fix(healthcheck): do not return true as response if the check if not …

…applicable (#3551)
juspay · Feb 13, 2024 · 6e103ce · 6e103ce
1 parent 02652a2
commit 6e103ce
Show file tree

Hide file tree

Showing 3 changed files with 90 additions and 67 deletions.
diff --git a/crates/api_models/src/health_check.rs b/crates/api_models/src/health_check.rs
@@ -2,7 +2,8 @@
 pub struct RouterHealthCheckResponse {
     pub database: bool,
     pub redis: bool,
-    pub locker: bool,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub vault: Option<bool>,
     #[cfg(feature = "olap")]
     pub analytics: bool,
     pub outgoing_request: bool,
@@ -17,4 +18,28 @@ pub struct SchedulerHealthCheckResponse {
     pub outgoing_request: bool,
 }
 
+pub enum HealthState {
+    Running,
+    Error,
+    NotApplicable,
+}
+
+impl From<HealthState> for bool {
+    fn from(value: HealthState) -> Self {
+        match value {
+            HealthState::Running => true,
+            HealthState::Error | HealthState::NotApplicable => false,
+        }
+    }
+}
+impl From<HealthState> for Option<bool> {
+    fn from(value: HealthState) -> Self {
+        match value {
+            HealthState::Running => Some(true),
+            HealthState::Error => Some(false),
+            HealthState::NotApplicable => None,
+        }
+    }
+}
+
 impl common_utils::events::ApiEventMetric for SchedulerHealthCheckResponse {}
diff --git a/crates/router/src/core/health_check.rs b/crates/router/src/core/health_check.rs
@@ -1,5 +1,6 @@
 #[cfg(feature = "olap")]
 use analytics::health_check::HealthCheck;
+use api_models::health_check::HealthState;
 use error_stack::ResultExt;
 use router_env::logger;
 
@@ -12,23 +13,27 @@ use crate::{
 
 #[async_trait::async_trait]
 pub trait HealthCheckInterface {
-    async fn health_check_db(&self) -> CustomResult<(), errors::HealthCheckDBError>;
-    async fn health_check_redis(&self) -> CustomResult<(), errors::HealthCheckRedisError>;
-    async fn health_check_locker(&self) -> CustomResult<(), errors::HealthCheckLockerError>;
-    async fn health_check_outgoing(&self) -> CustomResult<(), errors::HealthCheckOutGoing>;
+    async fn health_check_db(&self) -> CustomResult<HealthState, errors::HealthCheckDBError>;
+    async fn health_check_redis(&self) -> CustomResult<HealthState, errors::HealthCheckRedisError>;
+    async fn health_check_locker(
+        &self,
+    ) -> CustomResult<HealthState, errors::HealthCheckLockerError>;
+    async fn health_check_outgoing(&self)
+        -> CustomResult<HealthState, errors::HealthCheckOutGoing>;
     #[cfg(feature = "olap")]
-    async fn health_check_analytics(&self) -> CustomResult<(), errors::HealthCheckDBError>;
+    async fn health_check_analytics(&self)
+        -> CustomResult<HealthState, errors::HealthCheckDBError>;
 }
 
 #[async_trait::async_trait]
 impl HealthCheckInterface for app::AppState {
-    async fn health_check_db(&self) -> CustomResult<(), errors::HealthCheckDBError> {
+    async fn health_check_db(&self) -> CustomResult<HealthState, errors::HealthCheckDBError> {
         let db = &*self.store;
         db.health_check_db().await?;
-        Ok(())
+        Ok(HealthState::Running)
     }
 
-    async fn health_check_redis(&self) -> CustomResult<(), errors::HealthCheckRedisError> {
+    async fn health_check_redis(&self) -> CustomResult<HealthState, errors::HealthCheckRedisError> {
         let db = &*self.store;
         let redis_conn = db
             .get_redis_conn()
@@ -55,10 +60,12 @@ impl HealthCheckInterface for app::AppState {
 
         logger::debug!("Redis delete_key was successful");
 
-        Ok(())
+        Ok(HealthState::Running)
     }
 
-    async fn health_check_locker(&self) -> CustomResult<(), errors::HealthCheckLockerError> {
+    async fn health_check_locker(
+        &self,
+    ) -> CustomResult<HealthState, errors::HealthCheckLockerError> {
         let locker = &self.conf.locker;
         if !locker.mock_locker {
             let mut url = locker.host_rs.to_owned();
@@ -67,16 +74,19 @@ impl HealthCheckInterface for app::AppState {
             services::call_connector_api(self, request)
                 .await
                 .change_context(errors::HealthCheckLockerError::FailedToCallLocker)?
-                .ok();
+                .map_err(|_| {
+                    error_stack::report!(errors::HealthCheckLockerError::FailedToCallLocker)
+                })?;
+            Ok(HealthState::Running)
+        } else {
+            Ok(HealthState::NotApplicable)
         }
-
-        logger::debug!("Locker call was successful");
-
-        Ok(())
     }
 
     #[cfg(feature = "olap")]
-    async fn health_check_analytics(&self) -> CustomResult<(), errors::HealthCheckDBError> {
+    async fn health_check_analytics(
+        &self,
+    ) -> CustomResult<HealthState, errors::HealthCheckDBError> {
         let analytics = &self.pool;
         match analytics {
             analytics::AnalyticsProvider::Sqlx(client) => client
@@ -107,10 +117,14 @@ impl HealthCheckInterface for app::AppState {
                     .await
                     .change_context(errors::HealthCheckDBError::ClickhouseAnalyticsError)
             }
-        }
+        }?;
+
+        Ok(HealthState::Running)
     }
 
-    async fn health_check_outgoing(&self) -> CustomResult<(), errors::HealthCheckOutGoing> {
+    async fn health_check_outgoing(
+        &self,
+    ) -> CustomResult<HealthState, errors::HealthCheckOutGoing> {
         let request = services::Request::new(services::Method::Get, consts::OUTGOING_CALL_URL);
         services::call_connector_api(self, request)
             .await
@@ -125,6 +139,6 @@ impl HealthCheckInterface for app::AppState {
             })?;
 
         logger::debug!("Outgoing request successful");
-        Ok(())
+        Ok(HealthState::Running)
     }
 }
diff --git a/crates/router/src/routes/health.rs b/crates/router/src/routes/health.rs
@@ -45,7 +45,7 @@ async fn deep_health_check_func(state: app::AppState) -> RouterResponse<RouterHe
 
     logger::debug!("Database health check begin");
 
-    let db_status = state.health_check_db().await.map(|_| true).map_err(|err| {
+    let db_status = state.health_check_db().await.map_err(|err| {
         error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
             component: "Database",
             message: err.to_string()
@@ -56,64 +56,48 @@ async fn deep_health_check_func(state: app::AppState) -> RouterResponse<RouterHe
 
     logger::debug!("Redis health check begin");
 
-    let redis_status = state
-        .health_check_redis()
-        .await
-        .map(|_| true)
-        .map_err(|err| {
-            error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
-                component: "Redis",
-                message: err.to_string()
-            })
-        })?;
+    let redis_status = state.health_check_redis().await.map_err(|err| {
+        error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
+            component: "Redis",
+            message: err.to_string()
+        })
+    })?;
 
     logger::debug!("Redis health check end");
 
     logger::debug!("Locker health check begin");
 
-    let locker_status = state
-        .health_check_locker()
-        .await
-        .map(|_| true)
-        .map_err(|err| {
-            error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
-                component: "Locker",
-                message: err.to_string()
-            })
-        })?;
+    let locker_status = state.health_check_locker().await.map_err(|err| {
+        error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
+            component: "Locker",
+            message: err.to_string()
+        })
+    })?;
 
     #[cfg(feature = "olap")]
-    let analytics_status = state
-        .health_check_analytics()
-        .await
-        .map(|_| true)
-        .map_err(|err| {
-            error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
-                component: "Analytics",
-                message: err.to_string()
-            })
-        })?;
-
-    let outgoing_check = state
-        .health_check_outgoing()
-        .await
-        .map(|_| true)
-        .map_err(|err| {
-            error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
-                component: "Outgoing Request",
-                message: err.to_string()
-            })
-        })?;
+    let analytics_status = state.health_check_analytics().await.map_err(|err| {
+        error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
+            component: "Analytics",
+            message: err.to_string()
+        })
+    })?;
+
+    let outgoing_check = state.health_check_outgoing().await.map_err(|err| {
+        error_stack::report!(errors::ApiErrorResponse::HealthCheckError {
+            component: "Outgoing Request",
+            message: err.to_string()
+        })
+    })?;
 
     logger::debug!("Locker health check end");
 
     let response = RouterHealthCheckResponse {
-        database: db_status,
-        redis: redis_status,
-        locker: locker_status,
+        database: db_status.into(),
+        redis: redis_status.into(),
+        vault: locker_status.into(),
         #[cfg(feature = "olap")]
-        analytics: analytics_status,
-        outgoing_request: outgoing_check,
+        analytics: analytics_status.into(),
+        outgoing_request: outgoing_check.into(),
     };
 
     Ok(api::ApplicationResponse::Json(response))