users@glassfish.java.net

Re: DAS machine is single point of failure for entire clustered instances?

From: <glassfish_at_javadesktop.org>
Date: Fri, 06 Jun 2008 11:38:26 PDT

You are missing something substantial ;).

It is correct that DAS is the overseer of the "central repository" and that if it were to be brought
down, the administration/configuration changes would be stalled.

But that does not mean DAS has to be running for your clusters (and instances that form it) to
function and serve your applications. Once you reach a [i]steady state [/i] with your domain and
associated clusters, DAS can be effectively down. This is a significant point because loss of DAS
implies loss of administration, not loss of [i] service [/i] which is important to your customers (and
ultimately to you :) ).

What documentation gave you an impression that loss of DAS = loss of clusters?

Regards,
Kedar
[Message sent by forum member 'km' (km)]

http://forums.java.net/jive/thread.jspa?messageID=278862