feat(trace-protobuf)!: Add two fields to ClientGroupedStats [SVLS-8627] (#1630)

# What does this PR do?
Add two fields to `ClientGroupedStats`:
1. `service_source`
2. `span_derived_primary_tags`

It syncs the changes on datadog-agent. See pkg/proto/datadog/trace/stats.proto in DataDog/datadog-agent#42996.

# Motivation

To support span derived primary tags for trace stats generated in Lambda extension.

# Additional Notes

@Sam-Maya is driving this effort.

# How to test the change?

This PR only includes struct changes in this PR. Will set and test the struct in future PRs.


Co-authored-by: yiming.luo <yiming.luo@datadoghq.com>

Author: lym953

.github/workflows/verify-proto-files.yml

@@ -3,7 +3,7 @@ on:
   pull_request:
     types: [ opened, synchronize, reopened ]
 env:
-  DATADOG_AGENT_TAG: "7f6d07c93ba087f23f80a3f0c2da4b1f3dc664d7"
+  DATADOG_AGENT_TAG: "bdcdd8cf1ba4090a29b96d5669cfab5dd81814b1"
   rust_version: "1.84.1"
   CARGO_TERM_COLOR: always
   CARGO_INCREMENTAL: 0

libdd-trace-protobuf/build.rs

@@ -176,6 +176,11 @@ fn generate_protobuf() {
         "ClientGroupedStats.HTTP_endpoint",
         "#[serde(default)] #[serde(rename = \"HTTPEndpoint\")]",
     );
+    config.field_attribute("ClientGroupedStats.service_source", "#[serde(default)]");
+    config.field_attribute(
+        "ClientGroupedStats.span_derived_primary_tags",
+        "#[serde(default)]",
+    );
 
     config.field_attribute(
         "ClientGroupedStats.okSummary",

libdd-trace-protobuf/src/pb.rs

@@ -646,6 +646,18 @@ pub struct ClientGroupedStats {
     #[serde(default)]
     #[serde(rename = "HTTPEndpoint")]
     pub http_endpoint: ::prost::alloc::string::String,
+    /// @inject_tag: msg:"srv_src"
+    #[prost(string, tag = "21")]
+    #[serde(default)]
+    pub service_source: ::prost::alloc::string::String,
+    /// used to identify service override origin
+    /// span_derived_primary_tags are user-configured tags that are extracted from spans and used for stats aggregation
+    /// E.g., `aws.s3.bucket`, `http.url`, or any custom tag
+    #[prost(string, repeated, tag = "22")]
+    #[serde(default)]
+    pub span_derived_primary_tags: ::prost::alloc::vec::Vec<
+        ::prost::alloc::string::String,
+    >,
 }
 /// Trilean is an expanded boolean type that is meant to differentiate between being unset and false.
 #[derive(Clone, Copy, Debug, PartialEq, Eq, Hash, PartialOrd, Ord, ::prost::Enumeration)]

libdd-trace-protobuf/src/pb/stats.proto

@@ -103,4 +103,9 @@ message ClientGroupedStats {
 	string GRPC_status_code = 18;
 	string HTTP_method = 19; // HTTP method of the request
 	string HTTP_endpoint = 20; // Http route or quantized/simplified URL path
+	string service_source = 21; // @inject_tag: msg:"srv_src"
+	// used to identify service override origin
+	// span_derived_primary_tags are user-configured tags that are extracted from spans and used for stats aggregation
+	// E.g., `aws.s3.bucket`, `http.url`, or any custom tag
+	repeated string span_derived_primary_tags = 22;
 }

libdd-trace-protobuf/src/pb_test.rs

@@ -112,6 +112,8 @@ mod tests {
                     grpc_status_code: "0".to_string(),
                     http_endpoint: "/test".to_string(),
                     http_method: "GET".to_string(),
+                    service_source: "".to_string(),
+                    span_derived_primary_tags: vec![],
                 }],
                 agent_time_shift: 0,
             }],

libdd-trace-stats/src/span_concentrator/aggregation.rs

@@ -295,7 +295,9 @@ fn encode_grouped_stats(key: OwnedAggregationKey, group: GroupedStats) -> pb::Cl
         },
         http_method: key.http_method,
         http_endpoint: key.http_endpoint,
-        grpc_status_code: String::new(), // currently not used
+        grpc_status_code: String::new(),   // currently not used
+        service_source: String::new(),     // set by the agent
+        span_derived_primary_tags: vec![], // set by the agent
     }
 }
 

libdd-trace-utils/src/stats_utils.rs

@@ -209,6 +209,8 @@ mod mini_agent_tests {
                     grpc_status_code: "0".to_string(),
                     http_endpoint: "/test".to_string(),
                     http_method: "GET".to_string(),
+                    service_source: "".to_string(),
+                    span_derived_primary_tags: vec![],
                 }],
                 agent_time_shift: 0,
             }],