Hi Qumar,
The constraint of having standalone instances is based on fact:
1. Fail-over is restricted to cluster boundary
2. If standalone instances are mixed with cluster instances as part of
same cluster, then fail-over may happen outside cluster boundary,
resulting in session loss.
Thus by defining each of them as separate cluster, it provides correct
view for fail-over scenario.
Having said that, user can play with load-balancer xml and configure it
as they desire.
Hope this helps,
Kshitiz
glassfish_at_javadesktop.org wrote:
> Hi Kshitiz,
>
> Thank you very much for the usefull infoamtion.
> I bet, I can not find these information in docs..Thanks again!
>
> Now the cluster picture is clear to me..the only thing seems odd to me is that why can't I have all my stand alone server in a single cluster!!!
>
> Anway thanks for your help!
>
> Best Regards,
> Qumar
> [Message sent by forum member 'qumar_samba' (qumar_samba)]
>
> http://forums.java.net/jive/thread.jspa?messageID=324469
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: users-help_at_glassfish.dev.java.net
>
>