Faucet Filtering (#19988)

## Description 

Faucet filtering for paths /v1/gas, /gas, and group all /v1/status
More conditional checks to prevent panic

## Test plan 

How did you test the new or updated feature?

Tested locally, to confirm on testnet

## Release notes

Check each box that your changes affect. If none of the boxes relate to
your changes, release notes aren't required.

For each box you select, include information after the relevant heading
that describes the impact of your changes that a user might notice and
any actions they must take to implement updates.

- [ ] Protocol: 
- [ ] Nodes (Validators and Full nodes): 
- [ ] Indexer: 
- [ ] JSON-RPC: 
- [ ] GraphQL: 
- [ ] CLI: 
- [ ] Rust SDK:
- [ ] REST API:

crates/sui-faucet/src/metrics.rs

@@ -135,47 +135,83 @@ impl FaucetMetrics {
 
 impl MetricsCallbackProvider for RequestMetrics {
     fn on_request(&self, path: String) {
+        let normalized_path = normalize_path(&path);
+        if !is_path_tracked(normalized_path) {
+            return;
+        }
+
         self.total_requests_received
-            .with_label_values(&[path.as_str()])
+            .with_label_values(&[normalized_path])
             .inc();
     }
 
     fn on_response(&self, path: String, latency: Duration, _status: u16, grpc_status_code: Code) {
+        let normalized_path = normalize_path(&path);
+        if !is_path_tracked(normalized_path) {
+            return;
+        }
+
         self.process_latency
-            .with_label_values(&[path.as_str()])
+            .with_label_values(&[normalized_path])
             .observe(latency.as_secs_f64());
 
         match grpc_status_code {
             Code::Ok => {
                 self.total_requests_succeeded
-                    .with_label_values(&[path.as_str()])
+                    .with_label_values(&[normalized_path])
                     .inc();
             }
             Code::Unavailable | Code::ResourceExhausted => {
                 self.total_requests_shed
-                    .with_label_values(&[path.as_str()])
+                    .with_label_values(&[normalized_path])
                     .inc();
             }
             _ => {
                 self.total_requests_failed
-                    .with_label_values(&[path.as_str()])
+                    .with_label_values(&[normalized_path])
                     .inc();
             }
         }
     }
 
     fn on_start(&self, path: &str) {
+        let normalized_path = normalize_path(path);
+        if !is_path_tracked(normalized_path) {
+            return;
+        }
+
         self.current_requests_in_flight
-            .with_label_values(&[path])
+            .with_label_values(&[normalized_path])
             .inc();
     }
 
     fn on_drop(&self, path: &str) {
+        let normalized_path = normalize_path(path);
+        if !is_path_tracked(normalized_path) {
+            return;
+        }
+
         self.total_requests_disconnected
-            .with_label_values(&[path])
+            .with_label_values(&[normalized_path])
             .inc();
         self.current_requests_in_flight
-            .with_label_values(&[path])
+            .with_label_values(&[normalized_path])
             .dec();
     }
 }
+
+/// Normalizes the given path to handle variations across different deployments.
+/// Specifically, it trims dynamic segments from the `/v1/status/` endpoint.
+pub fn normalize_path(path: &str) -> &str {
+    if path.starts_with("/v1/status/") {
+        return "/v1/status";
+    }
+
+    path
+}
+
+/// Determines whether the given path should be tracked for metrics collection.
+/// Only specified paths relevant to monitoring are included.
+pub fn is_path_tracked(path: &str) -> bool {
+    matches!(path, "/v1/gas" | "/gas" | "/v1/status")
+}

crates/sui-faucet/src/metrics_layer.rs

@@ -13,7 +13,7 @@ use prometheus::{HistogramTimer, Registry};
 use tower::{load_shed::error::Overloaded, BoxError, Layer, Service, ServiceExt};
 use tracing::{error, info, warn};
 
-use crate::metrics::RequestMetrics;
+use crate::metrics::{is_path_tracked, normalize_path, RequestMetrics};
 use http::Request;
 
 /// Tower Layer for tracking metrics in Prometheus related to number, success-rate and latency of
@@ -81,16 +81,19 @@ where
 
         let future = Box::pin(async move {
             let resp = inner.oneshot(req).await;
-            match &resp {
-                Ok(resp) if !resp.status().is_success() => {
-                    metrics.failed(None, Some(resp.status()))
-                }
-                Ok(_) => metrics.succeeded(),
-                Err(err) => {
-                    if err.is::<Overloaded>() {
-                        metrics.shed();
-                    } else {
-                        metrics.failed(Some(err), None);
+
+            if let Some(metrics) = metrics {
+                match &resp {
+                    Ok(resp) if !resp.status().is_success() => {
+                        metrics.failed(None, Some(resp.status()))
+                    }
+                    Ok(_) => metrics.succeeded(),
+                    Err(err) => {
+                        if err.is::<Overloaded>() {
+                            metrics.shed();
+                        } else {
+                            metrics.failed(Some(err), None);
+                        }
                     }
                 }
             }
@@ -110,25 +113,31 @@ impl<Res> Future for RequestMetricsFuture<Res> {
 }
 
 impl MetricsGuard {
-    fn new(metrics: Arc<RequestMetrics>, path: &str) -> Self {
+    fn new(metrics: Arc<RequestMetrics>, path: &str) -> Option<Self> {
+        let normalized_path = normalize_path(path);
+
+        if !is_path_tracked(normalized_path) {
+            return None;
+        }
+
         metrics
             .total_requests_received
-            .with_label_values(&[path])
+            .with_label_values(&[normalized_path])
             .inc();
         metrics
             .current_requests_in_flight
-            .with_label_values(&[path])
+            .with_label_values(&[normalized_path])
             .inc();
-        MetricsGuard {
+        Some(MetricsGuard {
             timer: Some(
                 metrics
                     .process_latency
-                    .with_label_values(&[path])
+                    .with_label_values(&[normalized_path])
                     .start_timer(),
             ),
             metrics,
-            path: path.to_string(),
-        }
+            path: normalized_path.to_string(),
+        })
     }
 
     fn succeeded(mut self) {
@@ -183,22 +192,28 @@ impl MetricsGuard {
 
 impl Drop for MetricsGuard {
     fn drop(&mut self) {
-        self.metrics
+        if self
+            .metrics
             .current_requests_in_flight
-            .with_label_values(&[&self.path])
-            .dec();
-
-        // Request was still in flight when the guard was dropped, implying the client disconnected.
-        if let Some(timer) = self.timer.take() {
-            let elapsed = timer.stop_and_record();
+            .get_metric_with_label_values(&[&self.path])
+            .is_ok()
+        {
             self.metrics
-                .total_requests_disconnected
+                .current_requests_in_flight
                 .with_label_values(&[&self.path])
-                .inc();
-            info!(
-                "Request disconnected for path {} in {:.2}s",
-                self.path, elapsed
-            );
+                .dec();
+
+            if let Some(timer) = self.timer.take() {
+                let elapsed = timer.stop_and_record();
+                self.metrics
+                    .total_requests_disconnected
+                    .with_label_values(&[&self.path])
+                    .inc();
+                info!(
+                    "Request disconnected for path {} in {:.2}s",
+                    self.path, elapsed
+                );
+            }
         }
     }
 }