Re: Contributing

On Thu, 2002-10-10 at 11:05, Lars Marowsky-Bree wrote:
> On 2002-10-10T11:55:30,
>    Dhammika Pathirana <[email protected]> said:
> 
> > I'm looking to get some experience in cluster programming and would like 
> > to get involved. Though I'm quite conversant in linux programming I 
> > don't have any previous experience in clustering. I'd like to get some 
> > advice on possible starting points and possible individual projects.
> 
> Hi Dhammika, this is good news. (Able people willing to help always are!)
> 
> As you might have read, I personally don't consider Linux FailSafe the best
> place to apply much effort right now.
Yeah, FailSafe tried doing everything...even tried to create something
like a clustered filesystem(the cdb).

But personally, I would like that Linux become distributed, so that
low-level details like node memberships, synchronization primitives,
ipc-communication(i.e. distributed shared memory) are actually part of
the OS, and the end-user(programmer) will have a singular logical view
of the OS. This won't only benefit HA clusters, but HPCs or
load-balanced servers.

Then you could build additional layers on top that provide group
communication, atomic transactions with recovery semantics, clustered
filesystems, etc.

That's my biased opinion anyways, since I've worked with distributed OSs
like Amoeba.