feat(observability): OpenTelemetry golden-signals middleware

.gitignore

@@ -41,6 +41,7 @@ test_logs
 # pre-commit config
 .pre-commit-config.yaml
 .cargo
+!.cargo/config.toml
 
 #gradle files
 .gradle

Cargo.toml

@@ -63,13 +63,19 @@ diesel = { version = "2.2.4", package = "juspay_diesel", features = [
 ] }
 fred = { version = "9.2.1" }
 futures-util = "0.3.28"
+humantime = "2.1"
 inventory = "0.3"
 itertools = { version = "0.10.5" }
 jsonlogic = { version = "0.5.5", package = "juspay_jsonlogic" }
 jsonschema = "~0.17"
 leptos = { version = "0.6.11" }
 log = { version = "0.4.20", features = ["kv_unstable_serde"] }
 once_cell = { version = "1.18.0" }
+opentelemetry = { version = "0.27", default-features = false, features = ["metrics"] }
+opentelemetry_sdk = { version = "0.27", default-features = false, features = ["metrics", "rt-tokio"] }
+opentelemetry-prometheus = { version = "0.27", default-features = false }
+opentelemetry-otlp = { version = "0.27", default-features = false, features = ["metrics", "http-proto", "reqwest-client"] }
+prometheus = { version = "0.13", default-features = false }
 regex = "1.9.1"
 reqwest = { version = "0.11.18", features = ["json"] }
 rs-snowflake = "0.6.0"
@@ -79,6 +85,7 @@ serde_json = { version = "1.0.140" }
 secrecy = "0.10"
 strum = "0.25"
 strum_macros = "0.25"
+thiserror = "1"
 tokio = { version = "1.29.1", features = ["full"] }
 toml = { version = "0.8.8", features = ["preserve_order"] }
 tracing = "0.1.44"

README.md

@@ -150,6 +150,14 @@ Want a broader systems view? Open the [DeepWiki architecture guide](https://deep
 - [Context7 LLM-friendly docs](https://context7.com/juspay/superposition)
 - [DeepWiki repository guide](https://deepwiki.com/juspay/superposition)
 
+## Metrics & observability
+
+The HTTP API exposes Prometheus metrics on `SUPERPOSITION_METRICS_PORT` (default `9091`):
+
+```bash
+curl http://localhost:9091/metrics
+```
+
 ## Contributing
 
 We welcome contributions across the platform, clients, docs, and examples.

crates/experimentation_platform/src/api/experiment_groups/handlers.rs

@@ -64,9 +64,9 @@ use crate::api::{
     experiments::{
         cac_api::validate_context,
         helpers::{
-            validate_change_reason_with_function, hash,
-            validate_and_add_experiment_group_id,
+            hash, validate_and_add_experiment_group_id,
             validate_and_remove_experiment_group_id,
+            validate_change_reason_with_function,
         },
     },
 };

crates/experimentation_platform/src/api/experiments/handlers.rs

@@ -89,9 +89,9 @@ use crate::api::{
     },
     experiments::{
         helpers::{
-            validate_change_reason_with_function,
             get_control_overrides_from_exp_id, put_experiments_in_redis,
-            validate_control_overrides, validate_delete_experiment_variants,
+            validate_change_reason_with_function, validate_control_overrides,
+            validate_delete_experiment_variants,
         },
         types::StartedByChangeSet,
     },
@@ -170,7 +170,7 @@ async fn create_handler(
         &workspace_context,
         &change_reason,
         &state,
-        &user
+        &user,
     )
     .await?;
 
@@ -462,7 +462,7 @@ async fn conclude_handler(
         &workspace_context,
         &req.change_reason,
         &state,
-        &user
+        &user,
     )
     .await?;
 
@@ -745,7 +745,7 @@ async fn discard_handler(
         &workspace_context,
         &req.change_reason,
         &state,
-        &user
+        &user,
     )
     .await?;
 
@@ -1359,7 +1359,7 @@ async fn ramp_handler(
         &workspace_context,
         &change_reason,
         &state,
-        &user
+        &user,
     )
     .await?;
 
@@ -1568,7 +1568,7 @@ async fn update_handler(
         &workspace_context,
         &change_reason,
         &state,
-        &user
+        &user,
     )
     .await?;
 
@@ -1907,7 +1907,7 @@ async fn pause_handler(
         &workspace_context,
         &req.change_reason,
         &state,
-        &user
+        &user,
     )
     .await?;
 
@@ -2004,7 +2004,7 @@ async fn resume_handler(
         &workspace_context,
         &req.change_reason,
         &state,
-        &user
+        &user,
     )
     .await?;
 

crates/experimentation_platform/src/api/experiments/helpers.rs

@@ -790,7 +790,8 @@ pub async fn validate_change_reason_with_function(
         change_reason: change_reason.clone(),
     };
 
-    let headers_map = construct_header_map(workspace_context, vec![("x-user", user_str)])?;
+    let headers_map =
+        construct_header_map(workspace_context, vec![("x-user", user_str)])?;
 
     let response = http_client
         .post(&url)

crates/service_utils/Cargo.toml

@@ -21,10 +21,16 @@ diesel = { workspace = true }
 diesel-adapter = { version = "1.2.0" }
 fred = { workspace = true, features = ["metrics"] }
 futures-util = { workspace = true }
+humantime = { workspace = true }
 inventory = { workspace = true }
 log = { workspace = true }
 once_cell = { workspace = true }
 openidconnect = "3.5.0"
+opentelemetry = { workspace = true }
+opentelemetry_sdk = { workspace = true }
+opentelemetry-otlp = { workspace = true }
+opentelemetry-prometheus = { workspace = true }
+prometheus = { workspace = true }
 rand = "0.8"
 tokio = { workspace = true }
 tracing = { workspace = true }
@@ -42,6 +48,7 @@ superposition_types = { workspace = true, features = [
     "api",
     "diesel_derives",
 ] }
+thiserror = { workspace = true }
 url = { workspace = true }
 urlencoding = "~2.1.2"
 uuid = {workspace = true}

crates/service_utils/src/lib.rs

@@ -1,10 +1,15 @@
 #![deny(unused_crate_dependencies)]
+// opentelemetry_otlp is only used in cfg(not(test)) code; suppress the lint
+// when compiling tests.
+#[cfg(test)]
+use opentelemetry_otlp as _;
 pub mod aws;
 pub mod db;
 pub mod encryption;
 pub mod extensions;
 pub mod helpers;
 pub mod middlewares;
+pub mod observability;
 pub mod redis;
 pub mod registry;
 pub mod service;