Here we go.. again…
I was playing with this procedure, and seems like the Sun Cluster 3.2(SC) has a bug only fixed in the Express edition. Talking to Suraj Verma, he points me this link that explains more in depth the problem, and gaves me the idea to implement a GDS agent and do not use the Proxy_SMF_failover resource type.
Well, i did implement this agent for SC, but did not know anything about GDS implementation… thanks to Suraj (again), he gaves me some directions, and i found this page at (I have used this template).
So, i did create a Rsync agent without use the Proxy_SMF_failover RT, and here you can download the solaris package. For some reason the probe on port 873 was not working, so i did put the Probe_command=”/usr/bin/true” workaround (i will fix that soon).

Here is the command line:

clresource create -g yourresource-rg -t SUNW.gds -p \\
Port_List="873/tcp" -p Start_command="/usr/local/bin/gdsrsync-start" \\
-p Stop_command="/usr/local/bin/gdsrsync-stop" -p Probe_command=\\
"/usr/bin/telnet" -p Resource_dependencies=yourresource-hastorageplus-rs \\

ps.: Remember that you need to register the SUNW.gds RT if you did not use it yet.

Here you can download the gdsrsync-start and the gdsrsync-stop scripts. They are really simple scripts, feel free to enhance it and share it. :)

That’s all.