Implement a Metrics Snapshot/Export API

[grantfox-oss]

Implement a Metrics Snapshot/Export API

Labels: hard observability Official Campaign Maybe Rewarded
Component: src/metrics.rs

---

Problem

SyncEngineMetrics (src/metrics.rs) is just a bag of AtomicUsize fields with no way to read them together as a consistent snapshot, no way to reset them (e.g. for periodic reporting windows), and no export format. An embedding wallet or relay node has no way to surface "how many payments are queued right now," "how many conflicts have been detected this week," etc. to a dashboard, a support ticket, or a telemetry pipeline — all of which matter for an org running a "Relay Node Monitoring Dashboard" per the top-level roadmap.

Requirements

• Implement pub fn snapshot(&self) -> MetricsSnapshot, a plain-data struct capturing every counter's current value atomically-enough for reporting purposes (perfect cross-field atomicity isn't required — document that this is an approximate, independently-read snapshot, not a transactionally consistent one).
• Implement a Prometheus text-exposition-format serializer (pub fn to_prometheus_text(&self) -> String) — this is a well-specified, simple text format; look up the spec rather than inventing your own.
• Add whatever additional counters are missing but implied by other issues in this backlog (e.g. queue_evictions from issue Bound OutboundTxQueue Capacity with Priority-Aware Eviction #5, conflicts_detected/disputes_escalated are already present — audit SyncEngineMetrics for gaps against every module in src/ and fill them in).
• Consider a reset/windowed-counting mode (e.g. snapshot_and_reset()) for callers who want periodic deltas rather than lifetime totals, and document when each is appropriate.

Acceptance Criteria

• snapshot() returns accurate values matching the underlying atomics at call time.
• to_prometheus_text() output is valid Prometheus exposition format (verify against the spec, e.g. correct # TYPE/# HELP lines and metric naming conventions).
• Every counter that exists anywhere in this crate (including ones added by other landed issues) is represented in the snapshot — audit for completeness.
• cargo fmt, cargo clippy -D warnings, and cargo test all pass.

Required Tests

• test_snapshot_reflects_current_values
• test_prometheus_output_is_well_formed
• test_snapshot_and_reset_zeroes_counters

Notes

• Independent of all other issues — purely additive to src/metrics.rs.
• Estimated effort: small-medium. Good pickup for a contributor newer to Rust who still wants a genuinely useful, non-trivial task.

[Dev-journals]

Hi. I am a web3 fullstack developer. Kindly assign this issue to me. I have years of Rust development experience with deep expertise in metrics instrumentation, Prometheus integration, atomic operations, and building observability tooling for production-grade blockchain relay nodes and wallet infrastructure. My approach implements pub fn snapshot(&self) -> MetricsSnapshot returning a plain-data struct that captures every AtomicUsize counter's current value using Ordering::Relaxed loads, with documentation clearly stating this is an approximately consistent snapshot (not transactionally atomic across all counters) suitable for reporting and dashboards. I will implement pub fn to_prometheus_text(&self) -> String following the Prometheus text exposition specification, generating valid # TYPE, # HELP headers and correctly formatted metric lines with appropriate naming conventions (e.g., sync_engine_queue_evictions_total, sync_engine_conflicts_detected_total).
I will audit every module in src/ (queue, storage, conflict, sequence, etc.) to identify gaps and add any missing counters such as sequence_reservation_reconciliations, storage_query_latency, or queue_evictions—ensuring all relevant metrics are captured in the snapshot. I will implement snapshot_and_reset() for callers needing periodic windowed deltas (e.g., weekly reporting) while documenting when to use lifetime totals vs. periodic resets based on different monitoring requirements. I will write comprehensive tests covering snapshot accuracy, Prometheus output validity against the spec, and reset behavior correctly zeroes counters after snapshot. I am ready to deliver this observability enhancement within 2-3 hours with full test coverage and documentation for consumers

[grantfox-oss]

🦊 GrantFox — @Dev-journals has been assigned to this issue as part of the Official Campaign | FWC26 campaign!

Next steps:

1. Open a Pull Request referencing this issue (e.g., Closes #22)
2. Your PR will be reviewed by the StellarConduit maintainers

Good luck! Track your progress on GrantFox.