Fix actor shutdown to await merges

Author: rdettai

distribution/lambda/Makefile

@@ -56,7 +56,7 @@ indexer-package-path:
 searcher-package-path:
 	echo -n $(SEARCHER_PACKAGE_PATH)
 
-bootstrap: package check-env
+bootstrap:
 	cdk bootstrap aws://$$CDK_ACCOUNT/$$CDK_REGION
 
 deploy-hdfs: package check-env
@@ -65,17 +65,20 @@ deploy-hdfs: package check-env
 deploy-mock-data: package check-env
 	cdk deploy -a cdk/app.py MockDataStack
 
+print-mock-data-metastore: check-env
+	python -c 'from cdk import cli; cli.print_mock_data_metastore()'
+
 # address https://github.com/aws/aws-cdk/issues/20060
 before-destroy:
 	mkdir -p cdk.out
 	touch $(INDEXER_PACKAGE_PATH)
 	touch $(SEARCHER_PACKAGE_PATH)
 
-destroy-hdfs: before-destroy
+destroy-hdfs: before-destroy check-env
 	python -c 'from cdk import cli; cli.empty_hdfs_bucket()'
 	cdk destroy  --force -a cdk/app.py HdfsStack
 
-destroy-mock-data: before-destroy
+destroy-mock-data: before-destroy check-env
 	python -c 'from cdk import cli; cli.empty_mock_data_buckets()'
 	cdk destroy  --force -a cdk/app.py MockDataStack
 
@@ -108,6 +111,7 @@ bench-index:
 	done
 
 bench-search-term:
+	export QW_LAMBDA_LOG_SPAN_BOUNDARIES=true
 	mem_sizes=( 1024 2048 4096 8192 )
 	for mem_size in "$${mem_sizes[@]}"
 	do
@@ -117,6 +121,7 @@ bench-search-term:
 	done
 
 bench-search-histogram:
+	export QW_LAMBDA_LOG_SPAN_BOUNDARIES=true
 	mem_sizes=( 1024 2048 4096 8192 )
 	for mem_size in "$${mem_sizes[@]}"
 	do

distribution/lambda/cdk/cli.py

@@ -316,6 +316,15 @@ def empty_mock_data_buckets():
     _clean_s3_bucket(bucket_name)
 
 
+def print_mock_data_metastore():
+    bucket_name = _get_cloudformation_output_value(
+        app.MOCK_DATA_STACK_NAME, mock_data_stack.INDEX_STORE_BUCKET_NAME_EXPORT_NAME
+    )
+    s3 = session.client("s3")
+    response = s3.get_object(Bucket=bucket_name, Key="index/mock-sales/metastore.json")
+    print(response["Body"].read().decode())
+
+
 @cache
 def _git_commit():
     return subprocess.run(

distribution/lambda/cdk/stacks/examples/mock_data_stack.py

@@ -12,7 +12,7 @@
 from constructs import Construct
 import yaml
 
-from ..services.quickwit_service import QuickwitService
+from ..services import quickwit_service
 
 SEARCHER_FUNCTION_NAME_EXPORT_NAME = "mock-data-searcher-function-name"
 INDEX_STORE_BUCKET_NAME_EXPORT_NAME = "mock-data-index-store-bucket-name"
@@ -28,7 +28,7 @@ def __init__(
         scope: Construct,
         construct_id: str,
         index_id: str,
-        qw_svc: QuickwitService,
+        qw_svc: quickwit_service.QuickwitService,
         **kwargs,
     ):
         super().__init__(scope, construct_id, **kwargs)
@@ -83,7 +83,7 @@ def __init__(
         scope: Construct,
         construct_id: str,
         index_id: str,
-        qw_svc: QuickwitService,
+        qw_svc: quickwit_service.QuickwitService,
         api_key: str,
         **kwargs,
     ) -> None:
@@ -149,12 +149,15 @@ def __init__(
             "mock-data-index-config",
             path=index_config_local_path,
         )
-        qw_svc = QuickwitService(
+        lambda_env = quickwit_service.extract_local_env()
+        qw_svc = quickwit_service.QuickwitService(
             self,
             "Quickwit",
             index_id=index_id,
             index_config_bucket=index_config.s3_bucket_name,
             index_config_key=index_config.s3_object_key,
+            indexer_environment=lambda_env,
+            searcher_environment=lambda_env,
             indexer_package_location=indexer_package_location,
             searcher_package_location=searcher_package_location,
         )

distribution/lambda/cdk/stacks/services/indexer_service.py

@@ -32,7 +32,9 @@ def __init__(
                 "QW_LAMBDA_INDEX_CONFIG_URI": f"s3://{index_config_bucket}/{index_config_key}",
                 **environment,
             },
-            timeout=aws_cdk.Duration.minutes(15),
+            # use a strict timeout and retry policy to avoid unexpected costs
+            timeout=aws_cdk.Duration.minutes(1),
+            retry_attempts=0,
             reserved_concurrent_executions=1,
             memory_size=memory_size,
             ephemeral_storage_size=aws_cdk.Size.gibibytes(10),

distribution/lambda/cdk/stacks/services/quickwit_service.py

@@ -12,8 +12,12 @@
 
 
 def extract_local_env() -> dict[str, str]:
-    """Extracts local environment variables that start with QW_LAMBDA_"""
-    return {k: os.environ[k] for k in os.environ.keys() if k.startswith("QW_LAMBDA_")}
+    """Extracts local environment variables QW_LAMBDA_* and QW_DISABLE_TELEMETRY"""
+    return {
+        k: os.environ[k]
+        for k in os.environ.keys()
+        if (k.startswith("QW_LAMBDA_") or k == "QW_DISABLE_TELEMETRY")
+    }
 
 
 class QuickwitService(Construct):

quickwit/Cargo.lock

@@ -51,6 +51,7 @@ quickwit-doc-mapper = { workspace = true }
 quickwit-index-management = { workspace = true }
 quickwit-indexing = { workspace = true }
 quickwit-ingest = { workspace = true }
+quickwit-janitor = { workspace = true }
 quickwit-metastore = { workspace = true }
 quickwit-proto = { workspace = true }
 quickwit-rest-client = { workspace = true }

quickwit/quickwit-codegen/example/src/codegen/hello.rs

@@ -23,7 +23,7 @@ use quickwit_lambda::logger;
 
 #[tokio::main]
 async fn main() -> anyhow::Result<()> {
-    logger::setup_lambda_tracer()?;
+    logger::setup_lambda_tracer(tracing::Level::INFO)?;
     let func = service_fn(handler);
     lambda_runtime::run(func)
         .await