[helix] branch master updated: Add offline node purge timeout in cluster config (#1485)

[hzlu]

This is an automated email from the ASF dual-hosted git repository.

hzlu pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/helix.git


The following commit(s) were added to refs/heads/master by this push:
     new e391119  Add offline node purge timeout in cluster config (#1485)
e391119 is described below

commit e391119f82b1fe4e182c753fc48782224389bebc
Author: Meng Zhang <mnzh...@linkedin.com>
AuthorDate: Tue Oct 27 14:02:07 2020 -0700

    Add offline node purge timeout in cluster config (#1485)
    
    Add offline node purge timeout in cluster config
---
 .../java/org/apache/helix/model/ClusterConfig.java | 33 ++++++++++++++++++++--
 .../org/apache/helix/model/TestClusterConfig.java  | 23 +++++++++++++++
 2 files changed, 53 insertions(+), 3 deletions(-)

diff --git a/helix-core/src/main/java/org/apache/helix/model/ClusterConfig.java 
b/helix-core/src/main/java/org/apache/helix/model/ClusterConfig.java
index 66b1a98..ef87542 100644
--- a/helix-core/src/main/java/org/apache/helix/model/ClusterConfig.java
+++ b/helix-core/src/main/java/org/apache/helix/model/ClusterConfig.java
@@ -128,7 +128,13 @@ public class ClusterConfig extends HelixProperty {
     // offline for more than this specified time period, it's treated as 
offline for the rest of
     // the maintenance mode's duration even when it comes online.
     // The unit is milliseconds.
-    OFFLINE_NODE_TIME_OUT_FOR_MAINTENANCE_MODE
+    OFFLINE_NODE_TIME_OUT_FOR_MAINTENANCE_MODE,
+
+    // The time out window for offline nodes to be purged; if an offline node 
has been
+    // offline for more than this specified time period, and users call purge 
participant API,
+    // then the node will be removed.
+    // The unit is milliseconds.
+    OFFLINE_NODE_TIME_OUT_FOR_PURGE
   }
 
   public enum GlobalRebalancePreferenceKey {
@@ -157,7 +163,7 @@ public class ClusterConfig extends HelixProperty {
   private final static int MIN_REBALANCE_PREFERENCE = 0;
   public final static boolean DEFAULT_GLOBAL_REBALANCE_ASYNC_MODE_ENABLED = 
true;
   private static final int GLOBAL_TARGET_TASK_THREAD_POOL_SIZE_NOT_SET = -1;
-  private static final long OFFLINE_NODE_TIME_OUT_FOR_MAINTENANCE_MODE_NOT_SET 
= -1;
+  private static final long VALUE_NOT_SET = -1;
 
   /**
    * Instantiate for a specific cluster
@@ -934,7 +940,28 @@ public class ClusterConfig extends HelixProperty {
   public long getOfflineNodeTimeOutForMaintenanceMode() {
     return _record
         
.getLongField(ClusterConfigProperty.OFFLINE_NODE_TIME_OUT_FOR_MAINTENANCE_MODE.name(),
-            OFFLINE_NODE_TIME_OUT_FOR_MAINTENANCE_MODE_NOT_SET);
+            VALUE_NOT_SET);
+  }
+
+  /**
+   * Set the default time out window for offline nodes to be purged. If an 
offline node has been
+   * offline for more than this specified time period, when users call purge 
participants API,
+   * the node will be dropped.
+   * @param timeOut timeout window in milliseconds.
+   */
+  public void setOfflineNodeTimeOutForPurge(long timeOut) {
+    
_record.setLongField(ClusterConfigProperty.OFFLINE_NODE_TIME_OUT_FOR_PURGE.name(),
+        timeOut);
+  }
+
+  /**
+   * Get the default time out window for offline nodes to be purged.
+   * @return timeout window in milliseconds
+   */
+  public long getOfflineNodeTimeOutForPurge() {
+    return _record
+        
.getLongField(ClusterConfigProperty.OFFLINE_NODE_TIME_OUT_FOR_PURGE.name(),
+            VALUE_NOT_SET);
   }
 
   /**
diff --git 
a/helix-core/src/test/java/org/apache/helix/model/TestClusterConfig.java 
b/helix-core/src/test/java/org/apache/helix/model/TestClusterConfig.java
index 7ef1fa2..a9c26bf 100644
--- a/helix-core/src/test/java/org/apache/helix/model/TestClusterConfig.java
+++ b/helix-core/src/test/java/org/apache/helix/model/TestClusterConfig.java
@@ -320,6 +320,29 @@ public class TestClusterConfig {
             -1), 10000L);
   }
 
+
+  @Test
+  public void testGetOfflineNodeTimeOutForPurge() {
+    ClusterConfig testConfig = new ClusterConfig("testId");
+    Assert.assertEquals(testConfig.getOfflineNodeTimeOutForPurge(), -1);
+
+    testConfig.getRecord()
+        
.setLongField(ClusterConfig.ClusterConfigProperty.OFFLINE_NODE_TIME_OUT_FOR_PURGE
+                .name(),
+            10000L);
+    Assert.assertEquals(testConfig.getOfflineNodeTimeOutForPurge(), 10000L);
+  }
+
+  @Test
+  public void testSetOfflineNodeTimeOutForPurge() {
+    ClusterConfig testConfig = new ClusterConfig("testId");
+    testConfig.setOfflineNodeTimeOutForPurge(10000L);
+    Assert.assertEquals(testConfig.getRecord()
+        
.getLongField(ClusterConfig.ClusterConfigProperty.OFFLINE_NODE_TIME_OUT_FOR_PURGE
+                .name(),
+            -1), 10000L);
+  }
+
   @Test
   public void testAbnormalStatesResolverConfig() {
     ClusterConfig testConfig = new ClusterConfig("testConfig");