HDFS-16386. Reduce DataNode load when FsDatasetAsyncDiskService is working. (#3806)

(cherry picked from commit 746b328554)

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/DFSConfigKeys.java

@@ -131,6 +131,9 @@ public class DFSConfigKeys extends CommonConfigurationKeys {
   public static final long    DFS_DATANODE_MAX_LOCKED_MEMORY_DEFAULT = 0;
   public static final String  DFS_DATANODE_FSDATASETCACHE_MAX_THREADS_PER_VOLUME_KEY = "dfs.datanode.fsdatasetcache.max.threads.per.volume";
   public static final int     DFS_DATANODE_FSDATASETCACHE_MAX_THREADS_PER_VOLUME_DEFAULT = 4;
+  public static final String  DFS_DATANODE_FSDATASETASYNCDISK_MAX_THREADS_PER_VOLUME_KEY =
+      "dfs.datanode.fsdatasetasyncdisk.max.threads.per.volume";
+  public static final int     DFS_DATANODE_FSDATASETASYNCDISK_MAX_THREADS_PER_VOLUME_DEFAULT = 4;
   public static final String  DFS_DATANODE_LAZY_WRITER_INTERVAL_SEC = "dfs.datanode.lazywriter.interval.sec";
   public static final int     DFS_DATANODE_LAZY_WRITER_INTERVAL_DEFAULT_SEC = 60;
   public static final String  DFS_DATANODE_RAM_DISK_REPLICA_TRACKER_KEY = "dfs.datanode.ram.disk.replica.tracker";

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetAsyncDiskService.java

@@ -30,6 +30,8 @@ import java.util.concurrent.ThreadFactory;
 import java.util.concurrent.ThreadPoolExecutor;
 import java.util.concurrent.TimeUnit;
 
+import org.apache.hadoop.hdfs.DFSConfigKeys;
+import org.apache.hadoop.util.Preconditions;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 import org.apache.hadoop.hdfs.protocol.ExtendedBlock;
@@ -65,7 +67,7 @@ class FsDatasetAsyncDiskService {
   // ThreadPool core pool size
   private static final int CORE_THREADS_PER_VOLUME = 1;
   // ThreadPool maximum pool size
-  private static final int MAXIMUM_THREADS_PER_VOLUME = 4;
+  private final int maxNumThreadsPerVolume;
   // ThreadPool keep-alive time for threads over core pool size
   private static final long THREADS_KEEP_ALIVE_SECONDS = 60; 
   
@@ -90,6 +92,12 @@ class FsDatasetAsyncDiskService {
     this.datanode = datanode;
     this.fsdatasetImpl = fsdatasetImpl;
     this.threadGroup = new ThreadGroup(getClass().getSimpleName());
+    maxNumThreadsPerVolume = datanode.getConf().getInt(
+      DFSConfigKeys.DFS_DATANODE_FSDATASETASYNCDISK_MAX_THREADS_PER_VOLUME_KEY,
+          DFSConfigKeys.DFS_DATANODE_FSDATASETASYNCDISK_MAX_THREADS_PER_VOLUME_DEFAULT);
+    Preconditions.checkArgument(maxNumThreadsPerVolume > 0,
+        DFSConfigKeys.DFS_DATANODE_FSDATASETASYNCDISK_MAX_THREADS_PER_VOLUME_KEY +
+          " must be a positive integer.");
   }
 
   private void addExecutorForVolume(final FsVolumeImpl volume) {
@@ -110,7 +118,7 @@ class FsDatasetAsyncDiskService {
     };
 
     ThreadPoolExecutor executor = new ThreadPoolExecutor(
-        CORE_THREADS_PER_VOLUME, MAXIMUM_THREADS_PER_VOLUME,
+        CORE_THREADS_PER_VOLUME, maxNumThreadsPerVolume,
         THREADS_KEEP_ALIVE_SECONDS, TimeUnit.SECONDS,
         new LinkedBlockingQueue<Runnable>(), threadFactory);
 

hadoop-hdfs-project/hadoop-hdfs/src/main/resources/hdfs-default.xml

@@ -2660,6 +2660,16 @@
   </description>
 </property>
 
+<property>
+  <name>dfs.datanode.fsdatasetasyncdisk.max.threads.per.volume</name>
+  <value>4</value>
+  <description>
+    The maximum number of threads per volume used to process async disk
+    operations on the datanode. These threads consume I/O and CPU at the
+    same time. This will affect normal data node operations.
+  </description>
+</property>
+
 <property>
   <name>dfs.cachereport.intervalMsec</name>
   <value>10000</value>