Fix MeterService NPE as not initialized in so11y plugin. (#755)

Author: wu-sheng

CHANGES.md

@@ -8,6 +8,9 @@ Release Notes.
 * Add the virtual thread executor plugin
 * Fix Conflicts apm-jdk-threadpool-plugin conflicts with apm-jdk-forkjoinpool-plugin
 * Fix NPE in hikaricp-plugin if JDBC URL is not set
+* Agent kernel services could be not-booted-yet as ServiceManager#INSTANCE#boot executed after agent transfer
+  initialization. Delay so11y metrics#build when the services are not ready to avoid MeterService status is not
+  initialized.
 
 All issues and pull requests are [here](https://github.com/apache/skywalking/milestone/236?closed=1)
 

apm-sniffer/apm-agent-core/src/main/java/org/apache/skywalking/apm/agent/core/boot/ServiceManager.java

@@ -25,6 +25,7 @@
 import java.util.List;
 import java.util.Map;
 import java.util.ServiceLoader;
+import lombok.Getter;
 import org.apache.skywalking.apm.agent.core.logging.api.ILog;
 import org.apache.skywalking.apm.agent.core.logging.api.LogManager;
 import org.apache.skywalking.apm.agent.core.plugin.loader.AgentClassLoader;
@@ -37,6 +38,8 @@ public enum ServiceManager {
 
     private static final ILog LOGGER = LogManager.getLogger(ServiceManager.class);
     private Map<Class, BootService> bootedServices = Collections.emptyMap();
+    @Getter
+    private volatile boolean isBooted = false;
 
     public void boot() {
         bootedServices = loadAllServices();
@@ -127,6 +130,7 @@ private void onComplete() {
                 LOGGER.error(e, "Service [{}] AfterBoot process fails.", service.getClass().getName());
             }
         }
+        isBooted = true;
     }
 
     /**

apm-sniffer/apm-agent-core/src/main/java/org/apache/skywalking/apm/agent/core/so11y/AgentSo11y.java

@@ -22,6 +22,7 @@
 import java.util.List;
 import java.util.Map;
 import java.util.concurrent.ConcurrentHashMap;
+import org.apache.skywalking.apm.agent.core.boot.ServiceManager;
 import org.apache.skywalking.apm.agent.core.meter.Counter;
 import org.apache.skywalking.apm.agent.core.meter.Histogram;
 import org.apache.skywalking.apm.agent.core.meter.MeterFactory;
@@ -58,6 +59,12 @@ public class AgentSo11y {
     private static Histogram INTERCEPTOR_TIME_COST;
 
     public static void measureTracingContextCreation(boolean forceSampling, boolean ignoredTracingContext) {
+        if (!ServiceManager.INSTANCE.isBooted()) {
+            // Agent kernel services could be not-booted-yet as ServiceManager#INSTANCE#boot executed after agent
+            // transfer initialization.
+            // Skip when the services are not ready to avoid MeterService status is not initialized.
+            return;
+        }
         if (forceSampling) {
             if (ignoredTracingContext) {
                 if (PROPAGATED_IGNORE_CONTEXT_COUNTER == null) {
@@ -98,6 +105,12 @@ public static void measureTracingContextCreation(boolean forceSampling, boolean
     }
 
     public static void measureTracingContextCompletion(boolean ignoredTracingContext) {
+        if (!ServiceManager.INSTANCE.isBooted()) {
+            // Agent kernel services could be not-booted-yet as ServiceManager#INSTANCE#boot executed after agent
+            // transfer initialization.
+            // Skip when the services are not ready to avoid MeterService status is not initialized.
+            return;
+        }
         if (ignoredTracingContext) {
             if (FINISH_IGNORE_CONTEXT_COUNTER == null) {
                 FINISH_IGNORE_CONTEXT_COUNTER = MeterFactory.counter("finished_ignored_context_counter").build();
@@ -112,6 +125,12 @@ public static void measureTracingContextCompletion(boolean ignoredTracingContext
     }
 
     public static void measureLeakedTracingContext(boolean ignoredTracingContext) {
+        if (!ServiceManager.INSTANCE.isBooted()) {
+            // Agent kernel services could be not-booted-yet as ServiceManager#INSTANCE#boot executed after agent
+            // transfer initialization.
+            // Skip when the services are not ready to avoid MeterService status is not initialized.
+            return;
+        }
         if (ignoredTracingContext) {
             if (LEAKED_IGNORE_CONTEXT_COUNTER == null) {
                 LEAKED_IGNORE_CONTEXT_COUNTER = MeterFactory
@@ -132,6 +151,12 @@ public static void measureLeakedTracingContext(boolean ignoredTracingContext) {
     }
 
     public static void durationOfInterceptor(double timeCostInNanos) {
+        if (!ServiceManager.INSTANCE.isBooted()) {
+            // Agent kernel services could be not-booted-yet as ServiceManager#INSTANCE#boot executed after agent
+            // transfer initialization.
+            // Skip when the services are not ready to avoid MeterService status is not initialized.
+            return;
+        }
         if (INTERCEPTOR_TIME_COST == null) {
             INTERCEPTOR_TIME_COST = MeterFactory
                 .histogram("tracing_context_performance")
@@ -142,6 +167,12 @@ public static void durationOfInterceptor(double timeCostInNanos) {
     }
 
     public static void errorOfPlugin(String pluginName, String interType) {
+        if (!ServiceManager.INSTANCE.isBooted()) {
+            // Agent kernel services could be not-booted-yet as ServiceManager#INSTANCE#boot executed after agent
+            // transfer initialization.
+            // Skip when the services are not ready to avoid MeterService status is not initialized.
+            return;
+        }
         Counter counter = ERROR_COUNTER_CACHE.computeIfAbsent(pluginName + interType, key -> MeterFactory
             .counter("interceptor_error_counter")
             .tag("plugin_name", pluginName)