This content has been marked as final. Show 4 replies
Many of our customers manage their MDEX and ITL processes on different machines. Can you provide a little more detail on which data you want to update?
We will deploy one of the reference application "discover-data-pci" [Endeca\ToolsAndFrameworks\3.1.0\reference]
Till now we had MDEX and ITL in the same machine but now we will have both in separate machine and that is why have following questions
1) Shall we need to deploy the sample application in both the machines ,If not then in which I have to deploy either in the one having MDEX or one which have ITL
2) In order to customize the reference application with our requirements we are using baseline_update script which updates the data so we are looking for a way
so that when we run execute baseline_update script it should update the data in both the machines
Data here I am referring to either test_data OR the one which we will crawl in CAS console and pull from there
Please do let me know if any thing above is unclear
The right approach for this is to add MDEX (i.e. Dgraph) component you installed on machine_B to ITL (it will include forge and dgidx optionally dgraph) installed on machine_A and update the index rather than updating the input source data.
You can do this by configuring Dgraph component on machine_A by editing AppConfig.xml and use copyIndexToDgraphServers, applyIndex to update the indexes on machine_B. And your application can point to dgraph on machine_B.
The deployment template has this built into ./config/script/AppConfig.xml. Update the relevant <dgraph id="Dgraph1" /> to host="ITLHost", keep Dgraph2 as the MDEX server. Restart group property determines the restart order (integrate this with your load balancer). Have a read of the a deployment template usage guide - it has all of this in it.