feat(prover): export prover traces through OTLP (matter-labs#1427)

## What ❔

This PR adds an optional `tracing` layer to `vlog` that exports all
spans to a configurable OTEL collector.

This layer is then used in most prover components to export spans with
structured information about the block number that is currently being
processed.

## Why ❔

Better observability for prover and tracing for logs with a specific
block number

## Checklist

<!-- Check your PR fulfills the following items. -->
<!-- For draft PRs check the boxes as you complete them. -->

- [x] PR title corresponds to the body of PR (we generate changelog
entries from PRs).
- [ ] Tests for the changes have been added / updated.
- [ ] Documentation comments have been added / updated.
- [x] Code has been formatted via `zk fmt` and `zk lint`.
- [x] Spellcheck has been run via `zk spellcheck`.
- [x] Linkcheck has been run via `zk linkcheck`.

Cargo.toml

@@ -111,6 +111,9 @@ num = "0.4.0"
 num_cpus = "1.13"
 num_enum = "0.7.2"
 once_cell = "1"
+opentelemetry = "0.20.0"
+opentelemetry-otlp = "0.13.0"
+opentelemetry-semantic-conventions = "0.12.0"
 pin-project-lite = "0.2.13"
 pretty_assertions = "1"
 prost = "0.12.1"
@@ -145,6 +148,7 @@ tower = "0.4.13"
 tower-http = "0.4.1"
 tracing = "0.1"
 tracing-subscriber = "0.3"
+tracing-opentelemetry = "0.21.0"
 url = "2"
 web3 = "0.19.0"
 

checks-config/era.dic

@@ -921,3 +921,4 @@ p2p
 StorageProcessor
 StorageMarker
 SIGINT
+opentelemetry

core/bin/contract-verifier/src/verifier.rs

@@ -506,6 +506,7 @@ impl JobProcessor for ContractVerifier {
     #[allow(clippy::async_yields_async)]
     async fn process_job(
         &self,
+        _job_id: &Self::JobId,
         job: VerificationRequest,
         started_at: Instant,
     ) -> tokio::task::JoinHandle<anyhow::Result<()>> {

core/bin/external_node/src/config/observability.rs

@@ -35,5 +35,6 @@ pub fn observability_config_from_env() -> anyhow::Result<ObservabilityConfig> {
         sentry_url,
         sentry_environment,
         log_format,
+        opentelemetry: None,
     })
 }

core/lib/config/src/configs/mod.rs

@@ -14,7 +14,7 @@ pub use self::{
     fri_witness_generator::FriWitnessGeneratorConfig,
     fri_witness_vector_generator::FriWitnessVectorGeneratorConfig,
     object_store::ObjectStoreConfig,
-    observability::ObservabilityConfig,
+    observability::{ObservabilityConfig, OpentelemetryConfig},
     proof_data_handler::ProofDataHandlerConfig,
     snapshots_creator::SnapshotsCreatorConfig,
     utils::PrometheusConfig,

core/lib/config/src/configs/observability.rs

@@ -6,7 +6,17 @@ pub struct ObservabilityConfig {
     pub sentry_url: Option<String>,
     /// Name of the environment to use in Sentry.
     pub sentry_environment: Option<String>,
+    /// Opentelemetry configuration.
+    pub opentelemetry: Option<OpentelemetryConfig>,
     /// Format of the logs as expected by the `vlog` crate.
     /// Currently must be either `plain` or `json`.
     pub log_format: String,
 }
+
+#[derive(Debug, Clone, PartialEq)]
+pub struct OpentelemetryConfig {
+    /// Enables export of span data of specified level (and above) using opentelemetry exporters.
+    pub level: String,
+    /// Opentelemetry HTTP collector endpoint.
+    pub endpoint: String,
+}

core/lib/config/src/testonly.rs

@@ -754,6 +754,16 @@ impl RandomConfig for configs::ObservabilityConfig {
             sentry_url: g.gen(),
             sentry_environment: g.gen(),
             log_format: g.gen(),
+            opentelemetry: g.gen(),
+        }
+    }
+}
+
+impl RandomConfig for configs::OpentelemetryConfig {
+    fn sample(g: &mut Gen<impl Rng>) -> Self {
+        Self {
+            level: g.gen(),
+            endpoint: g.gen(),
         }
     }
 }

core/lib/env_config/src/observability.rs

@@ -1,4 +1,4 @@
-use zksync_config::configs::ObservabilityConfig;
+use zksync_config::configs::{ObservabilityConfig, OpentelemetryConfig};
 
 use crate::FromEnv;
 
@@ -33,11 +33,18 @@ impl FromEnv for ObservabilityConfig {
         } else {
             "plain".to_string()
         };
+        let opentelemetry_level = std::env::var("OPENTELEMETRY_LEVEL").ok();
+        let otlp_endpoint = std::env::var("OTLP_ENDPOINT").ok();
+        let opentelemetry = match (opentelemetry_level, otlp_endpoint) {
+            (Some(level), Some(endpoint)) => Some(OpentelemetryConfig { level, endpoint }),
+            _ => None,
+        };
 
         Ok(ObservabilityConfig {
             sentry_url,
             sentry_environment,
             log_format,
+            opentelemetry,
         })
     }
 }

core/lib/protobuf_config/src/observability.rs

@@ -1,5 +1,5 @@
 use anyhow::Context as _;
-use zksync_config::configs;
+use zksync_config::configs::{self};
 use zksync_protobuf::{required, ProtoRepr};
 
 use crate::proto::observability as proto;
@@ -11,6 +11,11 @@ impl ProtoRepr for proto::Observability {
             sentry_url: self.sentry_url.clone(),
             sentry_environment: self.sentry_environment.clone(),
             log_format: required(&self.log_format).context("log_format")?.clone(),
+            opentelemetry: self
+                .opentelemetry
+                .as_ref()
+                .map(|cfg| cfg.read().context("opentelemetry"))
+                .transpose()?,
         })
     }
 
@@ -19,6 +24,25 @@ impl ProtoRepr for proto::Observability {
             sentry_url: this.sentry_url.clone(),
             sentry_environment: this.sentry_environment.clone(),
             log_format: Some(this.log_format.clone()),
+            opentelemetry: this.opentelemetry.as_ref().map(ProtoRepr::build),
+        }
+    }
+}
+
+impl ProtoRepr for proto::Opentelemetry {
+    type Type = configs::OpentelemetryConfig;
+
+    fn read(&self) -> anyhow::Result<Self::Type> {
+        Ok(Self::Type {
+            level: required(&self.level).context("level")?.clone(),
+            endpoint: required(&self.endpoint).context("endpoint")?.clone(),
+        })
+    }
+
+    fn build(this: &Self::Type) -> Self {
+        Self {
+            level: Some(this.level.clone()),
+            endpoint: Some(this.endpoint.clone()),
         }
     }
 }

core/lib/protobuf_config/src/proto/observability.proto

@@ -6,4 +6,10 @@ message Observability {
   optional string sentry_url = 1; // optional
   optional string sentry_environment = 2; // optional
   optional string log_format = 3; // required
+  optional Opentelemetry opentelemetry = 4; // optional
+}
+
+message Opentelemetry {
+  optional string level = 1; // required
+  optional string endpoint = 2; // required
 }

core/lib/queued_job_processor/src/lib.rs

@@ -46,6 +46,7 @@ pub trait JobProcessor: Sync + Send {
     /// Function that processes a job
     async fn process_job(
         &self,
+        job_id: &Self::JobId,
         job: Self::Job,
         started_at: Instant,
     ) -> JoinHandle<anyhow::Result<Self::JobArtifacts>>;
@@ -83,7 +84,7 @@ pub trait JobProcessor: Sync + Send {
                     Self::SERVICE_NAME,
                     job_id
                 );
-                let task = self.process_job(job, started_at).await;
+                let task = self.process_job(&job_id, job, started_at).await;
 
                 self.wait_for_task(job_id, started_at, task)
                     .await

core/lib/vlog/Cargo.toml

@@ -19,5 +19,12 @@ tracing-subscriber = { workspace = true, features = [
     "time",
     "json",
 ] }
+tracing-opentelemetry.workspace = true
 sentry.workspace = true
 serde_json.workspace = true
+opentelemetry = { workspace = true, features = ["rt-tokio", "trace"] }
+opentelemetry-otlp = { workspace = true, features = [
+    "http-proto",
+    "reqwest-client",
+] }
+opentelemetry-semantic-conventions.workspace = true