Cross-Product Interop Matrix
Intent
Test across validators/servers/SDKs continuously to catch ecosystem differences early.
Structure
A cross-product interop matrix is a small harness that runs the same set of artifacts/tests across multiple validators, servers, and SDKs. Structurally it defines targets, execution environments, and a split between smoke and full suites.
- Target list: named validators/servers/SDKs with minimum supported versions
- Execution harness: scripted runs (often containerized) so results are comparable
- Suites: small smoke suite for PRs + larger suite for nightly/pre-release
- Reporting: normalized reports that show pass/fail by target and scenario
- Triage taxonomy: labels/owners that quickly route failures to profiling, terminology, or platform variance
Key Components
Matrix targets
- Choose representative validators/servers/SDKs
- Define minimum supported versions
- Separate required vs optional targets
- Keep the matrix small at first
- Expand as adoption grows
Docker harness
- Provide repeatable environments for each target
- Capture versions in Docker image tags
- Make runs deterministic and scriptable
- Store logs and reports as artifacts
- Allow local developer runs
Smoke vs full suites
- Keep a smoke suite that runs on every PR
- Run full suites nightly or before release
- Track coverage gaps explicitly
- Avoid flakiness by controlling dependencies
- Use failures to prioritize fixes
Triage labels
- Label failures by root cause (terminology, profile, server)
- Track whether failure is known variance or a bug
- Assign owners for each area
- Keep a dashboard of recurring issues
- Use labels to drive release readiness
Behavior
Detect variance early
Run a small subset early and often; run deeper checks on cadence. Treat variance as data, not as noise.
PR smoke run
- Run the smallest set of scenarios that catch common incompatibilities.
- Fail PRs only on agreed critical targets; keep others advisory at first.
- Record failure fingerprints for faster triage.
Nightly/pre-release
- Run full suite across all targets.
- Track trendlines (new failures vs known variance).
- Feed results back into patterns: add guidance where ecosystem behavior differs.
Benefits
- Avoid vendor-specific success
- Build confidence
Trade-offs
- Maintain environments/credentials
References
- Testing Platforms - Testing catalogue
- Inferno - Inferno host