Node deployment inventory with per-node client identification
Client distribution policy with target ratios and rebalancing
procedures
Live monitoring dashboards showing client distribution metrics
Historical monitoring data showing distribution over the audit
period
Orchestration configs (K8s, Ansible, Terraform) defining
distribution targets
Success Criteria
No single client represents more than 2/3 of deployment for any
protocol with multiple production-ready clients
Deployment inventory is accurate (confirmed by observation or system
evidence)
Operator monitors client distribution and has procedures to maintain
balance
Failure Criteria
Any of these circumstances is likely to mean the Requirement is not
met.
Single client exceeds 2/3 of deployment for a protocol with multiple
production-ready clients
Operator does not monitor client distribution and has no rebalancing
mechanism
Deployment inventory is inaccurate, preventing reliable
calculation
Justifiable Exceptions
Where multiple clients are available it is important to manage a
balanced set of clients, and any failure to meet this requirement needs
to be explicitly justified
Common Failure Patterns
Threshold met at initial deployment but distribution drifts over
time
Counting nodes rather than validators (or vice versa),
misrepresenting stake-weighted distribution;
The threshold met is globally but not within individual
protocols;
Temporary imbalances during migrations are not documented;
Ratios are calculated using plans but the actual running
configuration is different.