Skip to content

k8s/fabric-x-yugabyte.yaml

fabric-x-yugabyte.yaml deploys the Kubernetes sample with the committer storage layer switched from PostgreSQL to YugabyteDB.

Use it when you need to validate Fabric-X plus YugabyteDB on Kubernetes: workloads, services, NodePort exposure, TLS material, and committer-to-YugabyteDB references.

Network Diagram

The diagram below summarizes this inventory's Fabric-X services and how they fit together.

Kubernetes Fabric-X YugabyteDB inventory

Inventory Details

Fabric CA, CA databases, orderer, committer, YugabyteDB, load generator, node exporter, Prometheus, and Grafana use Kubernetes task paths. YugabyteDB master and tablet webserver ports are exposed through fixed NodePorts for inspection.

This inventory deploys these logical services as Kubernetes workloads and services:

  • 5 Fabric CA servers and 5 PostgreSQL databases for Fabric CA state.
  • 4 orderer groups. Each group has 1 router, 1 consenter, 1 assembler, and 1 batcher.
  • 1 committer with validator, verifier, coordinator, sidecar, and query service.
  • 1 YugabyteDB master and 1 YugabyteDB tablet in cluster 1.
  • 1 load generator.
  • Monitoring with node exporter, Prometheus, and Grafana.

Note

You can scale YugabyteDB for stronger performance by adding more master and tablet hosts. See the distributed Fabric-X inventory for a larger topology with replicated YugabyteDB masters and tablets.

flowchart TD
  all --> network
  network --> fabric_cas
  network --> fabric_x
  all --> load_generators
  all --> monitoring
  fabric_cas --> fabric_ca_servers
  fabric_cas --> fabric_ca_dbs["Fabric CA PostgreSQL DBs"]
  fabric_x --> fabric_x_orderers
  fabric_x --> fabric_x_committer
  fabric_x_orderers --> orderer_groups["fabric_x_orderer_1..4"]
  fabric_x_committer --> committer_services["validator, verifier, coordinator, sidecar, query service"]
  fabric_x_committer --> committer_dbs
  committer_dbs --> yugabyte_master["YugabyteDB master"]
  committer_dbs --> yugabyte_tablet["YugabyteDB tablet"]

The validator and query service both use yugabyte_cluster_ref_id: 1, which points them at the YugabyteDB hosts under committer_dbs.

PostgreSQL is still present for Fabric CA state, but the committer database is YugabyteDB. Monitoring omits the PostgreSQL exporter used by PostgreSQL-backed Kubernetes inventories.