Add timeouts and improve health checks

Changes:
- Health checks now have 5 second timeout (was no timeout)
- Chat requests have 120 second timeout (was no timeout)
- Health checks verify response body structure, not just status code
- Ollama: Verifies 'models' field exists in /api/tags response
- OpenAI: Verifies 'data' field exists in /models response

This prevents false-positive health checks where the endpoint
responds but actual API calls timeout (524 Gateway Timeout).

Co-authored-by: Qwen-Coder <qwen-coder@alibabacloud.com>

src/providers/ollama.rs

@@ -98,9 +98,20 @@ impl Provider for OllamaProvider {
     }
 
     async fn health_check(&self) -> Result<bool> {
+        // First check if the API is reachable
         let url = format!("{}/api/tags", self.base_url);
-        let response = self.apply_auth(self.client.get(&url)).send().await?;
-        Ok(response.status().is_success())
+        let response = self.apply_auth(self.client.get(&url))
+            .timeout(std::time::Duration::from_secs(5))
+            .send()
+            .await?;
+        
+        if !response.status().is_success() {
+            return Ok(false);
+        }
+
+        // Also verify we can parse the response (deeper health check)
+        let body: serde_json::Value = response.json().await?;
+        Ok(body.get("models").is_some())
     }
 
     async fn list_models(&self) -> Result<Vec<String>> {
@@ -159,6 +170,7 @@ impl Provider for OllamaProvider {
                 self.client
                     .post(&url)
                     .header("Content-Type", "application/json")
+                    .timeout(std::time::Duration::from_secs(120)) // 2 minute timeout for chat
             )
             .json(&ollama_request)
             .send()

src/providers/openai.rs

@@ -49,10 +49,17 @@ impl Provider for OpenAiProvider {
             .client
             .get(&url)
             .header("Authorization", format!("Bearer {}", self.api_key))
+            .timeout(std::time::Duration::from_secs(5))
             .send()
             .await?;
 
-        Ok(response.status().is_success())
+        if !response.status().is_success() {
+            return Ok(false);
+        }
+
+        // Verify we can parse the response
+        let body: serde_json::Value = response.json().await?;
+        Ok(body.get("data").is_some())
     }
 
     async fn list_models(&self) -> Result<Vec<String>> {
@@ -152,6 +159,7 @@ impl Provider for OpenAiProvider {
             .post(&url)
             .header("Authorization", format!("Bearer {}", self.api_key))
             .header("Content-Type", "application/json")
+            .timeout(std::time::Duration::from_secs(120)) // 2 minute timeout for chat
             .json(&req_body)
             .send()
             .await?;