this will break the dashboard

Author: sanderegg

services/autoscaling/src/simcore_service_autoscaling/modules/cluster_scaling/_auto_scaling_core.py

@@ -315,13 +315,13 @@ async def _try_attach_pending_ec2s(
     all_drained_nodes = (
         cluster.drained_nodes + cluster.hot_buffer_drained_nodes + new_found_instances
     )
-    drained_nodes, buffer_drained_nodes, _ = sort_drained_nodes(
+    drained_nodes, hot_buffer_drained_nodes, _ = sort_drained_nodes(
         app_settings, all_drained_nodes, allowed_instance_types
     )
     return dataclasses.replace(
         cluster,
         drained_nodes=drained_nodes,
-        buffer_drained_nodes=buffer_drained_nodes,
+        hot_buffer_drained_nodes=hot_buffer_drained_nodes,
         pending_ec2s=still_pending_ec2s,
     )
 
@@ -417,7 +417,7 @@ async def _activate_drained_nodes(
         cluster,
         active_nodes=cluster.active_nodes + activated_nodes,
         drained_nodes=remaining_drained_nodes,
-        buffer_drained_nodes=remaining_reserved_drained_nodes,
+        hot_buffer_drained_nodes=remaining_reserved_drained_nodes,
     )
 
 
@@ -483,7 +483,7 @@ async def _start_warm_buffer_instances(
 
     return dataclasses.replace(
         cluster,
-        buffer_ec2s=[
+        warm_buffer_ec2s=[
             i
             for i in cluster.warm_buffer_ec2s
             if i.ec2_instance.id not in started_instance_ids

services/autoscaling/src/simcore_service_autoscaling/modules/instrumentation/_constants.py

@@ -20,7 +20,7 @@
         "Number of EC2-backed docker nodes which are drained",
         EC2_INSTANCE_LABELS,
     ),
-    "buffer_drained_nodes": (
+    "hot_buffer_drained_nodes": (
         "Number of EC2-backed docker nodes which are drained and in buffer/reserve",
         EC2_INSTANCE_LABELS,
     ),
@@ -32,7 +32,7 @@
         "Number of EC2 instances that failed joining the cluster",
         EC2_INSTANCE_LABELS,
     ),
-    "buffer_ec2s": (
+    "warm_buffer_ec2s": (
         "Number of buffer EC2 instances prepared, stopped, and ready to be activated",
         EC2_INSTANCE_LABELS,
     ),
@@ -54,7 +54,7 @@
     ),
 }
 
-BUFFER_POOLS_METRICS_DEFINITIONS: Final[dict[str, tuple[str, tuple[str, ...]]]] = {
+WARM_BUFFER_POOLS_METRICS_DEFINITIONS: Final[dict[str, tuple[str, tuple[str, ...]]]] = {
     "ready_instances": (
         "Number of EC2 buffer instances that are ready for use",
         EC2_INSTANCE_LABELS,

services/autoscaling/src/simcore_service_autoscaling/modules/instrumentation/_models.py

@@ -6,10 +6,10 @@
 
 from ...models import Cluster, WarmBufferPoolManager
 from ._constants import (
-    BUFFER_POOLS_METRICS_DEFINITIONS,
     CLUSTER_METRICS_DEFINITIONS,
     EC2_INSTANCE_LABELS,
     METRICS_NAMESPACE,
+    WARM_BUFFER_POOLS_METRICS_DEFINITIONS,
 )
 from ._utils import TrackedGauge, create_gauge
 
@@ -19,10 +19,10 @@ class ClusterMetrics(MetricsBase):  # pylint: disable=too-many-instance-attribut
     active_nodes: TrackedGauge = field(init=False)
     pending_nodes: TrackedGauge = field(init=False)
     drained_nodes: TrackedGauge = field(init=False)
-    buffer_drained_nodes: TrackedGauge = field(init=False)
+    hot_buffer_drained_nodes: TrackedGauge = field(init=False)
     pending_ec2s: TrackedGauge = field(init=False)
     broken_ec2s: TrackedGauge = field(init=False)
-    buffer_ec2s: TrackedGauge = field(init=False)
+    warm_buffer_ec2s: TrackedGauge = field(init=False)
     disconnected_nodes: TrackedGauge = field(init=False)
     terminating_nodes: TrackedGauge = field(init=False)
     retired_nodes: TrackedGauge = field(init=False)
@@ -124,7 +124,7 @@ class BufferPoolsMetrics(MetricsBase):
 
     def __post_init__(self) -> None:
         buffer_pools_subsystem = f"{self.subsystem}_buffer_machines_pools"
-        for field_name, definition in BUFFER_POOLS_METRICS_DEFINITIONS.items():
+        for field_name, definition in WARM_BUFFER_POOLS_METRICS_DEFINITIONS.items():
             setattr(
                 self,
                 field_name,
@@ -169,7 +169,7 @@ def update_from_buffer_pool_manager(
     ) -> None:
         flat_pool = buffer_pool_manager.flatten_buffer_pool()
 
-        for field_name in BUFFER_POOLS_METRICS_DEFINITIONS:
+        for field_name in WARM_BUFFER_POOLS_METRICS_DEFINITIONS:
             tracked_gauge = getattr(self, field_name)
             assert isinstance(tracked_gauge, TrackedGauge)  # nosec
             instances = getattr(flat_pool, field_name)

services/autoscaling/tests/unit/conftest.py

@@ -849,10 +849,10 @@ def _creator(**cluter_overrides) -> Cluster:
                 active_nodes=[],
                 pending_nodes=[],
                 drained_nodes=[],
-                buffer_drained_nodes=[],
+                hot_buffer_drained_nodes=[],
                 pending_ec2s=[],
                 broken_ec2s=[],
-                buffer_ec2s=[],
+                warm_buffer_ec2s=[],
                 disconnected_nodes=[],
                 terminating_nodes=[],
                 retired_nodes=[],

services/autoscaling/tests/unit/test_modules_cluster_scaling_dynamic.py

@@ -1718,7 +1718,7 @@ async def test__find_terminateable_nodes_with_no_hosts(
             AssociatedInstance(node=host_node, ec2_instance=fake_ec2_instance_data())
         ],
         drained_nodes=[],
-        buffer_drained_nodes=[
+        hot_buffer_drained_nodes=[
             AssociatedInstance(node=host_node, ec2_instance=fake_ec2_instance_data())
         ],
     )
@@ -1752,7 +1752,7 @@ async def test__try_scale_down_cluster_with_no_nodes(
         drained_nodes=[
             create_associated_instance(drained_host_node, False)  # noqa: FBT003
         ],
-        buffer_drained_nodes=[
+        hot_buffer_drained_nodes=[
             create_associated_instance(drained_host_node, True)  # noqa: FBT003
         ],
     )
@@ -1793,7 +1793,7 @@ async def test__activate_drained_nodes_with_no_tasks(
         drained_nodes=[
             create_associated_instance(drained_host_node, True)  # noqa: FBT003
         ],
-        buffer_drained_nodes=[
+        hot_buffer_drained_nodes=[
             create_associated_instance(drained_host_node, True)  # noqa: FBT003
         ],
     )
@@ -2014,8 +2014,8 @@ async def test_warm_buffers_are_started_to_replace_missing_hot_buffers(
         expected_num_machines=0,
     )
     assert not analyzed_cluster.active_nodes
-    assert analyzed_cluster.buffer_ec2s
-    assert len(analyzed_cluster.buffer_ec2s) == len(buffer_machines)
+    assert analyzed_cluster.warm_buffer_ec2s
+    assert len(analyzed_cluster.warm_buffer_ec2s) == len(buffer_machines)
 
     # now we should have a warm buffer moved to the hot buffer
     await assert_autoscaled_dynamic_ec2_instances(
@@ -2053,14 +2053,14 @@ async def test_warm_buffers_are_started_to_replace_missing_hot_buffers(
         expected_num_machines=app_settings.AUTOSCALING_EC2_INSTANCES.EC2_INSTANCES_MACHINES_BUFFER,
     )
     assert not analyzed_cluster.active_nodes
-    assert len(analyzed_cluster.buffer_ec2s) == max(
+    assert len(analyzed_cluster.warm_buffer_ec2s) == max(
         0,
         buffer_count
         - app_settings.AUTOSCALING_EC2_INSTANCES.EC2_INSTANCES_MACHINES_BUFFER,
     ), (
         "the warm buffers were not used as expected there should be"
         f" {buffer_count - app_settings.AUTOSCALING_EC2_INSTANCES.EC2_INSTANCES_MACHINES_BUFFER} remaining, "
-        f"found {len(analyzed_cluster.buffer_ec2s)}"
+        f"found {len(analyzed_cluster.warm_buffer_ec2s)}"
     )
     assert (
         len(analyzed_cluster.pending_ec2s)
@@ -2189,8 +2189,8 @@ async def test_warm_buffers_only_replace_hot_buffer_if_service_is_started_issue7
     spied_cluster = assert_cluster_state(
         spied_cluster_analysis, expected_calls=1, expected_num_machines=5
     )
-    assert len(spied_cluster.buffer_drained_nodes) == num_hot_buffer
-    assert not spied_cluster.buffer_ec2s
+    assert len(spied_cluster.hot_buffer_drained_nodes) == num_hot_buffer
+    assert not spied_cluster.warm_buffer_ec2s
 
     # have a few warm buffers ready with the same type as the hot buffer machines
     await create_buffer_machines(
@@ -2234,8 +2234,8 @@ async def test_warm_buffers_only_replace_hot_buffer_if_service_is_started_issue7
     spied_cluster = assert_cluster_state(
         spied_cluster_analysis, expected_calls=1, expected_num_machines=5
     )
-    assert len(spied_cluster.buffer_drained_nodes) == num_hot_buffer
-    assert len(spied_cluster.buffer_ec2s) == buffer_count
+    assert len(spied_cluster.hot_buffer_drained_nodes) == num_hot_buffer
+    assert len(spied_cluster.warm_buffer_ec2s) == buffer_count
 
     #
     # BUG REPRODUCTION
@@ -2303,8 +2303,8 @@ async def test_warm_buffers_only_replace_hot_buffer_if_service_is_started_issue7
     spied_cluster = assert_cluster_state(
         spied_cluster_analysis, expected_calls=2, expected_num_machines=6
     )
-    assert len(spied_cluster.buffer_drained_nodes) == num_hot_buffer - 1
-    assert len(spied_cluster.buffer_ec2s) == buffer_count - 1
+    assert len(spied_cluster.hot_buffer_drained_nodes) == num_hot_buffer - 1
+    assert len(spied_cluster.warm_buffer_ec2s) == buffer_count - 1
     assert len(spied_cluster.active_nodes) == 1
     assert len(spied_cluster.pending_ec2s) == 1
 
@@ -2322,8 +2322,8 @@ async def _check_autoscaling_is_stable() -> None:
         spied_cluster = assert_cluster_state(
             spied_cluster_analysis, expected_calls=1, expected_num_machines=6
         )
-        assert len(spied_cluster.buffer_drained_nodes) == num_hot_buffer - 1
-        assert len(spied_cluster.buffer_ec2s) == buffer_count - 1
+        assert len(spied_cluster.hot_buffer_drained_nodes) == num_hot_buffer - 1
+        assert len(spied_cluster.warm_buffer_ec2s) == buffer_count - 1
         assert len(spied_cluster.active_nodes) == 1
         assert len(spied_cluster.pending_ec2s) == 1