It is possible to do this via node filters or attributes. We use differentiated fleets of client nodes that interact with server nodes to pull data processed at the server node level and then further process it at a cluster level. We assign a 'ROLE' attribute to each node and then filter those nodes from the Ignite cluster to form the compute projection to Apply() or Broadcast() a request to.
We do this in an auto scaled Kubernetes deployment on AWS. Cheers, Raymond. On Fri, Dec 11, 2020 at 9:37 AM akorensh <[email protected]> wrote: > Hi, > You could use node filters: > https://ignite.apache.org/docs/latest/services/services#node-filter > example of node filter: > > https://github.com/apache/ignite/blob/master/examples/src/main/java/org/apache/ignite/examples/computegrid/ComputeFibonacciContinuationExample.java > > In general, it is not recommended to segregate nodes based on individual > responsibility, as it negates the main advantage of automatic scaling, > failover and backup. > > see here: > > https://ignite.apache.org/docs/latest/distributed-computing/distributed-computing > > Thanks, Alex > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/ > -- <http://www.trimble.com/> Raymond Wilson Solution Architect, Civil Construction Software Systems (CCSS) 11 Birmingham Drive | Christchurch, New Zealand +64-21-2013317 Mobile [email protected] <https://worksos.trimble.com/?utm_source=Trimble&utm_medium=emailsign&utm_campaign=Launch>
