Struct IpfrsMetrics 

pub struct IpfrsMetrics {

    pub registry: Registry,
    pub blocks_added: Counter,
    pub blocks_fetched: Counter,
    pub blocks_deleted: Counter,
    pub block_add_bytes: Counter,
    pub block_fetch_latency: Histogram,
    pub cache_hit_rate: Gauge,
    pub dht_provide_calls: Counter,
    pub dht_find_providers_calls: Counter,
    pub dht_provider_records: Gauge,
    pub inference_sessions_started: Counter,
    pub inference_sessions_completed: Counter,
    pub inference_session_latency: Histogram,
    pub proof_tree_depth: Histogram,
    pub gossipsub_messages_sent: Counter,
    pub gossipsub_messages_received: Counter,
    pub gossipsub_mesh_peers: Gauge,
    pub storage_bytes_total: Gauge,
    pub storage_blocks_total: Gauge,
    pub gc_runs: Counter,
    pub gc_blocks_collected: Counter,
}

Global metrics registry for IPFRS.

Each field is a Prometheus primitive registered in a private registry so that multiple instances (e.g., in tests) do not conflict with each other.

use ipfrs_interface::metrics::IpfrsMetrics;

let m = IpfrsMetrics::new().expect("failed to create metrics registry");
m.blocks_added.inc();
let text = m.render();
assert!(text.contains("ipfrs_blocks_added_total"));

Fields

registry: Registry

Private registry — metrics exported via render().

blocks_added: Counter

Total blocks added to local storage.

blocks_fetched: Counter

Total blocks fetched from local storage or network.

blocks_deleted: Counter

Total blocks deleted.

block_add_bytes: Counter

Total bytes written via add_bytes / add_file.

block_fetch_latency: Histogram

Latency (seconds) for block fetch operations.

cache_hit_rate: Gauge

Rolling cache hit-rate (0.0 – 1.0).

dht_provide_calls: Counter

Number of DHT provide calls.

dht_find_providers_calls: Counter

Number of DHT find_providers calls.

dht_provider_records: Gauge

Current number of provider records held by this node.

inference_sessions_started: Counter

Inference sessions started.

inference_sessions_completed: Counter

Inference sessions completed successfully.

inference_session_latency: Histogram

End-to-end latency (seconds) for completed inference sessions.

proof_tree_depth: Histogram

Depth of proof trees emitted during inference.

gossipsub_messages_sent: Counter

Messages published to GossipSub topics.

gossipsub_messages_received: Counter

Messages received from GossipSub topics.

gossipsub_mesh_peers: Gauge

Current number of mesh peers across all topics.

storage_bytes_total: Gauge

Total bytes currently stored on disk (updated periodically).

storage_blocks_total: Gauge

Total number of blocks currently stored on disk.

gc_runs: Counter

Number of GC runs completed.

gc_blocks_collected: Counter

Total blocks collected (deleted) across all GC runs.

Implementations

impl IpfrsMetrics

pub fn new() -> Result<Self, Error>

Create a new metrics registry with all counters, gauges, and histograms initialised to zero.

Errors

Returns a prometheus::Error if any metric registration fails (which should only happen if two metrics share the same name).

pub fn render(&self) -> String

Render all registered metrics in Prometheus text exposition format.

Returns an empty string if encoding fails.

Trait Implementations

impl Default for IpfrsMetrics

fn default() -> Self

Returns the “default value” for a type.