[cgl_discussion] PoC concall minutes, 4/3

Peter Badovinatz tabmowzo at us.ibm.com
Fri Apr 4 12:41:22 PST 2003


Rod Van Meter wrote:
> <snip>
> AEM: discussion going on on cgl_discussion.  Added some new examples
> on SF page to see how it interacts w/ epoll & select.
> 
> Rusty: does AEM meet our (CGL) needs?  Is the plan to decide that,
> then shoot for 2.7 kernel?  Mika says we might use it as a means to
> encourage other implementations.  There are concerns with its
> intrusiveness into the kernel.  Dave says it makes good baseline.
> Mika doesn't want to drop until he sees something better.

I assume the current immature implementation doesn't meet what we
believe are our needs, but the capabilities of what AEM can be when
fully and properly implemented are what we want to match to our needs.

> 
> Mika says after public draft of clustering spec goes out, which calls
> for asynchronous events, might be a good time to take AEM to LKML and
> say, "Hey, this meets that need."  This should happen by the end of
> the month.

Actually, AEM is listed in the General System requirements spec, which
was just made publically available earlier today.  So it's already out
the door. 

My concerns about it are certainly being covered nicely in the ongoing
discussion.  Key (for me) is whether it offers something new and
essential that can't be gained with existing (or committed upcoming 2.6
features).  If so it becomes a requirement, and how to get it to
mainline proceeds on its own.

Peter
--
Peter R. Badovinatz aka 'Wombat' -- IBM Linux Technology Center
preferred: tabmowzo at us.ibm.com / alternate: wombat at us.ibm.com
These are my opinions and absolutely not official opinions of IBM, Corp.



More information about the cgl_discussion mailing list