Consider 3 machines (machine1, machine2, machine3):
1. On machine1 create folder: d:\scimoredb and d:\scimoredb\data ,d:\scimoredb\data1 (can be any drive)
2. Copy to the d:\scimoredb: scimoredb.exe,scimoreagent.exe,Scimore.Data.ScimoreClientNative.[x86/x64].dll.
3. Start commands prompt with admin rights and then execute installing instance/node command:
The command will create 2 distinct instances that will listen on 16250 and 16251 ports.
4. Repeat the 3 for machine2 and machine3. (change machine name in "-net/endpoint...")
5. start 6 scimoredb services on all 3 machines.
Start Scimore manager and connect to "machine1:16250". Open query window and execute following SQL:
After executing, connect to either: machine1:16250,machine1:16251,machine2:16250,...
Also, 3 partition groups mean that the data will be partitioned between 3 partitions and
nodes belonging to the partition will share the same segment of the data.
NOTE: in the current example every insert/delete/update the modifications will be applied to 2 machines, the write sclability will suffer. Hence, reads will benefit nicelly the current cluster setup.