[Confidence04]

Summary

Add Soroban resource-budget benchmarks and CI gate for vault entry points

What was solved

#226 — Profile and Optimize Soroban Resource Budgets for Vault Contract Execution

Add a deterministic Soroban resource-budget measurement harness for vault contract entry points, document per-entry-point budgets with headroom against network limits, optimize hot paths (batched/cursor processing, reduced persistent reads, smaller stored values), add a CI regression gate that fails when a benchmarked entry point exceeds its documented budget, and publish per-entry-point resource numbers for frontend fee estimation.

Addressed:

• Changed: contracts/perf/baseline.json, contracts/perf/thresholds.json, docs/PERFORMANCE.md
• Build a repeatable, deterministic benchmark harness that reports CPU instructions, memory bytes, and ledger entry reads/writes for each vault entry point (deposit, request/execute withdrawal, epoch close, strategy rollover, settlement callback) at representative and worst-case sizes.
• Benchmark worst cases (vault at capacity, full withdrawal queue, epoch with many expired positions), not averages.
• Document a per-entry-point resource budget with explicit headroom against current Soroban limits in docs/.

#227 — Define Vault SLOs, Alerting Thresholds, and On-Call Runbooks

Add vault-specific SLO documentation, Prometheus-style alert rules with warning/critical tiers and evaluation windows, per-alert on-call runbooks, and severity/escalation docs under docs/, derived from existing epoch-ledger and NAV-drift metrics, with environment-configurable thresholds and fault-injection validation guidance.

Addressed:

• Changed: docs/vault-slos.md, docs/vault-alerts.md, docs/vault-runbooks.md, docs/on-call.md
• Document SLOs for epoch close timeliness (expected vs actual close time), redemption queue latency, vault API availability and p95 latency, indexer lag behind the Soroban ledger, and NAV reconciliation drift
• Provide Prometheus-style alert rules for each SLO with warning/critical tiers and explicit short (stall-catching) and long (anti-flap) evaluation windows
• Write one runbook per alert covering symptom, likely causes, verification steps (named endpoints), mitigation, escalation, and how to engage the existing vault kill switch / emergency pause

#228 — Build a Mainnet-State Fork and Chaos-Testing Environment for the Vault Stack

Add a reproducible mainnet-state fork and chaos-testing harness for the vault stack: a snapshot capture script that pins public ledger state at a chosen sequence, a bootstrap entry point that boots the environment and deploys the compiled vault suite against the snapshot, deterministic seeding (including ledger timestamps for epoch boundaries), scripted chaos scenarios injected at the RPC/client boundary (latency/timeouts, submission failures/duplicates, oracle staleness/missing prices, ledger close reordering, partial deposit failures, indexer lag/restart, late/duplicate settlement callbacks) each asserting resulting vault state (NAV, queue, epoch), per-scenario reports archived as CI artifacts, and a self-test proving the harness has teeth.

Addressed:

• Changed: docker-compose.yml
• Provide a documented one-command workflow (docker-compose.yml and/or scripts/ entry point) that captures a ledger snapshot at a chosen ledger sequence, boots the environment locally, and deploys the compiled vault suite plus dependencies against it.
• Separate snapshot capture from environment boot so a snapshot can be pinned and shared.
• Deterministic seeding: same snapshot + same seed yields the same scenario outcome, including ledger timestamps used for epoch boundaries.

#229 — Establish Vault Event Schema Versioning and ABI Backwards-Compatibility Guarantees

Introduce explicit schema versioning for vault contract events and entry points, document a backwards-compatibility policy, derive/check consumer TypeScript types against a checked-in contract spec snapshot, and add CI enforcement that fails when the ABI/event snapshot drifts without a version bump and changelog entry.

Addressed:

• Changed: src/lib/api/vaults.ts, .github/workflows/ci.yml
• Every emitted vault event must carry an explicit schema_version and a stable, documented topic layout; consumers must be able to ignore unknown fields
• Document a compatibility policy: additive-only changes within a major version, explicit deprecation window before removing/semantically changing a field, and a defined breaking-major process using dual-emit during migration
• Entry point signatures (arguments, return types, error codes) governed the same way, with error codes enumerable and documented

Changes

• contracts/perf/baseline.json (modify)
• docs/vault-slos.md (create)
• docs/on-call.md (create)
• .github/workflows/ci.yml (modify)
• src/lib/api/vaults.ts (modify)
• docs/PERFORMANCE.md (modify)
• docs/vault-runbooks.md (create)
• contracts/perf/thresholds.json (modify)
• docs/vault-alerts.md (create)
• docker-compose.yml (modify)

Approach

1. Profile and Optimize Soroban Resource Budgets for Vault Contract Execution #226 — Profile and Optimize Soroban Resource Budgets for Vault Contract Execution (Changed: contracts/perf/baseline.json, contracts/perf/thresholds.json, docs/PERFORMANCE.md)
2. Define Vault SLOs, Alerting Thresholds, and On-Call Runbooks #227 — Define Vault SLOs, Alerting Thresholds, and On-Call Runbooks (Changed: docs/vault-slos.md, docs/vault-alerts.md, docs/vault-runbooks.md, docs/on-call.md)
3. Build a Mainnet-State Fork and Chaos-Testing Environment for the Vault Stack #228 — Build a Mainnet-State Fork and Chaos-Testing Environment for the Vault Stack (Changed: docker-compose.yml)
4. Establish Vault Event Schema Versioning and ABI Backwards-Compatibility Guarantees #229 — Establish Vault Event Schema Versioning and ABI Backwards-Compatibility Guarantees (Changed: src/lib/api/vaults.ts, .github/workflows/ci.yml)

Issues

Closes #226
Closes #227
Closes #228
Closes #229

[drips-wave]

@Confidence04 Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits