plugin: Remove ComputeUnit from config and PluginResponse (#743)

... and every where else that it filters into.

This is part of the process of moving the ComputeUnit source of truth
from the scheduler to the autoscaler-agent (#706).
We've already redefined the autoscaler-agent's own config as its source
of truth, so this is mostly just following through on the previous
deprecation.

deploy/scheduler/config_map.yaml

@@ -25,7 +25,6 @@ metadata:
 data:
   autoscale-enforcer-config.json: |
     {
-      "computeUnit": { "vCPUs": 0.25, "mem": "1Gi" },
       "nodeConfig": {
         "cpu": { "watermark": 0.9 },
         "memory": { "watermark": 0.9 },

pkg/agent/core/state_test.go

@@ -136,9 +136,8 @@ func Test_DesiredResourcesFromMetricsOrRequestedUpscaling(t *testing.T) {
 			// set lastApproved by simulating a scheduler request/response
 			state.Plugin().StartingRequest(now, c.schedulerApproved)
 			err := state.Plugin().RequestSuccessful(now, api.PluginResponse{
-				Permit:      c.schedulerApproved,
-				Migrate:     nil,
-				ComputeUnit: nil,
+				Permit:  c.schedulerApproved,
+				Migrate: nil,
 			})
 			if err != nil {
 				t.Errorf("state.Plugin().RequestSuccessful() failed: %s", err)
@@ -218,9 +217,8 @@ func doInitialPluginRequest(
 	a.Do(state.Plugin().StartingRequest, clock.Now(), resources)
 	clock.Inc(requestTime)
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resources,
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resources,
+		Migrate: nil,
 	})
 }
 
@@ -287,9 +285,8 @@ func TestBasicScaleUpAndDownFlow(t *testing.T) {
 	// should have nothing more to do; waiting on plugin request to come back
 	a.Call(nextActions).Equals(core.ActionSet{})
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(2),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(2),
+		Migrate: nil,
 	})
 
 	// Scheduler approval is done, now we should be making the request to NeonVM
@@ -396,9 +393,8 @@ func TestBasicScaleUpAndDownFlow(t *testing.T) {
 	// should have nothing more to do; waiting on plugin request to come back
 	a.Call(nextActions).Equals(core.ActionSet{})
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(1),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(1),
+		Migrate: nil,
 	})
 
 	// Finally, check there's no leftover actions:
@@ -460,9 +456,8 @@ func TestPeriodicPluginRequest(t *testing.T) {
 			clock.Inc(reqDuration)
 			a.Call(state.NextActions, clock.Now()).Equals(core.ActionSet{})
 			a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-				Permit:      resources,
-				Migrate:     nil,
-				ComputeUnit: nil,
+				Permit:  resources,
+				Migrate: nil,
 			})
 			clock.Inc(clockTick - reqDuration)
 		}
@@ -617,9 +612,8 @@ func TestDeniedDownscalingIncreaseAndRetry(t *testing.T) {
 	})
 	clockTick()
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(3),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(3),
+		Migrate: nil,
 	})
 	// ... And *now* there's nothing left to do but wait until downscale wait expires:
 	a.Call(nextActions).Equals(core.ActionSet{
@@ -661,9 +655,8 @@ func TestDeniedDownscalingIncreaseAndRetry(t *testing.T) {
 	})
 	clockTick()
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(3),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(3),
+		Migrate: nil,
 	})
 	a.Call(nextActions).Equals(core.ActionSet{
 		Wait: &core.ActionWait{Duration: duration("0.9s")}, // yep, still waiting on retrying vm-monitor downscaling
@@ -728,9 +721,8 @@ func TestDeniedDownscalingIncreaseAndRetry(t *testing.T) {
 	})
 	clockTick()
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(1),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(1),
+		Migrate: nil,
 	})
 	// And now there's truly nothing left to do. Back to waiting on plugin request tick :)
 	a.Call(nextActions).Equals(core.ActionSet{
@@ -795,9 +787,8 @@ func TestRequestedUpscale(t *testing.T) {
 		Wait: &core.ActionWait{Duration: duration("5.9s")}, // same waiting for requested upscale expiring
 	})
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(2),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(2),
+		Migrate: nil,
 	})
 
 	// After approval from the scheduler plugin, now need to make NeonVM request:
@@ -847,9 +838,8 @@ func TestRequestedUpscale(t *testing.T) {
 		Wait: &core.ActionWait{Duration: duration("0.9s")}, // waiting for requested upscale expiring
 	})
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(2),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(2),
+		Migrate: nil,
 	})
 
 	// Still should just be waiting on vm-monitor upscale expiring
@@ -992,9 +982,8 @@ func TestDownscalePivotBack(t *testing.T) {
 				*pluginWait = duration("4.9s") // reset because we just made a request
 				t.Log(" > finish plugin downscale")
 				a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-					Permit:      resForCU(1),
-					Migrate:     nil,
-					ComputeUnit: nil,
+					Permit:  resForCU(1),
+					Migrate: nil,
 				})
 			},
 			post: func(pluginWait *time.Duration) {
@@ -1011,9 +1000,8 @@ func TestDownscalePivotBack(t *testing.T) {
 				*pluginWait = duration("4.9s") // reset because we just made a request
 				t.Log(" > finish plugin upscale")
 				a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-					Permit:      resForCU(2),
-					Migrate:     nil,
-					ComputeUnit: nil,
+					Permit:  resForCU(2),
+					Migrate: nil,
 				})
 			},
 		},
@@ -1150,9 +1138,8 @@ func TestBoundsChangeRequiresDownsale(t *testing.T) {
 	a.Do(state.Plugin().StartingRequest, clock.Now(), resForCU(1))
 	clockTick()
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(1),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(1),
+		Migrate: nil,
 	})
 	// And then, we shouldn't need to do anything else:
 	a.Call(nextActions).Equals(core.ActionSet{
@@ -1222,9 +1209,8 @@ func TestBoundsChangeRequiresUpscale(t *testing.T) {
 	a.Do(state.Plugin().StartingRequest, clock.Now(), resForCU(3))
 	clockTick()
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(3),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(3),
+		Migrate: nil,
 	})
 	// Do NeonVM request for the upscaling
 	a.Call(nextActions).Equals(core.ActionSet{
@@ -1322,9 +1308,8 @@ func TestFailedRequestRetry(t *testing.T) {
 	a.Do(state.Plugin().StartingRequest, clock.Now(), resForCU(2))
 	clockTick()
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(2),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(2),
+		Migrate: nil,
 	})
 
 	// Now, after plugin request is successful, we should be making a request to NeonVM.
@@ -1407,9 +1392,8 @@ func TestMetricsConcurrentUpdatedDuringDownscale(t *testing.T) {
 	a.Do(state.Plugin().StartingRequest, clock.Now(), resForCU(3))
 	clockTick()
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(3),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(3),
+		Migrate: nil,
 	})
 
 	clockTick()
@@ -1502,9 +1486,8 @@ func TestMetricsConcurrentUpdatedDuringDownscale(t *testing.T) {
 	clockTick()
 
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(2),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(2),
+		Migrate: nil,
 	})
 	// Still waiting for NeonVM request to complete
 	a.Call(nextActions).Equals(core.ActionSet{
@@ -1528,9 +1511,8 @@ func TestMetricsConcurrentUpdatedDuringDownscale(t *testing.T) {
 	clockTick()
 
 	a.NoError(state.Plugin().RequestSuccessful, clock.Now(), api.PluginResponse{
-		Permit:      resForCU(1),
-		Migrate:     nil,
-		ComputeUnit: nil,
+		Permit:  resForCU(1),
+		Migrate: nil,
 	})
 	// Nothing left to do
 	a.Call(nextActions).Equals(core.ActionSet{

pkg/agent/runner.go

@@ -669,7 +669,7 @@ func (r *Runner) DoSchedulerRequest(
 	reqData := &api.AgentRequest{
 		ProtoVersion: PluginProtocolVersion,
 		Pod:          r.podName,
-		ComputeUnit:  &r.global.config.Scaling.ComputeUnit,
+		ComputeUnit:  r.global.config.Scaling.ComputeUnit,
 		Resources:    resources,
 		LastPermit:   lastPermit,
 		Metrics:      metrics,

pkg/api/VERSIONING.md

@@ -38,7 +38,7 @@ number.
 
 | Release | autoscaler-agent | Scheduler plugin |
 |---------|------------------|------------------|
-| _Current_ | v4.0 only | v1.0-v4.0 |
+| _Current_ | v4.0 only | v3.0-v4.0 |
 | v0.25.0 | v4.0 only | v1.0-v4.0 |
 | v0.24.0 | v4.0 only | v1.0-v4.0 |
 | v0.23.0 | **v4.0 only** | **v1.0-v4.0** |

pkg/api/types.go

@@ -170,7 +170,7 @@ type AgentRequest struct {
 	// If the requested resources are not a multiple of ComputeUnit, the scheduler plugin will make
 	// a best-effort attempt to return a value satisfying the request. Any approved increases will
 	// be a multiple of ComputeUnit, but otherwise the plugin does not check.
-	ComputeUnit *Resources `json:"computeUnit"`
+	ComputeUnit Resources `json:"computeUnit"`
 	// Resources gives a requested or notified change in resources allocated to the VM.
 	//
 	// The requested amount MAY be equal to the current amount, in which case it serves as a
@@ -390,17 +390,6 @@ type PluginResponse struct {
 	// Migrate, if present, notifies the autoscaler-agent that its VM will be migrated away,
 	// alongside whatever other information may be useful.
 	Migrate *MigrateResponse `json:"migrate,omitempty"`
-
-	// ComputeUnit is the minimum unit of resources that the scheduler is expecting to work with
-	//
-	// For example, if ComputeUnit is Resources{ VCPU: 2, Mem: 4Gi }, then all VMs are expected to
-	// have allocated vCPUs that are a multiple of two, with 2 GiB of memory per vCPU.
-	//
-	// This value may be different across nodes, and the scheduler expects that all AgentRequests
-	// will abide by the most recent ComputeUnit they've received.
-	//
-	// THIS FIELD IS DEPRECATED: See https://github.com/neondatabase/autoscaling/issues/706
-	ComputeUnit *Resources `json:"resourceUnit,omitempty"`
 }
 
 // MigrateResponse, when provided, is a notification to the autsocaler-agent that it will migrate

pkg/plugin/config.go

@@ -19,11 +19,6 @@ import (
 //////////////////
 
 type Config struct {
-	// ComputeUnit is the desired ratio between CPU and memory that autoscaler-agents should uphold
-	//
-	// This value is sent to autoscaler-agents in every response, as part of api.PluginResponse.
-	ComputeUnit api.Resources `json:"computeUnit"`
-
 	// NodeConfig defines our policies around node resources and scoring
 	NodeConfig nodeConfig `json:"nodeConfig"`
 
@@ -121,10 +116,6 @@ func (c *Config) migrationEnabled() bool {
 
 // if the returned error is not nil, the string is a JSON path to the invalid value
 func (c *Config) validate() (string, error) {
-	if err := c.ComputeUnit.ValidateNonZero(); err != nil {
-		return "computeUnit", err
-	}
-
 	if path, err := c.NodeConfig.validate(); err != nil {
 		return fmt.Sprintf("nodeConfig.%s", path), err
 	}

pkg/plugin/dumpstate.go

@@ -140,7 +140,6 @@ type podStateDump struct {
 type vmPodStateDump struct {
 	Name                     util.NamespacedName    `json:"name"`
 	TestingOnlyAlwaysMigrate bool                   `json:"testingOnlyAlwaysMigrate"`
-	MostRecentComputeUnit    *api.Resources         `json:"mostRecentComputeUnit"`
 	Metrics                  *api.Metrics           `json:"metrics"`
 	MqIndex                  int                    `json:"mqIndex"`
 	MigrationState           *podMigrationStateDump `json:"migrationState"`
@@ -246,11 +245,6 @@ func (s *podState) dump() podStateDump {
 
 func (s *vmPodState) dump() vmPodStateDump {
 	// Copy some of the "may be nil" pointer fields
-	var mostRecentComputeUnit *api.Resources
-	if s.mostRecentComputeUnit != nil {
-		mrcu := *s.mostRecentComputeUnit
-		mostRecentComputeUnit = &mrcu
-	}
 	var metrics *api.Metrics
 	if s.metrics != nil {
 		m := *s.metrics
@@ -266,7 +260,6 @@ func (s *vmPodState) dump() vmPodStateDump {
 	return vmPodStateDump{
 		Name:                     s.name,
 		TestingOnlyAlwaysMigrate: s.testingOnlyAlwaysMigrate,
-		MostRecentComputeUnit:    mostRecentComputeUnit,
 		Metrics:                  metrics,
 		MqIndex:                  s.mqIndex,
 		MigrationState:           migrationState,