Zebra 4.1.0 Release – Zcash Foundation

The Zcash Foundation is pleased to announce Zebra 4.1.0, a release focused on expanding observability with comprehensive Prometheus metrics and Grafana dashboards, while strengthening mempool compliance with zcashd. This release contains no breaking changes for node operators.

What’s New in 4.1.0

Observability: See Inside Your Node

Zebra 4.1.0 significantly expands monitoring capabilities to give operators deep visibility into node health, performance, and bottlenecks. We want to bring Zebra’s observability on par with other modern node client implementations, so we identified critical gaps and filled them.

Value Pool, RPC, and Peer Metrics

We are now exposing new Prometheus metrics across three key areas:

Value Pool Monitoring

• Pool balances: transparent, sprout, sapling, orchard, and deferred
• Total chain supply

RPC Performance Metrics

• Request counts by method and status
• Latency histograms for per-method performance analysis
• Active request gauge to track concurrent load
• Error tracking by method and error code

To help operators to detect RPC bottlenecks and client issues.

Peer Health Metrics

• Handshake latency histograms to identify slow peer connections
• Failure tracking by reason (timeout, invalid signature, protocol mismatch, etc.)

These are critical for understanding network health. 

Performance Histograms 

We added duration metrics to three critical performance areas:

Sync Pipeline Latency

• sync.stage.duration_seconds: Track time spent in obtain_tips and extend_tips stages
• sync.block.download.duration_seconds: Per-block download latency
• sync.block.verify.duration_seconds: Per-block verification time

To identify the cause of delays; network weather, block verification, or database operations.

Consensus Batch Verification

• zebra.consensus.batch.duration_seconds{verifier,result}: Detailed duration histograms, covering ed25519, redjubjub, redpallas, halo2, and groth16_sapling verifiers

To enable analysis of proof verification performance, post checkpoint synchronization.

RocksDB Database Metrics

• Total disk and live data sizes (updated every 30 seconds)
• memtable memory usage
• Breakdown of disk and memory usage per column family

To understand database growth and memory pressure during sync.

RocksDB I/O and Sync Distance 

Finally, we focused on observability in the state layer and chain synchronization progress

Sync Distance Metrics

• sync.estimated_distance_to_tip: How many blocks behind the network tip
• sync.estimated_network_tip_height: Estimated current network height

These metrics are critical for detecting stalled nodes and monitoring catch-up progress.

RocksDB I/O Performance

• zebra.state.rocksdb.batch_commit.duration_seconds: Write latency histograms

This enables detection of database bottlenecks during heavy sync load.

We are also now measuring bytes pending compaction as well as the active compaction status.

Documentation

For detailed documentation on configuration, deployment, and advanced features, visit the Zebra Book.

Get Involved

Zebra is an open-source project and we welcome contributions from the community:

Zebra is the Zcash Foundation’s independent, consensus-compatible implementation of a Zcash node, written in Rust for performance and safety.