No new COP IP addresses are allocated for scaling out and no COP IP addresses are removed for scaling in. Client machines can use the same group of COP entries to connect to the scaled system. In other words, scaling out or scaling in is transparent on client machines that have correctly configured COP entries.
If you did not enable Scale Out/In at the time of deployment, enter the following command to allocate enough COP entries to support the number of nodes on the system:
# tdc-network-copentries --allocate nodecountwhere nodecount is the number of nodes on the system after scaling out. The hosts file on all the TPA nodes is updated.
Type # tdc-network-copentries to display the COP entries in the correct order for the system, as shown in the following example:
COP entries: 10.1.1.4 databasecop1 10.1.1.8 databasecop2 10.1.1.5 databasecop3 10.1.1.9 databasecop4 10.1.1.6 databasecop5 10.1.1.10 databasecop6 10.1.1.7 databasecop7 10.1.1.11 databasecop8
- Assign the COP entries on your client machine.