Replace hyperlight-specific metrics with metrics crate

Signed-off-by: Ludvig Liljenberg <[email protected]>

Author: ludfjig

Cargo.lock

@@ -58,7 +58,6 @@ clippy target=default-target: (check target)
 test target=default-target features="": (test-inprocess target) (test-seccomp target features)
     cargo test {{ if features =="" {''} else if features=="no-default-features" {"--no-default-features" } else {"--no-default-features -F " + features } }}  --profile={{ if target == "debug" {"dev"} else { target } }}
     cargo test test_metrics {{ if features =="" {''} else if features=="no-default-features" {"--no-default-features" } else {"--no-default-features -F " + features } }}  --profile={{ if target == "debug" {"dev"} else { target } }} -- --ignored 
-    cargo test test_gather_metrics {{ if features =="" {''} else if features=="no-default-features" {"--no-default-features" } else {"--no-default-features -F " + features } }}  --profile={{ if target == "debug" {"dev"} else { target } }} -- --ignored
 
 test-inprocess target=default-target:
   {{ if target == "debug" { "cargo test --features='inprocess'; cargo test test_metrics --features='inprocess' -- --ignored; cargo test test_gather_metrics --features='inprocess' -- --ignored" } else {"echo 'inprocess tests are not run for release builds'" } }} 
@@ -83,4 +82,4 @@ bench target=default-target features="":
 bench-download os hypervisor tag="":
     gh release download {{ tag }} -D ./src/hyperlight_wasm/target/ -p benchmarks_{{ os }}_{{ hypervisor }}.tar.gz
     mkdir {{ mkdir-arg }} ./src/hyperlight_wasm/target/criterion
-    tar -zxvf ./src/hyperlight_wasm/target/benchmarks_{{ os }}_{{ hypervisor }}.tar.gz -C ./src/hyperlight_wasm/target/criterion/ --strip-components=1
+    tar -zxvf ./src/hyperlight_wasm/target/benchmarks_{{ os }}_{{ hypervisor }}.tar.gz -C ./src/hyperlight_wasm/target/criterion/ --strip-components=1

Justfile

@@ -2,42 +2,31 @@
 
 hyperlight-wasm provides the following observability features:
 
-* [Metrics](#metrics) metrics are provided using Prometheus.
+* [Metrics](#metrics) metrics are provided using `metrics` crate.
 
 ## Metrics
 
-Hyperlight-wasm provides metrics using Prometheus. The metrics are registered using either the [default_registry](https://docs.rs/prometheus/latest/prometheus/fn.default_registry.html) or a registry instance provided by the host application.
 
-To provide a registry to hyperlight_wasm, use the `set_metrics_registry` function and pass a reference to a registry with `static` lifetime:
+The following metrics are provided and are enabled by default:
 
-```rust
-use hyperlight_wasm::set_metrics_registry;
-use prometheus::Registry;
-use lazy_static::lazy_static;
+* `active_proto_wasm_sandboxes` - A gauge indicating the number of currently active proto wasm sandboxes
+* `active_wasm_sandboxes` - A gauge indicating the number of currently active wasm sandboxes
+* `active_loaded_wasm_sandboxes` - A gauge indicating the number of currently loaded wasm sandboxes
+* `proto_wasm_sandboxes_total` - A counter indicating the total number of proto wasm sandboxes created during the lifetime of the process
+* `wasm_sandboxes_total` - A counter indicating the total number of wasm sandboxes created during the lifetime of the process
+* `loaded_wasm_sandboxes_total` - A counter indicating the total number of loaded wasm sandboxes created during the lifetime of the process
+* `sandbox_loads_total` - A counter indicating how many times a wasm sandbox has been loaded into a loaded wasm sandbox during the lifetime of the process
+* `sandbox_unloads_total` - A counter indicating how many times a loaded wasm sandbox has been unloaded into a wasm sandbox during the lifetime of the process
 
-lazy_static! {
-    static ref REGISTRY: Registry = Registry::new();
-}
 
-set_metrics_registry(&REGISTRY);
-```
+In addition, regular Hyperlight provides the following metrics: 
 
-The following metrics are provided and are enabled by default:
+* `guest_errors_total` - A counter indicating how many times a guest error has occured 
+* `guest_cancellations_total` - The number of times guest execution has timed out
+
+If cargo feature `function_call_metrics` is enabled:
 
-* `hyperlight_guest_error_count` - a vector of counters that tracks the number of guest errors by code and message.
-* `hyperlight_number_of_cancelled_guest_execution` - a counter that tracks the number of guest executions that have been cancelled because the execution time exceeded the time allowed.
-* `current_number_of_wasm_sandboxes` - a gauge that tracks the current number of wasm sandboxes in this process.
-* `current_number_of_loaded_wasm_sandboxes` - a gauge that tracks the current number of loaded wasm sandboxes in this process.
-* `number_of_unloads_of_loaded_wasm_sandboxes` - a counter that tracks the number of times that unload_module has been called on a LoadedWasmSandbox.
-* `number_of_loads_of_wasm_sandboxes` - a counter that tracks the number of times that load_module has been called on a WasmSandbox.
-* `current_number_of_proto_wasm_sandboxes` - a gauge that tracks the current number of proto wasm sandboxes in this process.
-* `total_number_of_wasm_sandboxes` - a counter that tracks the total number of wasm sandboxes that have been created by this process.
-* `total_number_of_loaded_wasm_sandboxes` - a counter that tracks the total number of loaded wasm sandboxes that have been created by this process.
-* `total_number_of_proto_wasm_sandboxes` - a counter that tracks the total number of proto wasm sandboxes that have been created by this process.
-
-The following metrics are provided and are enabled by default using the feature `function_call_metrics` but can be disabled:
-
-* `hyperlight_guest_function_call_duration_microseconds` - a vector of histograms that tracks the execution time of guest functions in microseconds by function name. The histogram also tracks the number of calls to each function.
-* `hyperlight_host_function_calls_duration_microseconds` - a vector of histograms that tracks the execution time of host functions in microseconds by function name. The histogram also tracks the number of calls to each function.
+* `guest_call_duration_seconds` - Histogram for the execution time of guest function calls
+* `host_call_duration_seconds` - Histogram for the execution time of host function calls
 
 There is an example of how to gather metrics in the [examples/metrics](../src/hyperlight_wasm/examples/metrics) directory.

docs/observability.md

@@ -34,25 +34,23 @@ test = true
 hyperlight-host = { workspace = true }
 libc = { version = "0.2.171" }
 once_cell = "1.21.3"
-strum = "0.27"
 tracing = "0.1.27"
 log = "0.4.27"
 cfg-if = { version = "1" }
+metrics = "0.24.2"
 
 [target.'cfg(windows)'.dependencies]
-windows = { version = "0.61", features = [
-    "Win32_System_Threading",
-] }
+windows = { version = "0.61", features = ["Win32_System_Threading"] }
 page_size = "0.6.0"
 
 [dev-dependencies]
 examples_common = { path = "../examples_common" }
 criterion = { version = "0.5.1", features = ["html_reports"] }
-lazy_static = "1.4.0"
-prometheus = "0.13.3"
 crossbeam-queue = "0.3"
 blake3 = "1.8"
 toml = "0.8.19"
+metrics-util = "0.19.1"
+metrics-exporter-prometheus = "0.17"
 
 [build-dependencies]
 chrono = "0.4"