Backport of Spatel/net 1646 add max ejection percent and base ejectio…

…n time into release/1.0.x (#2106) * backport of commit 978c3fa * go mod tidy and update rand func --------- Co-authored-by: Semir Patel <semir.patel@hashicorp.com> Co-authored-by: Maliz <maliheh.monshizadeh@hashicorp.com>
hashicorp · May 9, 2023 · 2bfc943 · 2bfc943
1 parent 35a3a33
commit 2bfc943
Show file tree

Hide file tree

Showing 8 changed files with 94 additions and 19 deletions.
diff --git a/charts/consul/templates/crd-servicedefaults.yaml b/charts/consul/templates/crd-servicedefaults.yaml
@@ -351,6 +351,12 @@ spec:
                             how upstream proxy instances will be monitored for removal
                             from the load balancing pool.
                           properties:
+                            baseEjectionTime:
+                              description: The base time that a host is ejected for. The
+                                real time is equal to the base time multiplied by the number
+                                of times the host has been ejected and is capped by
+                                max_ejection_time (Default 300s). Defaults to 30s.
+                              type: string
                             enforcing_consecutive_5xx:
                               description: EnforcingConsecutive5xx is the % chance
                                 that a host will be actually ejected when an outlier
@@ -363,6 +369,12 @@ spec:
                                 sweeps. Each sweep may remove hosts or return hosts
                                 to the pool.
                               type: string
+                            maxEjectionPercent:
+                              description: The maximum % of an upstream cluster that
+                                can be ejected due to outlier detection. Defaults to
+                                10% but will eject at least one host regardless of the value.
+                              format: int32
+                              type: integer
                             maxFailures:
                               description: MaxFailures is the count of consecutive
                                 failures that results in a host being removed from

diff --git a/control-plane/api/v1alpha1/servicedefaults_types.go b/control-plane/api/v1alpha1/servicedefaults_types.go
@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"net"
 	"strings"
+	"time"
 
 	"github.com/google/go-cmp/cmp"
 	"github.com/google/go-cmp/cmp/cmpopts"
@@ -172,6 +173,13 @@ type PassiveHealthCheck struct {
 	// when an outlier status is detected through consecutive 5xx.
 	// This setting can be used to disable ejection or to ramp it up slowly.
 	EnforcingConsecutive5xx *uint32 `json:"enforcing_consecutive_5xx,omitempty"`
+	// The maximum % of an upstream cluster that can be ejected due to outlier detection.
+	// Defaults to 10% but will eject at least one host regardless of the value.
+	MaxEjectionPercent *uint32 `json:",omitempty" alias:"max_ejection_percent"`
+	// The base time that a host is ejected for. The real time is equal to the base time
+	// multiplied by the number of times the host has been ejected and is capped by
+	// max_ejection_time (Default 300s). Defaults to 30000ms or 30s.
+	BaseEjectionTime *time.Duration `json:",omitempty" alias:"base_ejection_time"`
 }
 
 type ServiceDefaultsDestination struct {
@@ -413,6 +421,8 @@ func (in *PassiveHealthCheck) toConsul() *capi.PassiveHealthCheck {
 		Interval:                in.Interval.Duration,
 		MaxFailures:             in.MaxFailures,
 		EnforcingConsecutive5xx: in.EnforcingConsecutive5xx,
+		MaxEjectionPercent:      in.MaxEjectionPercent,
+		BaseEjectionTime:        in.BaseEjectionTime,
 	}
 }
 

diff --git a/control-plane/api/v1alpha1/servicedefaults_types_test.go b/control-plane/api/v1alpha1/servicedefaults_types_test.go
@@ -85,6 +85,8 @@ func TestServiceDefaults_ToConsul(t *testing.T) {
 								},
 								MaxFailures:             uint32(20),
 								EnforcingConsecutive5xx: pointer.Uint32(100),
+								MaxEjectionPercent:      pointer.Uint32(10),
+								BaseEjectionTime:        pointer.Duration(10 * time.Second),
 							},
 							MeshGateway: MeshGateway{
 								Mode: "local",
@@ -110,6 +112,8 @@ func TestServiceDefaults_ToConsul(t *testing.T) {
 									},
 									MaxFailures:             uint32(10),
 									EnforcingConsecutive5xx: pointer.Uint32(60),
+									MaxEjectionPercent:      pointer.Uint32(20),
+									BaseEjectionTime:        pointer.Duration(20 * time.Second),
 								},
 								MeshGateway: MeshGateway{
 									Mode: "remote",
@@ -134,6 +138,8 @@ func TestServiceDefaults_ToConsul(t *testing.T) {
 									},
 									MaxFailures:             uint32(10),
 									EnforcingConsecutive5xx: pointer.Uint32(60),
+									MaxEjectionPercent:      pointer.Uint32(30),
+									BaseEjectionTime:        pointer.Duration(30 * time.Second),
 								},
 								MeshGateway: MeshGateway{
 									Mode: "remote",
@@ -197,6 +203,8 @@ func TestServiceDefaults_ToConsul(t *testing.T) {
 							Interval:                2 * time.Second,
 							MaxFailures:             uint32(20),
 							EnforcingConsecutive5xx: pointer.Uint32(100),
+							MaxEjectionPercent:      pointer.Uint32(10),
+							BaseEjectionTime:        pointer.Duration(10 * time.Second),
 						},
 						MeshGateway: capi.MeshGatewayConfig{
 							Mode: "local",
@@ -220,6 +228,8 @@ func TestServiceDefaults_ToConsul(t *testing.T) {
 								Interval:                2 * time.Second,
 								MaxFailures:             uint32(10),
 								EnforcingConsecutive5xx: pointer.Uint32(60),
+								MaxEjectionPercent:      pointer.Uint32(20),
+								BaseEjectionTime:        pointer.Duration(20 * time.Second),
 							},
 							MeshGateway: capi.MeshGatewayConfig{
 								Mode: "remote",
@@ -242,6 +252,8 @@ func TestServiceDefaults_ToConsul(t *testing.T) {
 								Interval:                2 * time.Second,
 								MaxFailures:             uint32(10),
 								EnforcingConsecutive5xx: pointer.Uint32(60),
+								MaxEjectionPercent:      pointer.Uint32(30),
+								BaseEjectionTime:        pointer.Duration(30 * time.Second),
 							},
 							MeshGateway: capi.MeshGatewayConfig{
 								Mode: "remote",
@@ -348,6 +360,8 @@ func TestServiceDefaults_MatchesConsul(t *testing.T) {
 								},
 								MaxFailures:             uint32(20),
 								EnforcingConsecutive5xx: pointer.Uint32(100),
+								MaxEjectionPercent:      pointer.Uint32(10),
+								BaseEjectionTime:        pointer.Duration(10 * time.Second),
 							},
 							MeshGateway: MeshGateway{
 								Mode: "local",
@@ -372,6 +386,8 @@ func TestServiceDefaults_MatchesConsul(t *testing.T) {
 									},
 									MaxFailures:             uint32(10),
 									EnforcingConsecutive5xx: pointer.Uint32(60),
+									MaxEjectionPercent:      pointer.Uint32(20),
+									BaseEjectionTime:        pointer.Duration(20 * time.Second),
 								},
 								MeshGateway: MeshGateway{
 									Mode: "remote",
@@ -395,6 +411,8 @@ func TestServiceDefaults_MatchesConsul(t *testing.T) {
 									},
 									MaxFailures:             uint32(10),
 									EnforcingConsecutive5xx: pointer.Uint32(60),
+									MaxEjectionPercent:      pointer.Uint32(30),
+									BaseEjectionTime:        pointer.Duration(30 * time.Second),
 								},
 								MeshGateway: MeshGateway{
 									Mode: "remote",
@@ -453,6 +471,8 @@ func TestServiceDefaults_MatchesConsul(t *testing.T) {
 							Interval:                2 * time.Second,
 							MaxFailures:             uint32(20),
 							EnforcingConsecutive5xx: pointer.Uint32(100),
+							MaxEjectionPercent:      pointer.Uint32(10),
+							BaseEjectionTime:        pointer.Duration(10 * time.Second),
 						},
 						MeshGateway: capi.MeshGatewayConfig{
 							Mode: "local",
@@ -475,6 +495,8 @@ func TestServiceDefaults_MatchesConsul(t *testing.T) {
 								Interval:                2 * time.Second,
 								MaxFailures:             uint32(10),
 								EnforcingConsecutive5xx: pointer.Uint32(60),
+								MaxEjectionPercent:      pointer.Uint32(20),
+								BaseEjectionTime:        pointer.Duration(20 * time.Second),
 							},
 							MeshGateway: capi.MeshGatewayConfig{
 								Mode: "remote",
@@ -496,6 +518,8 @@ func TestServiceDefaults_MatchesConsul(t *testing.T) {
 								Interval:                2 * time.Second,
 								MaxFailures:             uint32(10),
 								EnforcingConsecutive5xx: pointer.Uint32(60),
+								MaxEjectionPercent:      pointer.Uint32(30),
+								BaseEjectionTime:        pointer.Duration(30 * time.Second),
 							},
 							MeshGateway: capi.MeshGatewayConfig{
 								Mode: "remote",

diff --git a/control-plane/api/v1alpha1/zz_generated.deepcopy.go b/control-plane/api/v1alpha1/zz_generated.deepcopy.go
diff --git a/control-plane/config/crd/bases/consul.hashicorp.com_servicedefaults.yaml b/control-plane/config/crd/bases/consul.hashicorp.com_servicedefaults.yaml
@@ -245,6 +245,12 @@ spec:
                           upstream proxy instances will be monitored for removal from
                           the load balancing pool.
                         properties:
+                          baseEjectionTime:
+                            description: The base time that a host is ejected for. The
+                              real time is equal to the base time multiplied by the number
+                              of times the host has been ejected and is capped by
+                              max_ejection_time (Default 300s). Defaults to 30s.
+                            type: string
                           enforcing_consecutive_5xx:
                             description: EnforcingConsecutive5xx is the % chance that
                               a host will be actually ejected when an outlier status
@@ -256,6 +262,12 @@ spec:
                             description: Interval between health check analysis sweeps.
                               Each sweep may remove hosts or return hosts to the pool.
                             type: string
+                          maxEjectionPercent:
+                            description: The maximum % of an upstream cluster that
+                              can be ejected due to outlier detection. Defaults to
+                              10% but will eject at least one host regardless of the value.
+                            format: int32
+                            type: integer
                           maxFailures:
                             description: MaxFailures is the count of consecutive failures
                               that results in a host being removed from the pool.
@@ -344,6 +356,12 @@ spec:
                             how upstream proxy instances will be monitored for removal
                             from the load balancing pool.
                           properties:
+                            baseEjectionTime:
+                              description: The base time that a host is ejected for. The
+                                real time is equal to the base time multiplied by the number
+                                of times the host has been ejected and is capped by
+                                max_ejection_time (Default 300s). Defaults to 30s.
+                              type: string
                             enforcing_consecutive_5xx:
                               description: EnforcingConsecutive5xx is the % chance
                                 that a host will be actually ejected when an outlier
@@ -356,6 +374,12 @@ spec:
                                 sweeps. Each sweep may remove hosts or return hosts
                                 to the pool.
                               type: string
+                            maxEjectionPercent:
+                              description: The maximum % of an upstream cluster that
+                                can be ejected due to outlier detection. Defaults to
+                                10% but will eject at least one host regardless of the value.
+                              format: int32
+                              type: integer
                             maxFailures:
                               description: MaxFailures is the count of consecutive
                                 failures that results in a host being removed from

diff --git a/control-plane/go.mod b/control-plane/go.mod
@@ -6,12 +6,12 @@ require (
 	github.com/deckarep/golang-set v1.7.1
 	github.com/fsnotify/fsnotify v1.5.4
 	github.com/go-logr/logr v0.4.0
-	github.com/google/go-cmp v0.5.7
+	github.com/google/go-cmp v0.5.8
 	github.com/google/shlex v0.0.0-20191202100458-e7afc7fbc510
 	github.com/hashicorp/consul-k8s/control-plane/cni v0.0.0-20220831174802-b8af65262de8
 	github.com/hashicorp/consul-server-connection-manager v0.1.0
-	github.com/hashicorp/consul/api v1.17.0
-	github.com/hashicorp/consul/sdk v0.13.0
+	github.com/hashicorp/consul/api v1.10.1-0.20230426225948-5eaeb7b8e563
+	github.com/hashicorp/consul/sdk v0.13.1
 	github.com/hashicorp/go-discover v0.0.0-20200812215701-c4b85f6ed31f
 	github.com/hashicorp/go-hclog v1.2.2
 	github.com/hashicorp/go-multierror v1.1.1
@@ -112,13 +112,12 @@ require (
 	go.uber.org/atomic v1.7.0 // indirect
 	go.uber.org/multierr v1.6.0 // indirect
 	golang.org/x/crypto v0.0.0-20220411220226-7b82a4e95df4 // indirect
+	golang.org/x/exp v0.0.0-20230321023759-10a507213a29 // indirect
 	golang.org/x/net v0.7.0 // indirect
 	golang.org/x/oauth2 v0.0.0-20211104180415-d3ed0bb246c8 // indirect
 	golang.org/x/sync v0.0.0-20210220032951-036812b2e83c // indirect
 	golang.org/x/sys v0.5.0 // indirect
 	golang.org/x/term v0.5.0 // indirect
-	golang.org/x/tools v0.2.0 // indirect
-	golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 // indirect
 	google.golang.org/api v0.43.0 // indirect
 	google.golang.org/appengine v1.6.7 // indirect
 	google.golang.org/genproto v0.0.0-20210602131652-f16073e35f0c // indirect