Add Apalis job timeout (15m)

crates/core/src/config.rs

@@ -61,10 +61,20 @@ pub struct WorkerConfig {
     pub refresh_project_concurrency: usize,
     /// Number of retry attempts for failed jobs.
     pub retry_attempts: usize,
+    /// Maximum wall-clock seconds allowed for a single job attempt.
+    #[serde(default = "default_job_timeout_secs")]
+    pub job_timeout_secs: u64,
 }
 
 impl Default for WorkerConfig {
     fn default() -> Self {
-        Self { workflow_run_concurrency: 3, refresh_project_concurrency: 3, retry_attempts: 5 }
+        Self {
+            workflow_run_concurrency: 3,
+            refresh_project_concurrency: 3,
+            retry_attempts: 5,
+            job_timeout_secs: default_job_timeout_secs(),
+        }
     }
 }
+
+fn default_job_timeout_secs() -> u64 { 15 * 60 }

crates/jobs/src/lib.rs

@@ -86,8 +86,13 @@ pub fn create_monitor(
     context: JobContext,
     config: &WorkerConfig,
 ) -> Monitor {
-    let &WorkerConfig { workflow_run_concurrency, refresh_project_concurrency, retry_attempts } =
-        config;
+    let &WorkerConfig {
+        workflow_run_concurrency,
+        refresh_project_concurrency,
+        retry_attempts,
+        job_timeout_secs,
+    } = config;
+    let job_timeout = Duration::from_secs(job_timeout_secs);
 
     let backoff = ExponentialBackoffMaker::new(
         Duration::from_secs(1),
@@ -112,6 +117,7 @@ pub fn create_monitor(
         .register(move |_| {
             WorkerBuilder::new("workflow-run-worker")
                 .backend(storage1.workflow_run.clone())
+                .timeout(job_timeout)
                 .retry(retry1.clone())
                 .enable_tracing()
                 .catch_panic()
@@ -128,6 +134,7 @@ pub fn create_monitor(
         .register(move |_| {
             WorkerBuilder::new("refresh-project-worker")
                 .backend(storage2.refresh_project.clone())
+                .timeout(job_timeout)
                 .retry(retry2.clone())
                 .enable_tracing()
                 .catch_panic()