refactor: move port monitoring to pods parsing (#5100)

* refactor: move port monitoring to pods parsing

* style: fix overload and cli autocomplete

Co-authored-by: Jina Dev Bot <dev-bot@jina.ai>

docs/fundamentals/flow/executor-args.md

@@ -30,7 +30,7 @@
 | `shards` | The number of shards in the deployment running at the same time. For more details check https://docs.jina.ai/fundamentals/flow/create-flow/#complex-flow-topologies | `number` | `1` |
 | `replicas` | The number of replicas in the deployment | `number` | `1` |
 | `monitoring` | If set, spawn an http server with a prometheus endpoint to expose metrics | `boolean` | `False` |
-| `port_monitoring` | The port on which the prometheus server is exposed, default is a random port between [49152, 65535] | `number` | `random in [49152, 65535]` |
+| `port_monitoring` | The port on which the prometheus server is exposed, default is a random port between [49152, 65535] | `string` | `random in [49152, 65535]` |
 | `retries` | Number of retries per gRPC call. If <0 it defaults to max(3, num_replicas) | `number` | `-1` |
 | `floating` | If set, the current Pod/Deployment can not be further chained, and the next `.add()` will chain after the last Pod/Deployment not this current one. | `boolean` | `False` |
 | `install_requirements` | If set, install `requirements.txt` in the Hub Executor bundle to local | `boolean` | `False` |

docs/fundamentals/flow/gateway-args.md

@@ -43,6 +43,6 @@
 | `shards` | The number of shards in the deployment running at the same time. For more details check https://docs.jina.ai/fundamentals/flow/create-flow/#complex-flow-topologies | `number` | `1` |
 | `replicas` | The number of replicas in the deployment | `number` | `1` |
 | `monitoring` | If set, spawn an http server with a prometheus endpoint to expose metrics | `boolean` | `False` |
-| `port_monitoring` | The port on which the prometheus server is exposed, default is a random port between [49152, 65535] | `number` | `random in [49152, 65535]` |
+| `port_monitoring` | The port on which the prometheus server is exposed, default is a random port between [49152, 65535] | `string` | `random in [49152, 65535]` |
 | `retries` | Number of retries per gRPC call. If <0 it defaults to max(3, num_replicas) | `number` | `-1` |
 | `floating` | If set, the current Pod/Deployment can not be further chained, and the next `.add()` will chain after the last Pod/Deployment not this current one. | `boolean` | `False` |

jina/orchestrate/flow/base.py

@@ -165,7 +165,7 @@ def __init__(
         output_array_type: Optional[str] = None,
         polling: Optional[str] = 'ANY',
         port: Optional[int] = None,
-        port_monitoring: Optional[int] = None,
+        port_monitoring: Optional[str] = None,
         prefetch: Optional[int] = 1000,
         protocol: Optional[str] = 'GRPC',
         proxy: Optional[bool] = False,

jina/parsers/__init__.py

@@ -4,11 +4,10 @@
 from jina.parsers.orchestrate.runtimes.head import mixin_head_parser
 
 
-def set_pod_parser(parser=None, port_monitoring=True):
+def set_pod_parser(parser=None):
     """Set the parser for the Pod
 
     :param parser: an optional existing parser to build upon
-    :param port_monitoring: if to include the port parsing
     :return: the parser
     """
     if not parser:
@@ -33,7 +32,7 @@ def set_pod_parser(parser=None, port_monitoring=True):
     mixin_container_runtime_parser(parser)
     mixin_remote_runtime_parser(parser)
     mixin_distributed_feature_parser(parser)
-    mixin_pod_parser(parser, port_monitoring=port_monitoring)
+    mixin_pod_parser(parser)
     mixin_hub_pull_options_parser(parser)
     mixin_head_parser(parser)
 
@@ -51,7 +50,7 @@ def set_deployment_parser(parser=None):
 
         parser = set_base_parser()
 
-    set_pod_parser(parser, port_monitoring=False)
+    set_pod_parser(parser)
 
     from jina.parsers.orchestrate.deployment import mixin_base_deployment_parser
 

jina/parsers/orchestrate/deployment.py

@@ -60,11 +60,3 @@ def mixin_base_deployment_parser(parser):
         default=False,
         help='If set, connect to deployment using tls encryption',
     )
-
-    gp.add_argument(
-        '--port-monitoring',
-        type=str,
-        default=str(helper.random_port()),
-        dest='port_monitoring',
-        help=f'The port on which the prometheus server is exposed, default is a random port between [49152, 65535]',
-    )

jina/parsers/orchestrate/pod.py

@@ -6,10 +6,9 @@
 from jina.parsers.helper import _SHOW_ALL_ARGS, KVAppendAction, add_arg_group
 
 
-def mixin_pod_parser(parser, port_monitoring=True):
+def mixin_pod_parser(parser):
     """Mixing in arguments required by :class:`Pod` into the given parser.
     :param parser: the parser instance to which we add arguments
-    :param port_monitoring: if to include the port parsing
     """
 
     gp = add_arg_group(parser, title='Pod')
@@ -97,14 +96,13 @@ def mixin_pod_parser(parser, port_monitoring=True):
         help='If set, spawn an http server with a prometheus endpoint to expose metrics',
     )
 
-    if port_monitoring:
-        gp.add_argument(
-            '--port-monitoring',
-            type=int,
-            default=helper.random_port(),
-            dest='port_monitoring',
-            help=f'The port on which the prometheus server is exposed, default is a random port between [49152, 65535]',
-        )
+    gp.add_argument(
+        '--port-monitoring',
+        type=str,
+        default=str(helper.random_port()),
+        dest='port_monitoring',
+        help=f'The port on which the prometheus server is exposed, default is a random port between [49152, 65535]',
+    )
 
     gp.add_argument(
         '--retries',

jina/serve/runtimes/monitoring.py

@@ -17,4 +17,6 @@ def _setup_monitoring(self):
 
             from prometheus_client import start_http_server
 
-            start_http_server(self.args.port_monitoring, registry=self.metrics_registry)
+            start_http_server(
+                int(self.args.port_monitoring), registry=self.metrics_registry
+            )

jina_cli/autocomplete.py

@@ -301,6 +301,7 @@
             '--replicas',
             '--port',
             '--monitoring',
+            '--port-monitoring',
             '--retries',
             '--floating',
             '--install-requirements',
@@ -318,7 +319,6 @@
             '--external',
             '--deployment-role',
             '--tls',
-            '--port-monitoring',
         ],
         'client': [
             '--help',

tests/integration/monitoring/test_monitoring.py

@@ -139,15 +139,22 @@ def test_monitoring_head_few_port(port_generator, executor):
             assert resp.status_code == 200
 
 
-def test_monitoring_replicas_and_shards(port_generator, executor):
+@pytest.mark.parametrize('protocol', ['websocket', 'grpc', 'http', None])
+def test_monitoring_replicas_and_shards(port_generator, executor, protocol):
     n_shards = 2
     n_replicas = 3
     port_shards_list = [port_generator() for _ in range(n_shards * n_replicas)]
     port_head = port_generator()
     port_monitoring = ','.join([str(port) for port in [port_head] + port_shards_list])
     port1 = port_generator()
 
-    f = Flow(monitoring=True, port_monitoring=port1).add(
+    flow = (
+        Flow(protocol=protocol, monitoring=True, port_monitoring=port1)
+        if protocol
+        else Flow(monitoring=True, port_monitoring=port1)
+    )
+
+    f = flow.add(
         uses=executor,
         port_monitoring=port_monitoring,
         shards=n_shards,
@@ -186,9 +193,6 @@ def test_monitoring_replicas_and_shards(port_generator, executor):
         assert f'jina_sending_request_seconds' in str(resp.content)
 
 
-# GOOD
-
-
 def test_document_processed_total(port_generator, executor):
     port0 = port_generator()
     port1 = port_generator()
@@ -309,7 +313,6 @@ def _get_request_bytes_size():
         assert measured_request_bytes_sum > measured_request_bytes_sum_init
 
 
-# GOOD
 def test_failed_successful_request_count(port_generator, failing_executor):
     port0 = port_generator()
     port1 = port_generator()