Re: A proposal for a General Clustering Framework

[Peter Badovinatz <tabmowzo@yahoo.com>]

--- Alan Robertson <alanr@unix.sh> wrote:
> Alan Robertson wrote:
<snip>
> Sorry.
> 
> You could build either type of implementation on top of the heartbeat API
> easily.
> 
> Actually there are three kinds that come to mind:
> 
> 	A master-node concept, where all events are serialized through this one
> 		machine.
> 
> 	A distributed version with event-order coherency, where all events are seen
> 		in the same order by all machines.

These first two often break down to the same underlying implementation, or the
latter is a distributed agreement protocol of some sort.  Trade-offs here for
scaling, messages, etc.  I've built these kinds of systems.
> 
> 	A distributed version without event-order coherency.

This IS useful.  It is just useful for different things than the first two.  It
is very scalable, very flexible, and so long as it is separate from the ordered
mechanisms so the consumer can tell the message streams apart, you do want them
all present.
> 
> I suppose you could install all types simultaneously, but it sounds like it
> wouldn't be very useful, and would confuse the apps.
> 
> If you design the API properly, you should be able to hide the differences
> between the first two from the application.  The third one has different
> semantics, and shouldn't be made to look like the other two...  [and it's
> probably not that useful anyway...]

The first two should be invisible to the consumers, they get events, it is not
important to them how the ordering is controlled, so long as it works.  The
latter needs to be clearly separated such that consumers can get events about
which they care, but which may not indicate "bad" things, and they know that
this channel is unordered.  You want to be able to pass through large numbers
of events (10,000s) through this channel to support monitoring as well as
recovery.  

To try and order all of these is a scaling nightmare.
> 
> 
> 	-- Alan Robertson
> 	   alanr@unix.sh

Peter

=====
These have been the opinions of:
Peter R. Badovinatz -- (503)578-5530 (TL 775)
wombat@us.ibm.com/tabmowzo@yahoo.com
and in no way should be construed as official opinion of 
IBM, Corp.

__________________________________________________
Do You Yahoo!?
Get personalized email addresses from Yahoo! Mail - only $35 
a year!  http://personal.mail.yahoo.com/

Linux-cluster: generic cluster infrastructure for Linux
Archive:       http://mail.nl.linux.org/linux-cluster/