Hi all
I need ideas for an architecture formed by a Causal Cluster with the main graph and 10K machines in which to consult the graph and that can work automatically if communications with the main graph in the Cluster are ever broken.
There should be periodic updates of these mini graphs if the main graph changes.
It is not necessary for the machines to have a front, all the queries are from the back
Problems:
number of licenses to contract
the 10K machines do not have a large storage or processing capacity
Possible solutions?
install the 10k neo4j and use the Neo4j Streams CDC to keep them updated
install an Elastic on the 10K machines and keep it updated with Neo4j Transaction Handler
in general you can use Neo4j Fabric for federating many different databases and then select in your fabric queries dynamically which ones you want to address:
1)The different machines can be updated with the cluster, so that any change in the cluster is reflected in the 30K machines automatically? like a CDC
2)What is the main difference between this configuration and the default configuration, connecting the different neo4j instances of each machine with the Causal Cluster with kafka + cdc neo4j streams?