feat: make worker job polling interval config

- Previously, the value was hardcoded to 5000ms. This could lead to long
  wait between processing jobs and hence a delay in result being
  available in the coprocessor db.

- Now, make this interval configurable via command line argument. Can be
  fine tuned as per needed performance.

fhevm-engine/coprocessor/README.md

@@ -50,6 +50,7 @@ Reload database and apply schemas from scratch
 make recreate_db
 ```
 Run the server and background fhe worker
+
+```
+cargo run -- --run-server --run-bg-worker --worker-polling-interval-ms 1000
 ```
-cargo run -- --run-server --run-bg-worker
-```

fhevm-engine/coprocessor/src/daemon_cli.rs

@@ -11,6 +11,10 @@ pub struct Args {
     #[arg(long)]
     pub run_bg_worker: bool,
 
+    /// Polling interval for the background worker to fetch jobs
+    #[arg(long, default_value_t = 5000)]
+    pub worker_polling_interval_ms: u64,
+
     /// Generate fhe keys and exit
     #[arg(long)]
     pub generate_fhe_keys: bool,

fhevm-engine/coprocessor/src/tfhe_worker.rs

@@ -85,7 +85,7 @@ async fn tfhe_worker_cycle(
                     WORK_ITEMS_NOTIFICATIONS_COUNTER.inc();
                     info!(target: "tfhe_worker", "Received work_available notification from postgres");
                 },
-                _ = tokio::time::sleep(tokio::time::Duration::from_millis(5000)) => {
+                _ = tokio::time::sleep(tokio::time::Duration::from_millis(args.worker_polling_interval_ms)) => {
                     WORK_ITEMS_POLL_COUNTER.inc();
                     info!(target: "tfhe_worker", "Polling the database for more work on timer");
                 },