Fixing docker-compose example (#166)

* Fixing docker-compose example

* Removing some TODOs

* Do not hold guard across long running input -> embedding conversion

README.md

@@ -109,34 +109,28 @@ Below is an example `docker-compose.yaml` configuration to run both `ahnlich-db`
 services:  
   ahnlich_db:  
     image: ghcr.io/deven96/ahnlich-db:latest  
-    command: "ahnlich-db run --host 0.0.0.0 --enable-tracing"  
+    command: "ahnlich-db run --host 0.0.0.0 --enable-tracing --otel-endpoint http://jaeger:4317'"
     ports:  
       - "1369:1369"  
 
   ahnlich_ai:  
     image: ghcr.io/deven96/ahnlich-ai:latest  
-    command: "ahnlich-ai run --db-host ahnlich_db --host 0.0.0.0 --port 8880 --enable-tracing --supported-models all-minilm-l6-v2,resnet-50"  
+    command: "ahnlich-ai run --db-host ahnlich_db --host 0.0.0.0 -supported-models all-minilm-l6-v2,resnet-50 --enable-tracing ---otel-endpoint http://jaeger:4317'"
     ports:  
       - "1370:1370"  
-```  
-
----
-
-#### Environment Variables  
-
-The following environment variable can also be set:  
-
-- `DEMO_OTEL_URL`: Set the OpenTelemetry endpoint. Example:
-
-  ```bash  
-  DEMO_OTEL_URL=http://<otel-endpoint>  
-  ```  
-
-You can pass this variable directly to your Docker containers using `environment` in your Docker Compose file or when running `docker run`.  
-
----  
-
 
+  # optional jaeger service whenever --enable-tracing and
+  # --otel-endpoint is used
+  jaeger:
+    image: jaegertracing/all-in-one:${JAEGER_VERSION:-latest}
+    ports:
+      - "16686:16686"
+      - "1888:1888" # pprof extension
+      - "8888:8888" # Prometheus metrics exposed by the collector
+      - "8889:8889" # Prometheus exporter metrics
+      - "4317:4317" # otlp grpc
+      - "4318:4318" # otlp http
+```
 
 ### Contributing
 

ahnlich/ai/src/manager/mod.rs

@@ -29,7 +29,6 @@ use tracing_opentelemetry::OpenTelemetrySpanExt;
 type ModelThreadResponse = Result<Vec<StoreKey>, AIProxyError>;
 
 struct ModelThreadRequest {
-    // TODO: change this to a Vec of preprocessed input enum
     inputs: Vec<StoreInput>,
     response: oneshot::Sender<ModelThreadResponse>,
     preprocess_action: PreprocessAction,
@@ -198,8 +197,10 @@ impl Task for ModelThread {
     }
 
     async fn run(&self) -> TaskState {
-        if let Some(model_request) = self.request_receiver.lock().await.recv().await {
-            // TODO actually service model request in here and return
+        let mut guard = self.request_receiver.lock().await;
+        let request = guard.recv().await;
+        drop(guard);
+        if let Some(model_request) = request {
             let ModelThreadRequest {
                 inputs,
                 response,
@@ -233,8 +234,6 @@ impl ModelManager {
         model_config: ModelConfig,
         task_manager: Arc<TaskManager>,
     ) -> Result<Self, AIProxyError> {
-        // TODO: Actually load the model at this point, with supported models and throw up errors,
-        // also start up the various models with the task manager passed in
         let models = Cache::builder()
             .max_capacity(model_config.supported_models.len() as u64)
             .time_to_idle(Duration::from_secs(model_config.model_idle_time))
@@ -261,7 +260,7 @@ impl ModelManager {
         &self,
         model: &SupportedModels,
     ) -> Result<mpsc::Sender<ModelThreadRequest>, AIProxyError> {
-        let (request_sender, request_receiver) = mpsc::channel(100);
+        let (request_sender, request_receiver) = mpsc::channel(10000);
         // There may be other things needed to load a model thread
         let model_thread =
             ModelThread::new(*model, &self.config.model_cache_location, request_receiver)?;

ahnlich/dsl/src/ai.rs

@@ -237,7 +237,6 @@ pub fn parse_ai_query(input: &str) -> Result<Vec<AIQuery>, DslError> {
                     key: key.remove(0),
                 }
             }
-            // TODO: Introduce AIQuery::GetKey & AIQuery::ListClients
             Rule::create_non_linear_algorithm_index => {
                 let (store, non_linear_indices) =
                     parse_create_non_linear_algorithm_index(statement)?;

docker-compose.yml

@@ -19,8 +19,6 @@ services:
       context: ./ahnlich/
       args:
         - AHNLICH_BIN=ahnlich-db
-    environment:
-      - DEMO_OTEL_URL=http://jaeger:4317
     depends_on:
       jaeger:
         condition: service_started
@@ -35,8 +33,6 @@ services:
       context: ./ahnlich/
       args:
         - AHNLICH_BIN=ahnlich-ai
-    environment:
-      - DEMO_OTEL_URL=http://jaeger:4317
     depends_on:
       jaeger:
         condition: service_started