Re: alpha version of user area tools for 2.6
Moshe Bar <moshe-ay74M1d3r6RWk0Htik3J/[email protected]>
| Newsgroups | gmane.linux.cluster.openmosix.devel |
|---|---|
| Message-ID | <[email protected]> |
Tab
All good proposals, but let's not start doing stuff in Ocaml. It's
enough if the Xen team in Cambridge is fascinated by it, but we are
not going to go that way. :-)
Moshe
On Apr 19, 2006, at 5:43 AM, Vincent Hanquez wrote:
> On Tue, Apr 18, 2006 at 10:15:30AM +0200, David Santo Orcero wrote:
>> The most of the documentation, the scripts and the know-how will
>> stay
>> similar.
>
> everything is *DIFFERENT*. So I don't understand why you want to do
> this.
>
>> I do know that 2.4 clusters will stay a long time; that is why I
>> think
>> that compatibility is important. This will allow developers, system
>> administrators and documentators not to maintain two differents code
>> lines.
>
> That's useless to keep compatibility, for 2 different codebase that
> are going to
> drift away. You'll end up spending time where that's not necessary.
> Plus that give a great opportunity to actually break the link with
> mosix.
>
> If that's really necessary afterward (once everything else is
> working),
> if people want to spend time doing that, fine by me.
> But that's actually just will be something like:
>
> mosctl -> execve("omctl")
>
>> The development will continue slow (I am making this on my little
>> free
>> time); but at a constant rithm. It is expected to have all working
>> (with
>> node registering, cluster management and migration) for summer.
>
> loadbalancing and node registration are the only useful bit so far.
>
> node registration is actually pretty straightforward, receiving a
> message on the socket from X, you just add it to some list of node to
> monitor.
>
> load balancing is a bit more work, that can be split in 2:
> * reading local processes information, to choose process that can be
> migrated and that have good reason to be.
> * exchanging information (cpu, memory, processes) with other node
> to be able to send some process to the correct place.
>
> The link between the two is "just" a function that look like (OCaml
> notation):
> local_info -> remote_info -> (pid, where_to)
>
> This doesn't have to be smart at initial implementation anyway.
>
> As well I suggest writing the load balancer in a high level language.
> that's 1 week work using something like OCaml.
>
> --
> Vincent Hanquez
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting
> language
> that extends applications into web and mobile media. Attend the
> live webcast
> and join the prime developer group breaking into this new coding
> territory!
> http://sel.as-us.falkag.net/sel?
> cmd=lnk&kid=110944&bid=241720&dat=121642
> _______________________________________________
> openMosix-devel mailing list
> openMosix-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/openmosix-devel
>
-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642