We have started testing and are getting ready to implement TCT in other environments in the next few months. The testing we are doing is in our sandbox environment so we can easily tear it down and reset it back up again. We are using a TS7700 Grid as an Object Store. With Two clusters in our primary site and one in our DR site. We have the policy setup that both the Elk River(primary) clusters copy to the DR site. But neither primary cluster is supposed to copy to the other. Nor does the DR cluster copy to the primary.
Our Mkcld command was like this
mkcloudserver -type TS7700 -primary7700IPs 10.200.192.xyz,10.200.194.xyz -secondary7700IPs 10.200.192.xyz,10.200.194.xyz -nossl outechsm
Even though it's only one of the two clusters entered as both primary and secondary it appears we are using the forking mechanism and all objects are on all three clusters. So my impression is that because of entering secondary IPs even though there the same as the primary they are grid links and would force a copy to the other primary library. Can you confirm first of all if my thoughts are correct and if this would be the correct config to avoid forking.
mkcloudserver -type TS7700 -primary7700IPs 10.200.192.xyz,10.200.194.xyz,10.200.192.xyy,10.200.194.xyy -nossl outechsm
Let me know if you need any other information.
------------------------------
John Benik
------------------------------