distributed/fabric-x.yaml

fabric-x.yaml is a multi-machine SSH inventory for performance-oriented Fabric-X evaluation.

This inventory is not ready to run as-is. Replace the host_machine_* placeholders with real machines and confirm SSH access before using it.

Network Diagram

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

Inventory Details

Ansible reaches the target machines over SSH. The Fabric-X services, YugabyteDB, load generators, and monitoring components run as containers on remote hosts.

The environment file distributed/group_vars/all/env.yaml defines the remote connection defaults, placeholder machines, and deployment directories.

This inventory describes a larger container-based Fabric-X deployment:

• No Fabric CA services. Crypto material is generated on the control node with cryptogen.
• 4 orderer groups. Each group has 1 router, 1 consenter, 1 assembler, and 2 batchers.
• 7 validators, 7 verifiers, 1 coordinator, 1 sidecar, and 1 query service.
• 3 YugabyteDB masters and 7 YugabyteDB tablets.
• 2 load generators.
• Monitoring with 16 node exporters, Prometheus, and Grafana.

flowchart TD
  all --> network
  network --> fabric_x
  all --> load_generators
  all --> monitoring
  fabric_x --> fabric_x_orderers
  fabric_x --> fabric_x_committers
  fabric_x_committers --> fabric_x_committer
  fabric_x_orderers --> fabric_x_orderer_1
  fabric_x_orderers --> fabric_x_orderer_2
  fabric_x_orderers --> fabric_x_orderer_3
  fabric_x_orderers --> fabric_x_orderer_4
  fabric_x_committer --> committer_dbs
  committer_dbs --> yb_masters
  committer_dbs --> yb_tablets

This is a performance-oriented reference topology, not a small development sample. It scales validators, verifiers, batchers, load generators, and YugabyteDB tablets across 16 remote machine placeholders.

Fabric CA is intentionally omitted so large performance runs do not spend time starting CA services or enrolling identities. TLS and mTLS still use centrally generated cryptogen material.