feat: gcp spot instances

Remove node kubernetes resource on spot node eviction event.

Signed-off-by: Serge Logvinov <serge.logvinov@sinextra.dev>

cmd/talos-cloud-controller-manager/main.go

@@ -62,6 +62,7 @@ func main() {
 		Constructor: nodeIpamController.startNodeIpamControllerWrapper,
 	}
 
+	app.ControllersDisabledByDefault.Insert(kcmnames.NodeLifecycleController)
 	app.ControllersDisabledByDefault.Insert(kcmnames.NodeIpamController)
 	controllerAliases["nodeipam"] = kcmnames.NodeIpamController
 	command := app.NewCloudControllerManagerCommand(ccmOptions, cloudInitializer, controllerInitializers, controllerAliases, fss, wait.NeverStop)

go.mod

@@ -1,9 +1,9 @@
 module github.com/siderolabs/talos-cloud-controller-manager
 
-go 1.22.4
+go 1.22.5
 
 require (
-	github.com/cosi-project/runtime v0.5.0
+	github.com/cosi-project/runtime v0.5.2
 	github.com/siderolabs/go-retry v0.3.3
 	github.com/siderolabs/net v0.4.0
 	github.com/siderolabs/talos/pkg/machinery v1.7.5
@@ -18,7 +18,7 @@ require (
 	k8s.io/component-base v0.30.2
 	k8s.io/controller-manager v0.30.2
 	k8s.io/klog/v2 v2.130.1
-	k8s.io/utils v0.0.0-20240502163921-fe8a2dddb1d0
+	k8s.io/utils v0.0.0-20240711033017-18e509b52bc8
 )
 
 require (
@@ -122,7 +122,7 @@ require (
 	golang.org/x/time v0.5.0 // indirect
 	google.golang.org/genproto/googleapis/api v0.0.0-20240624140628-dc46fd24d27d // indirect
 	google.golang.org/genproto/googleapis/rpc v0.0.0-20240624140628-dc46fd24d27d // indirect
-	google.golang.org/grpc v1.64.0 // indirect
+	google.golang.org/grpc v1.65.0 // indirect
 	google.golang.org/protobuf v1.34.2 // indirect
 	gopkg.in/inf.v0 v0.9.1 // indirect
 	gopkg.in/natefinch/lumberjack.v2 v2.2.1 // indirect

go.sum

@@ -1,4 +1,4 @@
-cloud.google.com/go/compute v1.25.1 h1:ZRpHJedLtTpKgr3RV1Fx23NuaAEN1Zfx9hw1u4aJdjU=
+cloud.google.com/go/compute v1.18.0 h1:FEigFqoDbys2cvFkZ9Fjq4gnHBP55anJ0yQyau2f9oY=
 cloud.google.com/go/compute/metadata v0.3.0 h1:Tz+eQXMEqDIKRsmY3cHTL6FVaynIjX2QxYC4trgAKZc=
 cloud.google.com/go/compute/metadata v0.3.0/go.mod h1:zFmK7XCadkQkj6TtorcaGlCW1hT1fIilQDwofLpJ20k=
 github.com/Azure/go-ansiterm v0.0.0-20230124172434-306776ec8161 h1:L/gRVlceqvL25UVaW/CKtUDjefjrs0SPonmDGUVOYP0=
@@ -34,8 +34,8 @@ github.com/cilium/ebpf v0.12.3/go.mod h1:TctK1ivibvI3znr66ljgi4hqOT8EYQjz1KWBfb1
 github.com/cloudflare/circl v1.3.3/go.mod h1:5XYMA4rFBvNIrhs50XuiBJ15vF2pZn4nnUKZrLbUZFA=
 github.com/cloudflare/circl v1.3.9 h1:QFrlgFYf2Qpi8bSpVPK1HBvWpx16v/1TZivyo7pGuBE=
 github.com/cloudflare/circl v1.3.9/go.mod h1:PDRU+oXvdD7KCtgKxW95M5Z8BpSCJXQORiZFnBQS5QU=
-github.com/cncf/xds/go v0.0.0-20240318125728-8a4994d93e50 h1:DBmgJDC9dTfkVyGgipamEh2BpGYxScCH1TOF1LL1cXc=
+github.com/cncf/xds/go v0.0.0-20240423153145-555b57ec207b h1:ga8SEFjZ60pxLcmhnThWgvH2wg8376yUJmPhEH4H3kw=
-github.com/cncf/xds/go v0.0.0-20240318125728-8a4994d93e50/go.mod h1:5e1+Vvlzido69INQaVO6d87Qn543Xr6nooe9Kz7oBFM=
+github.com/cncf/xds/go v0.0.0-20240423153145-555b57ec207b/go.mod h1:W+zGtBO5Y1IgJhy4+A9GOqVhqLpfZi+vwmdNXUehLA8=
 github.com/containerd/go-cni v1.1.10 h1:c2U73nld7spSWfiJwSh/8W9DK+/qQwYM2rngIhCyhyg=
 github.com/containerd/go-cni v1.1.10/go.mod h1:/Y/sL8yqYQn1ZG1om1OncJB1W4zN3YmjfP/ShCzG/OY=
 github.com/containernetworking/cni v1.2.2 h1:9IbP6KJQQxVKo4hhnm8r50YcVKrJbJu3Dqw+Rbt1vYk=
@@ -44,8 +44,10 @@ github.com/coreos/go-semver v0.3.1 h1:yi21YpKnrx1gt5R+la8n5WgS0kCrsPp33dmEyHReZr
 github.com/coreos/go-semver v0.3.1/go.mod h1:irMmmIw/7yzSRPWryHsK7EYSg09caPQL03VsM8rvUec=
 github.com/coreos/go-systemd/v22 v22.5.0 h1:RrqgGjYQKalulkV8NGVIfkXQf6YYmOyiJKk8iXXhfZs=
 github.com/coreos/go-systemd/v22 v22.5.0/go.mod h1:Y58oyj3AT4RCenI/lSvhwexgC+NSVTIJ3seZv2GcEnc=
-github.com/cosi-project/runtime v0.5.0 h1:Thin3w4ZQbRgHIv2h6a+zNGvp7N85eO8VEz0Kp859Jo=
+github.com/cosi-project/runtime v0.5.1 h1:07r4lk8sgiyhLdRuqZidB6qV3jFpKYhGccWdhTYHYcc=
-github.com/cosi-project/runtime v0.5.0/go.mod h1:GH/EZSmCeMpOFNeZYZnlmdB14mj0zLyREA4PqOI0ubg=
+github.com/cosi-project/runtime v0.5.1/go.mod h1:m+bkfUzKYeUyoqYAQBxdce3bfgncG8BsqcbfKRbvJKs=
+github.com/cosi-project/runtime v0.5.2 h1:3UlNiXF1/jexLYu2grd8jdTR8/FrACOUdN1mxRnCKZs=
+github.com/cosi-project/runtime v0.5.2/go.mod h1:m+bkfUzKYeUyoqYAQBxdce3bfgncG8BsqcbfKRbvJKs=
 github.com/cpuguy83/go-md2man/v2 v2.0.3/go.mod h1:tgQtvFlXSQOSOSIRvRPT7W67SCa46tRHOmNcaadrF8o=
 github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ33E=
 github.com/creack/pty v1.1.18 h1:n56/Zwd5o6whRC5PMGretI4IdRLlmBXYNjScPaBgsbY=
@@ -372,8 +374,8 @@ google.golang.org/genproto/googleapis/api v0.0.0-20240624140628-dc46fd24d27d h1:
 google.golang.org/genproto/googleapis/api v0.0.0-20240624140628-dc46fd24d27d/go.mod h1:Od4k8V1LQSizPRUK4OzZ7TBE/20k+jPczUDAEyvn69Y=
 google.golang.org/genproto/googleapis/rpc v0.0.0-20240624140628-dc46fd24d27d h1:k3zyW3BYYR30e8v3x0bTDdE9vpYFjZHK+HcyqkrppWk=
 google.golang.org/genproto/googleapis/rpc v0.0.0-20240624140628-dc46fd24d27d/go.mod h1:Ue6ibwXGpU+dqIcODieyLOcgj7z8+IcskoNIgZxtrFY=
-google.golang.org/grpc v1.64.0 h1:KH3VH9y/MgNQg1dE7b3XfVK0GsPSIzJwdF617gUSbvY=
+google.golang.org/grpc v1.65.0 h1:bs/cUb4lp1G5iImFFd3u5ixQzweKizoZJAwBNLR42lc=
-google.golang.org/grpc v1.64.0/go.mod h1:oxjF8E3FBnjp+/gVFYdWacaLDx9na1aqy9oovLpxQYg=
+google.golang.org/grpc v1.65.0/go.mod h1:WgYC2ypjlB0EiQi6wdKixMqukr6lBc0Vo+oOgjrM5ZQ=
 google.golang.org/protobuf v1.34.2 h1:6xV6lTsCfpGD21XK49h7MhtcApnLqkfYgPcdHftf6hg=
 google.golang.org/protobuf v1.34.2/go.mod h1:qYOHts0dSfpeUzUFpOMr/WGzszTmLH+DiWniOlNbLDw=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
@@ -415,6 +417,8 @@ k8s.io/kube-openapi v0.0.0-20240620174524-b456828f718b h1:Q9xmGWBvOGd8UJyccgpYlL
 k8s.io/kube-openapi v0.0.0-20240620174524-b456828f718b/go.mod h1:UxDHUPsUwTOOxSU+oXURfFBcAS6JwiRXTYqYwfuGowc=
 k8s.io/utils v0.0.0-20240502163921-fe8a2dddb1d0 h1:jgGTlFYnhF1PM1Ax/lAlxUPE+KfCIXHaathvJg1C3ak=
 k8s.io/utils v0.0.0-20240502163921-fe8a2dddb1d0/go.mod h1:OLgZIPagt7ERELqWJFomSt595RzquPNLL48iOWgYOg0=
+k8s.io/utils v0.0.0-20240711033017-18e509b52bc8 h1:pUdcCO1Lk/tbT5ztQWOBi5HBgbBP1J8+AsQnQCKsi8A=
+k8s.io/utils v0.0.0-20240711033017-18e509b52bc8/go.mod h1:OLgZIPagt7ERELqWJFomSt595RzquPNLL48iOWgYOg0=
 sigs.k8s.io/apiserver-network-proxy/konnectivity-client v0.30.3 h1:2770sDpzrjjsAtVhSeUFseziht227YAWYHLGNM8QPwY=
 sigs.k8s.io/apiserver-network-proxy/konnectivity-client v0.30.3/go.mod h1:Ve9uj1L+deCXFrPOk1LpFXqTg7LCFzFso6PA48q/XZw=
 sigs.k8s.io/json v0.0.0-20221116044647-bc3834ca7abd h1:EDPBXCAspyGV4jQlpZSudPeMmr1bNJefnuqLsRAsHZo=

pkg/talos/instances.go

@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"maps"
 	"strings"
+	"time"
 
 	"github.com/siderolabs/talos-cloud-controller-manager/pkg/metrics"
 	"github.com/siderolabs/talos-cloud-controller-manager/pkg/transformer"
@@ -22,6 +23,20 @@ type instances struct {
 	c *client
 }
 
+var (
+	uninitializedTaint = &v1.Taint{
+		Key:    cloudproviderapi.TaintExternalCloudProvider,
+		Effect: v1.TaintEffectNoSchedule,
+	}
+
+	notReadyTaint = &v1.Taint{
+		Key:    v1.TaintNodeNotReady,
+		Effect: v1.TaintEffectNoExecute,
+	}
+
+	initializedNodeDelay = time.Second * 30
+)
+
 func newInstances(client *client) *instances {
 	return &instances{
 		c: client,
@@ -33,6 +48,36 @@ func newInstances(client *client) *instances {
 func (i *instances) InstanceExists(_ context.Context, node *v1.Node) (bool, error) {
 	klog.V(4).InfoS("instances.InstanceExists() called", "node", klog.KRef("", node.Name))
 
+	if node.Spec.ProviderID == "" {
+		return true, nil
+	}
+
+	notReady := false
+
+	for _, taint := range node.Spec.Taints {
+		if taint.MatchTaint(uninitializedTaint) {
+			return true, nil
+		}
+
+		if taint.MatchTaint(notReadyTaint) {
+			notReady = true
+		}
+	}
+
+	delay := time.Since(node.ObjectMeta.CreationTimestamp.Time)
+	if delay < initializedNodeDelay {
+		klog.V(4).InfoS("instances.InstanceExists() wait initialized node delay", "node", klog.KRef("", node.Name), "delay", delay)
+
+		return true, nil
+	}
+
+	if notReady {
+		if node.Labels[ClusterNodePlatformLabel] == "gcp" &&
+			node.Labels[ClusterNodeLifeCycleLabel] == "spot" {
+			return false, nil
+		}
+	}
+
 	return true, nil
 }