feat: add Better Stack as provider

.github/workflows/test-go.yml

@@ -19,12 +19,18 @@ jobs:
           go-version: '1.23'
 
       - name: Make test
+        uses: nick-fields/retry@v3
         env:
           PINGDOM_API_TOKEN: ${{ secrets.PINGDOM_API_TOKEN }}
-        run: |
-          make test
-          echo "removing generated code from coverage results"
-          mv cover.out cover.out.tmp && grep -vP "uptime-operator/(api/v1alpha1|cmd|test/utils)/" cover.out.tmp > cover.out
+          BETTERSTACK_API_TOKEN: ${{ secrets.BETTERSTACK_API_TOKEN }}
+        with:
+          timeout_minutes: 5
+          max_attempts: 2
+          retry_on: error
+          command: |
+            make test
+            echo "removing generated code from coverage results"
+            mv cover.out cover.out.tmp && grep -vP "uptime-operator/(api/v1alpha1|cmd|test/utils)/" cover.out.tmp > cover.out
 
       - name: Update coverage report
         uses: ncruces/go-coverage-report@v0

.golangci.yml

@@ -47,6 +47,10 @@ linters-settings:
   gomoddirectives:
     replace-allow-list:
       - github.com/abbot/go-http-auth
+  tagliatelle:
+    case:
+      rules:
+        json: snake  # since betterstack uses snake instead of camel case for JSON
 
 linters:
   disable-all: true

README.md

@@ -10,6 +10,7 @@
 Kubernetes Operator to watch [Traefik](https://github.com/traefik/traefik) IngressRoute(s) and register these with a (SaaS) uptime monitoring provider.
 Currently supported providers are:
 - [Pingdom](https://www.pingdom.com/)
+- [Better Stack](https://betterstack.com/)
 - Mock (for testing purposes)
 
 Submit a PR when you wish to add another provider!
@@ -61,44 +62,46 @@ USAGE:
    <uptime-controller-manager> [OPTIONS]
 
 OPTIONS:
+  -betterstack-api-token string
+    	The API token to authenticate with Better Stack. Only applies when 'uptime-provider' is 'betterstack'
   -enable-deletes
-        Allow the operator to delete checks from the uptime provider when ingress routes are removed.
+    	Allow the operator to delete checks from the uptime provider when ingress routes are removed.
   -enable-http2
-        If set, HTTP/2 will be enabled for the metrics and webhook servers.
+    	If set, HTTP/2 will be enabled for the metrics and webhook servers.
   -health-probe-bind-address string
-        The address the probe endpoint binds to. (default ":8081")
+    	The address the probe endpoint binds to. (default ":8081")
   -kubeconfig string
-        Paths to a kubeconfig. Only required if out-of-cluster.
+    	Paths to a kubeconfig. Only required if out-of-cluster.
   -leader-elect
-        Enable leader election for controller manager. Enabling this will ensure there is only one active controller manager.
+    	Enable leader election for controller manager. Enabling this will ensure there is only one active controller manager.
   -metrics-bind-address string
-        The address the metric endpoint binds to. (default ":8080")
+    	The address the metric endpoint binds to. (default ":8080")
   -metrics-secure
-        If set the metrics endpoint is served securely.
+    	If set the metrics endpoint is served securely.
   -namespace value
-        Namespace(s) to watch for changes. Specify this flag multiple times for each namespace to watch. When not provided all namespaces will be watched.
+    	Namespace(s) to watch for changes. Specify this flag multiple times for each namespace to watch. When not provided all namespaces will be watched.
   -pingdom-alert-integration-ids value
-        One or more IDs of Pingdom integrations (like slack channels) to alert. Only applies when 'uptime-provider' is 'pingdom'
+    	One or more IDs of Pingdom integrations (like slack channels) to alert. Only applies when 'uptime-provider' is 'pingdom'
   -pingdom-alert-user-ids value
-        One or more IDs of Pingdom users to alert. Only applies when 'uptime-provider' is 'pingdom'
+    	One or more IDs of Pingdom users to alert. Only applies when 'uptime-provider' is 'pingdom'
   -pingdom-api-token string
-        The API token to authenticate with Pingdom. Only applies when 'uptime-provider' is 'pingdom'
+    	The API token to authenticate with Pingdom. Only applies when 'uptime-provider' is 'pingdom'
   -slack-channel string
-        The Slack Channel ID for posting updates when uptime checks are mutated.
+    	The Slack Channel ID for posting updates when uptime checks are mutated.
   -slack-webhook-url string
-        The webhook URL required to post messages to the given Slack channel.
+    	The webhook URL required to post messages to the given Slack channel.
   -uptime-provider string
-        Name of the (SaaS) uptime monitoring provider to use. (default "mock")
+    	Name of the (SaaS) uptime monitoring provider to use. (default "mock")
   -zap-devel
-        Development Mode defaults(encoder=consoleEncoder,logLevel=Debug,stackTraceLevel=Warn). Production Mode defaults(encoder=jsonEncoder,logLevel=Info,stackTraceLevel=Error) (default true)
+    	Development Mode defaults(encoder=consoleEncoder,logLevel=Debug,stackTraceLevel=Warn). Production Mode defaults(encoder=jsonEncoder,logLevel=Info,stackTraceLevel=Error) (default true)
   -zap-encoder value
-        Zap log encoding (one of 'json' or 'console')
+    	Zap log encoding (one of 'json' or 'console')
   -zap-log-level value
-        Zap Level to configure the verbosity of logging. Can be one of 'debug', 'info', 'error', or any integer value > 0 which corresponds to custom debug levels of increasing verbosity
+    	Zap Level to configure the verbosity of logging. Can be one of 'debug', 'info', 'error', or any integer value > 0 which corresponds to custom debug levels of increasing verbosity
   -zap-stacktrace-level value
-        Zap Level at and above which stacktraces are captured (one of 'info', 'error', 'panic').
+    	Zap Level at and above which stacktraces are captured (one of 'info', 'error', 'panic').
   -zap-time-encoding value
-        Zap time encoding (one of 'epoch', 'millis', 'nano', 'iso8601', 'rfc3339' or 'rfc3339nano'). Defaults to 'epoch'.
+    	Zap time encoding (one of 'epoch', 'millis', 'nano', 'iso8601', 'rfc3339' or 'rfc3339nano'). Defaults to 'epoch'.
 ```
 
 ## Develop

cmd/main.go

@@ -22,7 +22,9 @@ import (
 	"os"
 
 	"github.com/PDOK/uptime-operator/internal/service"
-	"github.com/PDOK/uptime-operator/internal/service/providers"
+	p "github.com/PDOK/uptime-operator/internal/service/providers"
+	"github.com/PDOK/uptime-operator/internal/service/providers/betterstack"
+	"github.com/PDOK/uptime-operator/internal/service/providers/pingdom"
 	"github.com/PDOK/uptime-operator/internal/util"
 	"github.com/peterbourgon/ff"
 	"sigs.k8s.io/controller-runtime/pkg/cache"
@@ -73,6 +75,9 @@ func main() {
 	var pingdomAPIToken string
 	var pingdomAlertUserIDs util.SliceFlag
 	var pingdomAlertIntegrationIDs util.SliceFlag
+	var betterstackAPIToken string
+
+	// Default kubebuilder
 	flag.StringVar(&metricsAddr, "metrics-bind-address", ":8080",
 		"The address the metric endpoint binds to.")
 	flag.StringVar(&probeAddr, "health-probe-bind-address", ":8081",
@@ -86,6 +91,8 @@ func main() {
 		"If set, HTTP/2 will be enabled for the metrics and webhook servers.")
 	flag.BoolVar(&enableDeletes, "enable-deletes", false,
 		"Allow the operator to delete checks from the uptime provider when ingress routes are removed.")
+
+	// General uptime-operator
 	flag.Var(&namespaces, "namespace", "Namespace(s) to watch for changes. "+
 		"Specify this flag multiple times for each namespace to watch. When not provided all namespaces will be watched.")
 	flag.StringVar(&slackChannel, "slack-channel", "",
@@ -94,13 +101,19 @@ func main() {
 		"The webhook URL required to post messages to the given Slack channel.")
 	flag.StringVar(&uptimeProvider, "uptime-provider", "mock",
 		"Name of the (SaaS) uptime monitoring provider to use.")
+
+	// Pingdom specific
 	flag.StringVar(&pingdomAPIToken, "pingdom-api-token", "",
 		"The API token to authenticate with Pingdom. Only applies when 'uptime-provider' is 'pingdom'")
 	flag.Var(&pingdomAlertUserIDs, "pingdom-alert-user-ids",
 		"One or more IDs of Pingdom users to alert. Only applies when 'uptime-provider' is 'pingdom'")
 	flag.Var(&pingdomAlertIntegrationIDs, "pingdom-alert-integration-ids",
 		"One or more IDs of Pingdom integrations (like slack channels) to alert. Only applies when 'uptime-provider' is 'pingdom'")
 
+	// Better Stack specific
+	flag.StringVar(&betterstackAPIToken, "betterstack-api-token", "",
+		"The API token to authenticate with Better Stack. Only applies when 'uptime-provider' is 'betterstack'")
+
 	opts := zap.Options{
 		Development: true,
 	}
@@ -119,7 +132,10 @@ func main() {
 	}
 
 	var uptimeProviderSettings any
-	if uptimeProvider == "pingdom" {
+	uptimeProviderID := p.UptimeProviderID(uptimeProvider)
+
+	// Optional provider specific flag handling
+	if uptimeProviderID == p.ProviderPingdom {
 		alertUserIDs, err := util.StringsToInts(pingdomAlertUserIDs)
 		if err != nil {
 			setupLog.Error(err, "Unable to parse 'pingdom-alert-user-ids' flag")
@@ -130,18 +146,23 @@ func main() {
 			setupLog.Error(err, "Unable to parse 'pingdom-alert-integration-ids' flag")
 			os.Exit(1)
 		}
-		uptimeProviderSettings = providers.PingdomSettings{
+		uptimeProviderSettings = pingdom.Settings{
 			APIToken:       pingdomAPIToken,
 			UserIDs:        alertUserIDs,
 			IntegrationIDs: alertIntegrationIDs,
 		}
+	} else if uptimeProviderID == p.ProviderBetterStack {
+		uptimeProviderSettings = betterstack.Settings{
+			APIToken: betterstackAPIToken,
+		}
 	}
 
+	// Setup controller
 	if err = (&controller.IngressRouteReconciler{
 		Client: mgr.GetClient(),
 		Scheme: mgr.GetScheme(),
 		UptimeCheckService: service.New(
-			service.WithProviderAndSettings(uptimeProvider, uptimeProviderSettings),
+			service.WithProviderAndSettings(uptimeProviderID, uptimeProviderSettings),
 			service.WithSlack(slackWebhookURL, slackChannel),
 			service.WithDeletes(enableDeletes),
 		),

internal/model/check.go

@@ -32,9 +32,9 @@ type UptimeCheck struct {
 	URL               string            `json:"url"`
 	Tags              []string          `json:"tags"`
 	Interval          int               `json:"resolution"`
-	RequestHeaders    map[string]string `json:"request_headers"`     //nolint:tagliatelle // grandfathered in
-	StringContains    string            `json:"string_contains"`     //nolint:tagliatelle // grandfathered in
-	StringNotContains string            `json:"string_not_contains"` //nolint:tagliatelle // grandfathered in
+	RequestHeaders    map[string]string `json:"request_headers"`
+	StringContains    string            `json:"string_contains"`
+	StringNotContains string            `json:"string_not_contains"`
 }
 
 func NewUptimeCheck(ingressName string, annotations map[string]string) (*UptimeCheck, error) {

internal/service/providers/betterstack/betterstack.go

@@ -0,0 +1,124 @@
+package betterstack
+
+import (
+	"context"
+	"fmt"
+	classiclog "log"
+	"net/http"
+	"strconv"
+	"time"
+
+	"github.com/PDOK/uptime-operator/internal/model"
+	p "github.com/PDOK/uptime-operator/internal/service/providers"
+	"sigs.k8s.io/controller-runtime/pkg/log"
+)
+
+const betterStackBaseURL = "https://uptime.betterstack.com"
+
+type Settings struct {
+	APIToken string
+	PageSize int
+}
+
+type BetterStack struct {
+	client Client
+}
+
+// New creates a BetterStack
+func New(settings Settings) *BetterStack {
+	if settings.APIToken == "" {
+		classiclog.Fatal("Better Stack API token is not provided")
+	}
+	if settings.PageSize < 1 {
+		settings.PageSize = 50 // default https://betterstack.com/docs/uptime/api/pagination/
+	}
+	return &BetterStack{
+		Client{
+			httpClient: &http.Client{Timeout: time.Duration(5) * time.Minute},
+			settings:   settings,
+		},
+	}
+}
+
+// CreateOrUpdateCheck create the given check with Better Stack, or update an existing check. Needs to be idempotent!
+func (b *BetterStack) CreateOrUpdateCheck(ctx context.Context, check model.UptimeCheck) (err error) {
+	existingCheckID, err := b.findCheck(check)
+	if err != nil {
+		return fmt.Errorf("failed to find check %s, error: %w", check.ID, err)
+	}
+	if existingCheckID == p.CheckNotFound { //nolint:nestif // clean enough
+		log.FromContext(ctx).Info("creating check", "check", check)
+		monitorID, err := b.client.createMonitor(check)
+		if err != nil {
+			return fmt.Errorf("failed to create monitor for check %s, error: %w", check.ID, err)
+		}
+		if err = b.client.createMetadata(check.ID, monitorID, check.Tags); err != nil {
+			return fmt.Errorf("failed to create metadata for check %s, error: %w", check.ID, err)
+		}
+	} else {
+		log.FromContext(ctx).Info("updating check", "check", check, "betterstack ID", existingCheckID)
+		existingMonitor, err := b.client.getMonitor(existingCheckID)
+		if err != nil {
+			return fmt.Errorf("failed to get monitor for check %s, error: %w", check.ID, err)
+		}
+		if err = b.client.updateMonitor(check, existingMonitor); err != nil {
+			return fmt.Errorf("failed to update monitor for check %s (betterstack ID: %d), "+
+				"error: %w", check.ID, existingCheckID, err)
+		}
+		if err = b.client.updateMetadata(check.ID, existingCheckID, check.Tags); err != nil {
+			return fmt.Errorf("failed to update metdata for check %s (betterstack ID: %d), "+
+				"error: %w", check.ID, existingCheckID, err)
+		}
+	}
+	return err
+}
+
+// DeleteCheck deletes the given check from Better Stack
+func (b *BetterStack) DeleteCheck(ctx context.Context, check model.UptimeCheck) error {
+	log.FromContext(ctx).Info("deleting check", "check", check)
+
+	existingCheckID, err := b.findCheck(check)
+	if err != nil {
+		return fmt.Errorf("failed to find check %s, error: %w", check.ID, err)
+	}
+	if existingCheckID == p.CheckNotFound {
+		log.FromContext(ctx).Info(fmt.Sprintf("check with ID '%s' is already deleted", check.ID))
+		return nil
+	}
+	if err = b.client.deleteMetadata(check.ID, existingCheckID); err != nil {
+		return fmt.Errorf("failed to delete metadata for check %s (betterstack ID: %d), "+
+			"error: %w", check.ID, existingCheckID, err)
+	}
+	if err = b.client.deleteMonitor(existingCheckID); err != nil {
+		return fmt.Errorf("failed to delete monitor for check %s (betterstack ID: %d), "+
+			"error: %w", check.ID, existingCheckID, err)
+	}
+	return nil
+}
+
+func (b *BetterStack) findCheck(check model.UptimeCheck) (int64, error) {
+	result := p.CheckNotFound
+	metadata, err := b.client.listMetadata()
+	if err != nil {
+		return result, err
+	}
+	for {
+		for _, md := range metadata.Data {
+			if md.Attributes != nil && md.Attributes.Key == check.ID {
+				result, err = strconv.ParseInt(md.Attributes.OwnerID, 10, 64)
+				if err != nil {
+					return result, fmt.Errorf("failed to parse monitor ID %s to integer", md.Attributes.OwnerID)
+				}
+				return result, nil
+			}
+		}
+		if !metadata.HasNext() {
+			break // exit infinite loop
+		}
+		metadata, err = metadata.Next(b.client)
+		if err != nil {
+			return result, err
+		}
+	}
+	return result, nil
+}