Add support for multiple nodes

README.md

@@ -17,18 +17,16 @@ Available Commands:
   snapshot    Checks the status of Elasticsearch snapshots
 
 Flags:
-  -H, --hostname string    Hostname of the Elasticsearch instance (CHECK_ELASTICSEARCH_HOSTNAME) (default "localhost")
-  -p, --port int           Port of the Elasticsearch instance (default 9200)
-  -U, --username string    Username for HTTP Basic Authentication (CHECK_ELASTICSEARCH_USERNAME)
-  -P, --password string    Password for HTTP Basic Authentication (CHECK_ELASTICSEARCH_PASSWORD)
-  -S, --tls                Use a HTTPS connection
-      --insecure           Skip the verification of the server's TLS certificate
-      --ca-file string     Specify the CA File for TLS authentication (CHECK_ELASTICSEARCH_CA_FILE)
-      --cert-file string   Specify the Certificate File for TLS authentication (CHECK_ELASTICSEARCH_CERT_FILE)
-      --key-file string    Specify the Key File for TLS authentication (CHECK_ELASTICSEARCH_KEY_FILE)
-  -t, --timeout int        Timeout in seconds for the CheckPlugin (default 30)
-  -h, --help               help for check_elasticsearch
-  -v, --version            version for check_elasticsearch
+  -H, --hostname stringArray   URL of an Elasticsearch instance. Can be used multiple times. (default [http://localhost:9200])
+  -U, --username string        Username for HTTP Basic Authentication (CHECK_ELASTICSEARCH_USERNAME)
+  -P, --password string        Password for HTTP Basic Authentication (CHECK_ELASTICSEARCH_PASSWORD)
+      --insecure               Skip the verification of the server's TLS certificate
+      --ca-file string         Specify the CA File for TLS authentication (CHECK_ELASTICSEARCH_CA_FILE)
+      --cert-file string       Specify the Certificate File for TLS authentication (CHECK_ELASTICSEARCH_CERT_FILE)
+      --key-file string        Specify the Key File for TLS authentication (CHECK_ELASTICSEARCH_KEY_FILE)
+  -t, --timeout int            Timeout in seconds for the plugin (default 30)
+  -h, --help                   help for check_elasticsearch
+  -v, --version                version for check_elasticsearch
 ```
 
 The check plugin respects the environment variables `HTTP_PROXY`, `HTTPS_PROXY` and `NO_PROXY`.
@@ -54,14 +52,22 @@ Examples:
 Elasticsearch cluster with green status (all nodes are running):
 
 ```
-$ check_elasticsearch health -U exampleuser -P examplepassword -S --insecure
+$ check_elasticsearch health -U exampleuser -P examplepassword
+[OK] - Cluster es-example-cluster is green | status=0 nodes=3 data_nodes=3 active_primary_shards=10 active_shards=20
+```
+
+When you have multiple cluster nodes:
+
+```
+$ check_elasticsearch health -U exampleuser -P examplepassword \
+--hostname "https://node1:9200" --hostname "https://node2:9200" --hostname "https://node3:9200"
 [OK] - Cluster es-example-cluster is green | status=0 nodes=3 data_nodes=3 active_primary_shards=10 active_shards=20
 ```
 
 Elasticsearch cluster with yellow status (not all nodes are running):
 
 ```
-$ check_elasticsearch health -U exampleuser -P examplepassword -S --insecure
+$ check_elasticsearch health -U exampleuser -P examplepassword
 [WARNING] - Cluster es-example-cluster is yellow | status=1 nodes=2 data_nodes=2 active_primary_shards=10 active_shards=13```
 ```
 

cmd/config.go

@@ -7,7 +7,6 @@ import (
 	"net/url"
 	"os"
 	"reflect"
-	"strconv"
 	"time"
 
 	"github.com/NETWAYS/check_elasticsearch/internal/client"
@@ -16,15 +15,13 @@ import (
 )
 
 type Config struct {
+	Hostname []string
 	Bearer   string // Currently unused in CLI
-	Hostname string `env:"CHECK_ELASTICSEARCH_HOSTNAME"`
 	CAFile   string `env:"CHECK_ELASTICSEARCH_CA_FILE"`
 	CertFile string `env:"CHECK_ELASTICSEARCH_CERT_FILE"`
 	KeyFile  string `env:"CHECK_ELASTICSEARCH_KEY_FILE"`
 	Username string `env:"CHECK_ELASTICSEARCH_USERNAME"`
 	Password string `env:"CHECK_ELASTICSEARCH_PASSWORD"`
-	Port     int
-	TLS      bool
 	Insecure bool
 }
 
@@ -65,13 +62,16 @@ func loadFromEnv(config any) {
 var cliConfig Config
 
 func (c *Config) NewClient() *client.Client {
-	u := url.URL{
-		Scheme: "http",
-		Host:   c.Hostname + ":" + strconv.Itoa(c.Port),
-	}
+	urls := make([]*url.URL, 0, len(c.Hostname))
+
+	for _, host := range c.Hostname {
+		u, errParse := url.Parse(host)
+
+		if errParse != nil {
+			check.ExitError(errParse)
+		}
 
-	if c.TLS {
-		u.Scheme = "https"
+		urls = append(urls, u)
 	}
 
 	// Create TLS configuration for default RoundTripper
@@ -110,5 +110,5 @@ func (c *Config) NewClient() *client.Client {
 		rt = checkhttpconfig.NewBasicAuthRoundTripper(c.Username, c.Password, rt)
 	}
 
-	return client.NewClient(u.String(), rt)
+	return client.NewClient(urls, rt)
 }

cmd/health.go

@@ -15,8 +15,8 @@ The cluster health status is:
 	green = OK
 	yellow = WARNING
 	red = CRITICAL`,
-	Example: "  check_elasticsearch health --hostname \"127.0.0.1\" --port 9200 --username \"exampleUser\"  " +
-		"--password \"examplePass\" --tls --insecure",
+	Example: "  check_elasticsearch health --hostname \"https://localhost:9200\" --username \"exampleUser\"  " +
+		"--password \"examplePass\" --insecure",
 	Run: func(_ *cobra.Command, _ []string) {
 		client := cliConfig.NewClient()
 
@@ -26,6 +26,11 @@ The cluster health status is:
 		}
 
 		var rc int
+		// How we map cluster states:
+		// green = 0
+		// yellow = 1
+		// red = 2
+		// unknown = 3
 		switch health.Status {
 		case "green":
 			rc = check.OK
@@ -42,12 +47,7 @@ The cluster health status is:
 			output = "Cluster " + health.ClusterName + " is " + health.Status
 		}
 
-		// green = 0
-		// yellow = 1
-		// red = 2
-		// unknown = 3
 		p := perfdata.PerfdataList{
-			{Label: "status", Value: rc},
 			{Label: "nodes", Value: health.NumberOfNodes},
 			{Label: "data_nodes", Value: health.NumberOfDataNodes},
 			{Label: "active_primary_shards", Value: health.ActivePrimaryShards},

cmd/health_test.go

@@ -3,19 +3,18 @@ package cmd
 import (
 	"net/http"
 	"net/http/httptest"
-	"net/url"
 	"os/exec"
 	"strings"
 	"testing"
 )
 
 func TestHealth_ConnectionRefused(t *testing.T) {
 
-	cmd := exec.Command("go", "run", "../main.go", "health", "--port", "9999")
+	cmd := exec.Command("go", "run", "../main.go", "health", "--hostname", "http://localhost:9999")
 	out, _ := cmd.CombinedOutput()
 
 	actual := string(out)
-	expected := "[UNKNOWN] - could not fetch cluster health: Get \"http://localhost:9999/_cluster/health\": dial"
+	expected := "[UNKNOWN] - could not fetch cluster health: no node reachable (*errors.errorString)"
 
 	if !strings.Contains(actual, expected) {
 		t.Error("\nActual: ", actual, "\nExpected: ", expected)
@@ -48,7 +47,7 @@ func TestHealthCmd(t *testing.T) {
 				w.Write([]byte(`The Authorization header wasn't set`))
 			})),
 			args:     []string{"run", "../main.go", "health", "--username", "username", "--password", "password"},
-			expected: "[OK] - Cluster test is green | status=0 nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\n",
+			expected: "[OK] - Cluster test is green | nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\n",
 		},
 		{
 			name: "health-invalid",
@@ -58,7 +57,7 @@ func TestHealthCmd(t *testing.T) {
 				w.Write([]byte(`{}`))
 			})),
 			args:     []string{"run", "../main.go", "health"},
-			expected: "[UNKNOWN] - Cluster status unknown | status=3 nodes=0 data_nodes=0 active_primary_shards=0 active_shards=0\nexit status 3\n",
+			expected: "[UNKNOWN] - Cluster status unknown | nodes=0 data_nodes=0 active_primary_shards=0 active_shards=0\nexit status 3\n",
 		},
 		{
 			name: "health-404",
@@ -78,7 +77,7 @@ func TestHealthCmd(t *testing.T) {
 				w.Write([]byte(`{"cluster_name":"test","status":"foobar","timed_out":false,"number_of_nodes":1,"number_of_data_nodes":1,"active_primary_shards":3}`))
 			})),
 			args:     []string{"run", "../main.go", "health"},
-			expected: "[UNKNOWN] - Cluster test is foobar | status=3 nodes=1 data_nodes=1 active_primary_shards=3 active_shards=0\nexit status 3\n",
+			expected: "[UNKNOWN] - Cluster test is foobar | nodes=1 data_nodes=1 active_primary_shards=3 active_shards=0\nexit status 3\n",
 		},
 		{
 			name: "health-ok",
@@ -88,7 +87,7 @@ func TestHealthCmd(t *testing.T) {
 				w.Write([]byte(`{"cluster_name":"test","status":"green","timed_out":false,"number_of_nodes":1,"number_of_data_nodes":1,"active_primary_shards":3,"active_shards":3,"relocating_shards":0,"initializing_shards":0,"unassigned_shards":0,"delayed_unassigned_shards":0,"number_of_pending_tasks":0,"number_of_in_flight_fetch":0,"task_max_waiting_in_queue_millis":0,"active_shards_percent_as_number":100.0}`))
 			})),
 			args:     []string{"run", "../main.go", "health"},
-			expected: "[OK] - Cluster test is green | status=0 nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\n",
+			expected: "[OK] - Cluster test is green | nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\n",
 		},
 		{
 			name: "health-yellow",
@@ -98,7 +97,7 @@ func TestHealthCmd(t *testing.T) {
 				w.Write([]byte(`{"cluster_name":"test","status":"yellow","timed_out":false,"number_of_nodes":1,"number_of_data_nodes":1,"active_primary_shards":3,"active_shards":3,"relocating_shards":0,"initializing_shards":0,"unassigned_shards":0,"delayed_unassigned_shards":0,"number_of_pending_tasks":0,"number_of_in_flight_fetch":0,"task_max_waiting_in_queue_millis":0,"active_shards_percent_as_number":100.0}`))
 			})),
 			args:     []string{"run", "../main.go", "health"},
-			expected: "[WARNING] - Cluster test is yellow | status=1 nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\nexit status 1\n",
+			expected: "[WARNING] - Cluster test is yellow | nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\nexit status 1\n",
 		},
 		{
 			name: "health-red",
@@ -108,17 +107,15 @@ func TestHealthCmd(t *testing.T) {
 				w.Write([]byte(`{"cluster_name":"test","status":"red","timed_out":false,"number_of_nodes":1,"number_of_data_nodes":1,"active_primary_shards":3,"active_shards":3,"relocating_shards":0,"initializing_shards":0,"unassigned_shards":0,"delayed_unassigned_shards":0,"number_of_pending_tasks":0,"number_of_in_flight_fetch":0,"task_max_waiting_in_queue_millis":0,"active_shards_percent_as_number":100.0}`))
 			})),
 			args:     []string{"run", "../main.go", "health"},
-			expected: "[CRITICAL] - Cluster test is red | status=2 nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\nexit status 2\n",
+			expected: "[CRITICAL] - Cluster test is red | nodes=1 data_nodes=1 active_primary_shards=3 active_shards=3\nexit status 2\n",
 		},
 	}
 
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
 			defer test.server.Close()
 
-			// We need the random Port extracted
-			u, _ := url.Parse(test.server.URL)
-			cmd := exec.Command("go", append(test.args, "--port", u.Port())...)
+			cmd := exec.Command("go", append(test.args, "--hostname", test.server.URL)...)
 			out, _ := cmd.CombinedOutput()
 
 			actual := string(out)

cmd/ingest_test.go

@@ -3,19 +3,18 @@ package cmd
 import (
 	"net/http"
 	"net/http/httptest"
-	"net/url"
 	"os/exec"
 	"strings"
 	"testing"
 )
 
 func TestIngest_ConnectionRefused(t *testing.T) {
 
-	cmd := exec.Command("go", "run", "../main.go", "ingest", "--port", "9999")
+	cmd := exec.Command("go", "run", "../main.go", "ingest", "--hostname", "http://localhost:9999")
 	out, _ := cmd.CombinedOutput()
 
 	actual := string(out)
-	expected := "[UNKNOWN] - could not fetch cluster nodes statistics: Get \"http://localhost:9999/_nodes/stats\": dial"
+	expected := "[UNKNOWN] - could not fetch cluster nodes statistics: no node reachable (*errors.errorString)"
 
 	if !strings.Contains(actual, expected) {
 		t.Error("\nActual: ", actual, "\nExpected: ", expected)
@@ -82,9 +81,7 @@ func TestIngestCmd(t *testing.T) {
 		t.Run(test.name, func(t *testing.T) {
 			defer test.server.Close()
 
-			// We need the random Port extracted
-			u, _ := url.Parse(test.server.URL)
-			cmd := exec.Command("go", append(test.args, "--port", u.Port())...)
+			cmd := exec.Command("go", append(test.args, "--hostname", test.server.URL)...)
 			out, _ := cmd.CombinedOutput()
 
 			actual := string(out)

cmd/query_test.go

@@ -3,20 +3,19 @@ package cmd
 import (
 	"net/http"
 	"net/http/httptest"
-	"net/url"
 	"os/exec"
 	"strings"
 	"testing"
 )
 
 func TestQuery_ConnectionRefused(t *testing.T) {
 
-	cmd := exec.Command("go", "run", "../main.go", "query", "--port", "9999")
+	cmd := exec.Command("go", "run", "../main.go", "query", "--hostname", "http://localhost:9999")
 	out, _ := cmd.CombinedOutput()
 
 	actual := string(out)
 
-	expected := "[UNKNOWN] - could not execute search request: Get \"http://localhost:9999/_all/_search?size=1&track_total_hits=true\": dial"
+	expected := "[UNKNOWN] - could not execute search request: no node reachable (*errors.errorString)"
 
 	if !strings.Contains(actual, expected) {
 		t.Error("\nActual: ", actual, "\nExpected: ", expected)
@@ -137,9 +136,7 @@ exit status 1
 		t.Run(test.name, func(t *testing.T) {
 			defer test.server.Close()
 
-			// We need the random Port extracted
-			u, _ := url.Parse(test.server.URL)
-			cmd := exec.Command("go", append(test.args, "--port", u.Port())...)
+			cmd := exec.Command("go", append(test.args, "--hostname", test.server.URL)...)
 			out, _ := cmd.CombinedOutput()
 
 			actual := string(out)

cmd/root.go

@@ -41,16 +41,12 @@ func init() {
 	})
 
 	pfs := rootCmd.PersistentFlags()
-	pfs.StringVarP(&cliConfig.Hostname, "hostname", "H", "localhost",
-		"Hostname of the Elasticsearch instance (CHECK_ELASTICSEARCH_HOSTNAME)")
-	pfs.IntVarP(&cliConfig.Port, "port", "p", 9200,
-		"Port of the Elasticsearch instance")
+	pfs.StringArrayVarP(&cliConfig.Hostname, "hostname", "H", []string{"http://localhost:9200"},
+		"URL of an Elasticsearch instance. Can be used multiple times.")
 	pfs.StringVarP(&cliConfig.Username, "username", "U", "",
 		"Username for HTTP Basic Authentication (CHECK_ELASTICSEARCH_USERNAME)")
 	pfs.StringVarP(&cliConfig.Password, "password", "P", "",
 		"Password for HTTP Basic Authentication (CHECK_ELASTICSEARCH_PASSWORD)")
-	pfs.BoolVarP(&cliConfig.TLS, "tls", "S", false,
-		"Use a HTTPS connection")
 	pfs.BoolVar(&cliConfig.Insecure, "insecure", false,
 		"Skip the verification of the server's TLS certificate")
 	pfs.StringVarP(&cliConfig.CAFile, "ca-file", "", "",
@@ -60,7 +56,7 @@ func init() {
 	pfs.StringVarP(&cliConfig.KeyFile, "key-file", "", "",
 		"Specify the Key File for TLS authentication (CHECK_ELASTICSEARCH_KEY_FILE)")
 	pfs.IntVarP(&timeout, "timeout", "t", timeout,
-		"Timeout in seconds for the CheckPlugin")
+		"Timeout in seconds for the plugin")
 
 	rootCmd.Flags().SortFlags = false
 	pfs.SortFlags = false

cmd/snapshot_test.go

@@ -3,19 +3,18 @@ package cmd
 import (
 	"net/http"
 	"net/http/httptest"
-	"net/url"
 	"os/exec"
 	"strings"
 	"testing"
 )
 
 func TestSnapshot_ConnectionRefused(t *testing.T) {
 
-	cmd := exec.Command("go", "run", "../main.go", "snapshot", "--port", "9999")
+	cmd := exec.Command("go", "run", "../main.go", "snapshot", "--hostname", "http://localhost:9999")
 	out, _ := cmd.CombinedOutput()
 
 	actual := string(out)
-	expected := "[UNKNOWN] - could not fetch snapshots: Get \"http://localhost:9999/_snapshot/*/*?order=desc\": dial"
+	expected := "[UNKNOWN] - could not fetch snapshots: no node reachable (*errors.errorString)"
 
 	if !strings.Contains(actual, expected) {
 		t.Error("\nActual: ", actual, "\nExpected: ", expected)
@@ -154,9 +153,7 @@ func TestSnapshotCmd(t *testing.T) {
 		t.Run(test.name, func(t *testing.T) {
 			defer test.server.Close()
 
-			// We need the random Port extracted
-			u, _ := url.Parse(test.server.URL)
-			cmd := exec.Command("go", append(test.args, "--port", u.Port())...)
+			cmd := exec.Command("go", append(test.args, "--hostname", test.server.URL)...)
 			out, _ := cmd.CombinedOutput()
 
 			actual := string(out)