more tunes on the grpc options

Author: liferoad

sdks/python/apache_beam/runners/worker/channel_factory.py

@@ -23,17 +23,24 @@
 
 class GRPCChannelFactory(grpc.StreamStreamClientInterceptor):
   DEFAULT_OPTIONS = [
-      ("grpc.keepalive_time_ms",
-       120000),  # Increased to 120s to further reduce ping frequency
-      ("grpc.keepalive_timeout_ms", 300000),
-      ("grpc.http2.max_pings_without_data",
-       0),  # Allow unlimited pings without data
-      ("grpc.keepalive_permit_without_calls",
-       True),  # Allow keepalive pings when no calls
-      ("grpc.http2.min_recv_ping_interval_without_data_ms",
-       300000),  # 5 minutes
-      ("grpc.http2.min_sent_ping_interval_without_data_ms",
-       60000),  # Increased to 60s to be more conservative
+      # Default: 30000ms (30s), increased to 180s to reduce ping frequency
+      ("grpc.keepalive_time_ms", 180000),
+      # Default: 5000ms (5s), increased to 10 minutes for stability
+      ("grpc.keepalive_timeout_ms", 600000),
+      # Default: 2, set to 0 to allow unlimited pings without data
+      ("grpc.http2.max_pings_without_data", 0),
+      # Default: False, set to True to allow keepalive pings when no calls
+      ("grpc.keepalive_permit_without_calls", True),
+      # Default: 300000ms (5min), increased to 10min for stability
+      ("grpc.http2.min_recv_ping_interval_without_data_ms", 600000),
+      # Default: 300000ms (5min), increased to 120s for conservative pings
+      ("grpc.http2.min_sent_ping_interval_without_data_ms", 120000),
+      # Default: 2, set to 0 to allow unlimited ping strikes
+      ("grpc.http2.max_ping_strikes", 0),
+      # Default: 0 (disabled), enable socket reuse for better handling
+      ("grpc.so_reuseport", 1),
+      # Default: 0 (disabled), set 30s TCP timeout for connection control
+      ("grpc.tcp_user_timeout_ms", 30000),
   ]
 
   def __init__(self):

sdks/python/apache_beam/utils/subprocess_server.py

@@ -188,14 +188,24 @@ def start(self):
       channel_options = [
           ("grpc.max_receive_message_length", -1),
           ("grpc.max_send_message_length", -1),
-          ("grpc.keepalive_time_ms",
-           120000),  # Increased to 120s to further reduce ping frequency
-          ("grpc.keepalive_timeout_ms", 300000),
+          # Default: 30000ms (30s), increased to 180s to reduce ping frequency
+          ("grpc.keepalive_time_ms", 180000),
+          # Default: 5000ms (5s), increased to 10 minutes for stability
+          ("grpc.keepalive_timeout_ms", 600000),
+          # Default: 2, set to 0 to allow unlimited pings without data
           ("grpc.http2.max_pings_without_data", 0),
+          # Default: False, set to True to allow keepalive pings when no calls
           ("grpc.keepalive_permit_without_calls", True),
-          ("grpc.http2.min_recv_ping_interval_without_data_ms", 300000),
-          ("grpc.http2.min_sent_ping_interval_without_data_ms",
-           60000),  # Increased to 60s to be more conservative
+          # Default: 300000ms (5min), increased to 10min for stability
+          ("grpc.http2.min_recv_ping_interval_without_data_ms", 600000),
+          # Default: 300000ms (5min), increased to 120s for conservative pings
+          ("grpc.http2.min_sent_ping_interval_without_data_ms", 120000),
+          # Default: 2, set to 0 to allow unlimited ping strikes
+          ("grpc.http2.max_ping_strikes", 0),
+          # Default: 0 (disabled), enable socket reuse for better handling
+          ("grpc.so_reuseport", 1),
+          # Default: 0 (disabled), set 30s TCP timeout for connection control
+          ("grpc.tcp_user_timeout_ms", 30000),
       ]
       self._grpc_channel = grpc.insecure_channel(
           endpoint, options=channel_options)