Can anyone see any immediate road blocks to doing this?=20 Except for exposing the same api, no. This would probably also include a cross platform GUI interface at some point. This interface would be specific to KVM and the subtle nuances that make it unique. I've been looking for a good excuse to use FLTK on a project anyway ;-)=20 Why do you need a gui? What's your application/plan for it? Thanks for your consideration,=20 -G=20 =09 =09 =09 =09 Regards, Dor |
From: <ge...@co...> - 2007-09-04 09:35:38
|
Hi Dor, I appreciate your comments. The GUI interface would be an attempt to give KVM the same kind of features that some of the commercial virtualization products provide. A single GUI interface to manage multiple virtual host systems. The goal would be to create a tightly integrated solution that would allow management of even the most subtle KVM features, options, and functions. The exposure of the host / guest communication facilities would be a good example. It might also be interesting to take a look at providing a guest utility library as well. I'm very excited about the KVM project and the tremendous potential it opens up in virtualization. It's really got the gears in my head turning. Thanks, -G "Dor Laor" <dor...@qu...> 09/04/2007 02:38 AM To <ge...@co...>, <kvm...@li...> cc Subject RE: [kvm-devel] libqemu => libkvm? Hi Folks, I've been spending some time playing around with KVM and I am starting to become interested in making a contribution to the project. I'm considering the idea of writing a library interface to libqemu in a KVM context. I am aware of libvirt, but that isn't what I'm thinking here. This lib would be very specific to KVM. I have been poking around in the source a bit and it seems like it would be possible. Before I go further, I thought I would ask a few questions. I would also appreciate any and all opinions on the subject. 1. Is anyone already thinking about , or actually doing this? That's a real interesting idea. Actually this is the first time I discovered people use libqemu apart from qemu itslef. We didn't thought about this yet. 2. Does this sound useful? It does create some interesting VM-process merges although a regular VM has monitor interface or dbus interface so many actions can be done by tcp to qemu. Actually there is work in progress to add dbus interface to qemu. I might not fit every thing but can help for some cases. 3. Is the libqemu API stable enough at this point to write code against? I believe KVM triggers some changes to it. 4. Can anyone see any immediate road blocks to doing this? Except for exposing the same api, no. This would probably also include a cross platform GUI interface at some point. This interface would be specific to KVM and the subtle nuances that make it unique. I've been looking for a good excuse to use FLTK on a project anyway ;-) Why do you need a gui? What's your application/plan for it? Thanks for your consideration, -G Regards, Dor |
From: Dor L. <dor...@qu...> - 2007-09-04 12:16:07
|
Hi Dor,=20 I appreciate your comments. The GUI interface would be an attempt to give KVM the same kind of features that some of the commercial virtualization products provide. A single GUI interface to manage multiple virtual host systems. The goal would be to create a tightly integrated solution that would allow management of even the most subtle KVM features, options, and functions. The exposure of the host / guest communication facilities would be a good example. It might also be interesting to take a look at providing a guest utility library as well. I'm very excited about the KVM project and the tremendous potential it opens up in virtualization. It's really got the gears in my head turning.=20 Thanks,=20 -G =20 While it's fun to have a new GUI dedicated to KVM, IMHO the virt-manager can be enhanced for everyones need. The community will benefit of one good GUI=20 instead of several parallel efforts. |
From: Daniel V. <vei...@re...> - 2007-09-04 12:42:25
|
On Tue, Sep 04, 2007 at 05:15:58AM -0700, Dor Laor wrote: > > While it's fun to have a new GUI dedicated to KVM, IMHO the > virt-manager can > > be enhanced for everyones need. The community will benefit of one good > GUI > > instead of several parallel efforts. Thanks ! If there is something missing in libvirt or virt-manager, we are definitely open to suggestions, usually take patches, and eager to get more people helping to build and run those projects in a fully open way, yours, Daniel -- Red Hat Virtualization group http://redhat.com/virtualization/ Daniel Veillard | virtualization library http://libvirt.org/ vei...@re... | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/ |
From: Anthony L. <an...@co...> - 2007-09-04 13:55:28
|
On Mon, 2007-09-03 at 18:49 -0500, ge...@co... wrote: > > Hi Folks, > > I've been spending some time playing around with KVM and I am starting > to become interested in making a contribution to the project. I'm > considering the idea of writing a library interface to libqemu in a > KVM context. What do you mean by libqemu? The libqemu in the QEMU project is a library for the CPU translation. It actually doesn't really serve a purpose in the context of KVM. It is not the sort of thing you'd write a GUI tool around. Regards, Anthony Liguori > I am aware of libvirt, but that isn't what I'm thinking here. This > lib would be very specific to KVM. I have been poking around in the > source a bit and it seems like it would be possible. Before I go > further, I thought I would ask a few questions. I would also > appreciate any and all opinions on the subject. > > 1. Is anyone already thinking about , or actually doing this? > 2. Does this sound useful? > 3. Is the libqemu API stable enough at this point to write code > against? > 4. Can anyone see any immediate road blocks to doing this? > > This would probably also include a cross platform GUI interface at > some point. This interface would be specific to KVM and the subtle > nuances that make it unique. I've been looking for a good excuse to > use FLTK on a project anyway ;-) > > Thanks for your consideration, > -G > > > > > > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ kvm-devel mailing list kvm...@li... https://lists.sourceforge.net/lists/listinfo/kvm-devel |
From: <ge...@co...> - 2007-09-04 14:16:51
|
Hi Anthony, What I'm suggesting is a library for KVM that sits between the command line interface and the functions for creating, destroying, starting, stopping, migrating, etc. When I looked at the code it appeared that KVM was actually a modified QEMU. I didn't realize at the time that there was actually a libqemu. The subject line of my post was referring to whether the API would be considered libqemu instead of libkvm. Sorry for the confusion. Thanks, -G Anthony Liguori <an...@co...> 09/04/2007 08:55 AM To ge...@co... cc kvm...@li... Subject Re: [kvm-devel] libqemu => libkvm? On Mon, 2007-09-03 at 18:49 -0500, ge...@co... wrote: > > Hi Folks, > > I've been spending some time playing around with KVM and I am starting > to become interested in making a contribution to the project. I'm > considering the idea of writing a library interface to libqemu in a > KVM context. What do you mean by libqemu? The libqemu in the QEMU project is a library for the CPU translation. It actually doesn't really serve a purpose in the context of KVM. It is not the sort of thing you'd write a GUI tool around. Regards, Anthony Liguori > I am aware of libvirt, but that isn't what I'm thinking here. This > lib would be very specific to KVM. I have been poking around in the > source a bit and it seems like it would be possible. Before I go > further, I thought I would ask a few questions. I would also > appreciate any and all opinions on the subject. > > 1. Is anyone already thinking about , or actually doing this? > 2. Does this sound useful? > 3. Is the libqemu API stable enough at this point to write code > against? > 4. Can anyone see any immediate road blocks to doing this? > > This would probably also include a cross platform GUI interface at > some point. This interface would be specific to KVM and the subtle > nuances that make it unique. I've been looking for a good excuse to > use FLTK on a project anyway ;-) > > Thanks for your consideration, > -G > > > > > > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ kvm-devel mailing list kvm...@li... https://lists.sourceforge.net/lists/listinfo/kvm-devel |