diff --git hadoop-yarn-project/hadoop-yarn/hadoop-yarn-api/src/main/java/org/apache/hadoop/yarn/conf/YarnConfiguration.java hadoop-yarn-project/hadoop-yarn/hadoop-yarn-api/src/main/java/org/apache/hadoop/yarn/conf/YarnConfiguration.java index a18ef7c..8cc0703 100644 --- hadoop-yarn-project/hadoop-yarn/hadoop-yarn-api/src/main/java/org/apache/hadoop/yarn/conf/YarnConfiguration.java +++ hadoop-yarn-project/hadoop-yarn/hadoop-yarn-api/src/main/java/org/apache/hadoop/yarn/conf/YarnConfiguration.java @@ -402,7 +402,7 @@ private static void addDeprecatedKeys() { public static final boolean DEFAULT_RM_RECOVERY_ENABLED = false; public static final String YARN_FAIL_FAST = YARN_PREFIX + "fail-fast"; - public static final boolean DEFAULT_YARN_FAIL_FAST = true; + public static final boolean DEFAULT_YARN_FAIL_FAST = false; public static final String RM_FAIL_FAST = RM_PREFIX + "fail-fast"; diff --git hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/resources/yarn-default.xml hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/resources/yarn-default.xml index 62ba599..7070cd7 100644 --- hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/resources/yarn-default.xml +++ hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/resources/yarn-default.xml @@ -343,9 +343,12 @@ Should YARN fail fast if it encounters any errors. + This is a global config for all other components including RM,NM etc. + If no value is set for component-specific config (e.g yarn.resourcemanager.fail-fast), + this value will be the default. yarn.fail-fast - true + false diff --git hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/RMStateStore.java hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/RMStateStore.java index b7f1e6c..6e3aa69 100644 --- hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/RMStateStore.java +++ hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/RMStateStore.java @@ -45,6 +45,7 @@ import org.apache.hadoop.yarn.api.records.ApplicationSubmissionContext; import org.apache.hadoop.yarn.api.records.ReservationId; import org.apache.hadoop.yarn.api.records.impl.pb.ApplicationSubmissionContextPBImpl; +import org.apache.hadoop.yarn.conf.HAUtil; import org.apache.hadoop.yarn.conf.YarnConfiguration; import org.apache.hadoop.yarn.event.AsyncDispatcher; import org.apache.hadoop.yarn.event.Dispatcher; @@ -190,11 +191,12 @@ public void transition(RMStateStore store, RMStateStoreEvent event) { LOG.info("Storing info for app: " + appId); try { store.storeApplicationStateInternal(appId, appState); - store.notifyApplication(new RMAppEvent(appId, - RMAppEventType.APP_NEW_SAVED)); } catch (Exception e) { LOG.error("Error storing app: " + appId, e); store.notifyStoreOperationFailed(e); + } finally { + store.notifyApplication(new RMAppEvent(appId, + RMAppEventType.APP_NEW_SAVED)); } }; } @@ -215,13 +217,14 @@ public void transition(RMStateStore store, RMStateStoreEvent event) { LOG.info("Updating info for app: " + appId); try { store.updateApplicationStateInternal(appId, appState); + } catch (Exception e) { + LOG.error("Error updating app: " + appId, e); + store.notifyStoreOperationFailed(e); + } finally { if (((RMStateUpdateAppEvent) event).isNotifyApplication()) { store.notifyApplication(new RMAppEvent(appId, RMAppEventType.APP_UPDATE_SAVED)); } - } catch (Exception e) { - LOG.error("Error updating app: " + appId, e); - store.notifyStoreOperationFailed(e); } }; } @@ -266,12 +269,13 @@ public void transition(RMStateStore store, RMStateStoreEvent event) { } store.storeApplicationAttemptStateInternal(attemptState.getAttemptId(), attemptState); - store.notifyApplicationAttempt(new RMAppAttemptEvent - (attemptState.getAttemptId(), - RMAppAttemptEventType.ATTEMPT_NEW_SAVED)); } catch (Exception e) { LOG.error("Error storing appAttempt: " + attemptState.getAttemptId(), e); store.notifyStoreOperationFailed(e); + } finally { + store.notifyApplicationAttempt(new RMAppAttemptEvent + (attemptState.getAttemptId(), + RMAppAttemptEventType.ATTEMPT_NEW_SAVED)); } }; } @@ -293,12 +297,13 @@ public void transition(RMStateStore store, RMStateStoreEvent event) { } store.updateApplicationAttemptStateInternal(attemptState.getAttemptId(), attemptState); - store.notifyApplicationAttempt(new RMAppAttemptEvent - (attemptState.getAttemptId(), - RMAppAttemptEventType.ATTEMPT_UPDATE_SAVED)); } catch (Exception e) { LOG.error("Error updating appAttempt: " + attemptState.getAttemptId(), e); store.notifyStoreOperationFailed(e); + } finally { + store.notifyApplicationAttempt(new RMAppAttemptEvent + (attemptState.getAttemptId(), + RMAppAttemptEventType.ATTEMPT_UPDATE_SAVED)); } }; } @@ -1013,18 +1018,20 @@ protected void handleStoreEvent(RMStateStoreEvent event) { */ protected void notifyStoreOperationFailed(Exception failureCause) { LOG.error("State store operation failed ", failureCause); - if (failureCause instanceof StoreFencedException) { + if (HAUtil.isHAEnabled(getConfig())) { + LOG.warn("State-store fenced ! Transitioning RM to standby"); updateFencedState(); Thread standByTransitionThread = new Thread(new StandByTransitionThread()); standByTransitionThread.setName("StandByTransitionThread Handler"); standByTransitionThread.start(); + } else if (YarnConfiguration.shouldRMFailFast(getConfig())) { + LOG.warn("Fail RM now due to state-store error!"); + rmDispatcher.getEventHandler().handle( + new RMFatalEvent(RMFatalEventType.STATE_STORE_OP_FAILED, + failureCause)); } else { - if (YarnConfiguration.shouldRMFailFast(getConfig())) { - rmDispatcher.getEventHandler().handle( - new RMFatalEvent(RMFatalEventType.STATE_STORE_OP_FAILED, - failureCause)); - } + LOG.warn("Skip the state-store error."); } } diff --git hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/ZKRMStateStore.java hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/ZKRMStateStore.java index 0550087..945405e 100644 --- hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/ZKRMStateStore.java +++ hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/recovery/ZKRMStateStore.java @@ -1049,7 +1049,7 @@ public void run() { LOG.info(VerifyActiveStatusThread.class.getName() + " thread " + "interrupted! Exiting!"); } catch (Exception e) { - notifyStoreOperationFailed(new StoreFencedException()); + notifyStoreOperationFailed(e); } } }