open-vm-tools-devel

[open-vm-tools-devel] Getting involved

[Elliot L. <el...@vm...>]

Hi all,

Now that the code has been out for a little while, and the initial
excitement has died down, I wanted to start a discussion to find out
what everyone's impressions had been so far, and what you wanted out of
the project.

If you are looking to get involved, please don't hesitate to speak up.
Here are some specific things I know can use some work:
. Code review
   . Every check-in inside VMware gets reviewed, and it'd be great to be
able to work with the community on doing reviews of the tools. One
specific area that could especially use review is the kernel modules -
any code in there that you can suggest improvements for?

. Implementing a CLI for the tools. Some of this type of work has
already been done (see
http://chitchat.at.infoseek.co.jp/vmware/ ).

. Improved logging. I've heard it mentioned that someone inside VMware
had possibly started on integrating log4c into the tools, so ask for
more details if you're interested.

. Ports to new OS's
   . FreeBSD
   . Solaris
   . Your favorite OS

FreeBSD & Solaris mainly need kernel module attention. I don't want to
discourage people from going ahead and making wonderful things happen
here, but at the same time, please be aware that there are other efforts
going on here, so speak up if you're interested.

. Making the tools work with qemu, KVM, etc. Anthony Liguori

. Packaging on OS's where the code should already work
   . Debian
   . Fedora
Talk to Denis Leroy, and see
https://bugzilla.redhat.com/show_bug.cgi?id=294341 - what's needed for
the time being is someone to maintain the package in livna.
   . Gentoo
Talk to Elias Probst, and see
http://bugs.gentoo.org/show_bug.cgi?id=192377
   . SuSE
   . Ubuntu
See https://launchpad.net/open-vm-tools

Of course, if you have your own ideas, bring them on!

Best,
-- Elliot

Re: [open-vm-tools-devel] Getting involved

[Justin M. F. <jmf...@li...>]

> . Implementing a CLI for the tools. Some of this type of work has
> already been done (see
> http://chitchat.at.infoseek.co.jp/vmware/ ).
> 
This is indeed interesting and a CLI would be most useful from a
software appliance standpoint.

> . Making the tools work with qemu, KVM, etc. Anthony Liguori
> 
Also interesting work, particularly with KVM.

> . Packaging on OS's where the code should already work
>    . Debian
>    . Fedora

Add rPath to this list.  I am working on this packaging righ tnow.


One thing I am curious on.  The kernel modules... Are there any
intentions to upstream them? I know the virt I/O bits could probably
eventually replace some of these (vmblock/vmnet).  I am guessing if
there was a desire to get the vm* drivers upstream, earlier rather than
later would be best...

Justin

Re: [open-vm-tools-devel] Getting involved

[Adar D. <ad...@vm...>]

> One thing I am curious on.  The kernel modules... Are there any
> intentions to upstream them? I know the virt I/O bits could probably
> eventually replace some of these (vmblock/vmnet).  I am guessing if
> there was a desire to get the vm* drivers upstream, earlier=20
> rather than
> later would be best...

There was a brief discussion[1] about virtio a few weeks back. Note that
vmblock isn't a block device driver (a bit confusing given its name), so =
I
don't think it's necessarily relevant to virtio.

We are definitely interested in upstreaming the kernel modules, but =
it'll
take time and effort. The code must be made conformant to the Linux =
kernel
coding style, and then any design or implementation issues would have to =
be
worked out. There've been rumblings that some (such as vmblock) may need =
to
be reimplemented outright. See Elliot's e-mail where he referenced a =
Fedora
packaging bug; there was some brief discussion about this in that bug =
report.

[1]
http://sourceforge.net/mailarchive/forum.php?thread_name=3D1189726910.598=
2.50.c
amel%40bodhitayantram.eng.vmware.com&forum_name=3Dopen-vm-tools-devel

Re: [open-vm-tools-devel] Getting involved

[Elliot L. <el...@vm...>]

On Mon, 2007-10-01 at 18:42 -0500, Justin M. Forbes wrote: 
> > . Implementing a CLI for the tools. Some of this type of work has
> > already been done (see
> > http://chitchat.at.infoseek.co.jp/vmware/ ).
> > 
> This is indeed interesting and a CLI would be most useful from a
> software appliance standpoint.
> 
> > . Making the tools work with qemu, KVM, etc. Anthony Liguori
> > 
> Also interesting work, particularly with KVM.
> 
> > . Packaging on OS's where the code should already work
> >    . Debian
> >    . Fedora
> 
> Add rPath to this list.  I am working on this packaging righ tnow.
> 
> 
> One thing I am curious on.  The kernel modules... Are there any
> intentions to upstream them? I know the virt I/O bits could probably
> eventually replace some of these (vmblock/vmnet).  I am guessing if
> there was a desire to get the vm* drivers upstream, earlier rather than
> later would be best...

Hey Justin, 

Sorry for the lag... I don't have a good official answer off the top of
my head, but those of us inside of VMware have definitely done some
thinking about upstreaming, and I'll do what I can to keep the issue on
the agenda. Hopefully we'll have something more definitive to share with
the rest of the group at some point.

Any particular reason you think earlier is better than later? (Other
than the benefits that come from having the modules upstream in
general... :)

Best,
-- Elliot

Re: [open-vm-tools-devel] Getting involved

[Charles N W. <ch...@th...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1



Elliot Lee wrote:
> On Mon, 2007-10-01 at 18:42 -0500, Justin M. Forbes wrote: 
>>> . Implementing a CLI for the tools. Some of this type of work has
>>> already been done (see
>>> http://chitchat.at.infoseek.co.jp/vmware/ ).
<snip>

> One thing I am curious on.  The kernel modules... Are there any
>> intentions to upstream them? I know the virt I/O bits could probably
>> eventually replace some of these (vmblock/vmnet).  I am guessing if
>> there was a desire to get the vm* drivers upstream, earlier rather than
>> later would be best...
> 
> Hey Justin, 
> 
> Sorry for the lag... I don't have a good official answer off the top of
> my head, but those of us inside of VMware have definitely done some
> thinking about upstreaming, and I'll do what I can to keep the issue on
> the agenda. Hopefully we'll have something more definitive to share with
> the rest of the group at some point.
> 
> Any particular reason you think earlier is better than later? (Other
> than the benefits that come from having the modules upstream in
> general... :)

I think (in regards to the I/O bits getting upstreamed) that it will
help the other hyper visor projects that are going on.
There have been a number of discussions on the kern virt list recently
regarding lguest and VMI. Very interesting stuff.

 I think adding the vmware tools into the mix could lead to some useful
cross polination.

Rusty Russel, are you on this list? :)

- --
Charles N Wyble ch...@th... (818) 280 7059
http://www.thewybles.com/~charles

President and CEO of Known Element Enterprises and affiliated companies.
http://www.knownelement.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHBBgnkQPZV56XDBMRAvZBAKCW+Egx25U5AObTM8eKAfMJcbvx3ACfdBJv
A4p4A9iJhVZIi7oQDlpGhbo=
=cviN
-----END PGP SIGNATURE-----

Re: [open-vm-tools-devel] Getting involved

[Eric A. <and...@vn...>]

Elliot Lee wrote:
> Hi all,
> 
> Now that the code has been out for a little while, and the initial
> excitement has died down, I wanted to start a discussion to find out
> what everyone's impressions had been so far, and what you wanted out of
> the project.
> 
> If you are looking to get involved, please don't hesitate to speak up.
> Here are some specific things I know can use some work:
> . Code review
>    . Every check-in inside VMware gets reviewed, and it'd be great to be
> able to work with the community on doing reviews of the tools. One
> specific area that could especially use review is the kernel modules -
> any code in there that you can suggest improvements for?
> 
> . Implementing a CLI for the tools. Some of this type of work has
> already been done (see
> http://chitchat.at.infoseek.co.jp/vmware/ ).
> 
> . Improved logging. I've heard it mentioned that someone inside VMware
> had possibly started on integrating log4c into the tools, so ask for
> more details if you're interested.
> 
> . Ports to new OS's
>    . FreeBSD
>    . Solaris
>    . Your favorite OS
> 
> FreeBSD & Solaris mainly need kernel module attention. I don't want to
> discourage people from going ahead and making wonderful things happen
> here, but at the same time, please be aware that there are other efforts
> going on here, so speak up if you're interested.


I'm interested in helping out in the FreeBSD efforts.  Please let me 
know what the current state of things are for the FreeBSD tools.  I 
remember it being said that somebody (intern at VMWare?) had worked on 
porting the kernel modules (vmblock and vmhgfs) to FreeBSD over the 
summer, and either completed them or nearly did.  I would be interested 
in testing them, cleaning up code, and making any necessary changes to 
port to FreeBSD -CURRENT (7.x).


Thanks!
Eric

Re: [open-vm-tools-devel] Getting involved

[Eric A. <and...@fr...>]

Elliot Lee wrote:
> Hi all,
> 
> Now that the code has been out for a little while, and the initial
> excitement has died down, I wanted to start a discussion to find out
> what everyone's impressions had been so far, and what you wanted out of
> the project.
> 
> If you are looking to get involved, please don't hesitate to speak up.
> Here are some specific things I know can use some work:
> . Code review
>    . Every check-in inside VMware gets reviewed, and it'd be great to be
> able to work with the community on doing reviews of the tools. One
> specific area that could especially use review is the kernel modules -
> any code in there that you can suggest improvements for?
> 
> . Implementing a CLI for the tools. Some of this type of work has
> already been done (see
> http://chitchat.at.infoseek.co.jp/vmware/ ).
> 
> . Improved logging. I've heard it mentioned that someone inside VMware
> had possibly started on integrating log4c into the tools, so ask for
> more details if you're interested.
> 
> . Ports to new OS's
>    . FreeBSD
>    . Solaris
>    . Your favorite OS
> 
> FreeBSD & Solaris mainly need kernel module attention. I don't want to
> discourage people from going ahead and making wonderful things happen
> here, but at the same time, please be aware that there are other efforts
> going on here, so speak up if you're interested.


I'm interested in helping out in the FreeBSD efforts.  Please let me
know what the current state of things are for the FreeBSD tools.  I
remember it being said that somebody (intern at VMWare?) had worked on
porting the kernel modules (vmblock and vmhgfs) to FreeBSD over the
summer, and either completed them or nearly did.  I would be interested
in testing them, cleaning up code, and making any necessary changes to
port to FreeBSD -CURRENT (7.x).


Thanks!
Eric