hadoop-common-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From lium...@apache.org
Subject [40/50] [abbrv] hadoop git commit: HDFS-11383. Intern strings in BlockLocation and ExtendedBlock. Contributed by Misha Dmitriev.
Date Tue, 06 Jun 2017 08:28:30 GMT
HDFS-11383. Intern strings in BlockLocation and ExtendedBlock. Contributed by Misha Dmitriev.


Project: http://git-wip-us.apache.org/repos/asf/hadoop/repo
Commit: http://git-wip-us.apache.org/repos/asf/hadoop/commit/7101477e
Tree: http://git-wip-us.apache.org/repos/asf/hadoop/tree/7101477e
Diff: http://git-wip-us.apache.org/repos/asf/hadoop/diff/7101477e

Branch: refs/heads/HADOOP-13345
Commit: 7101477e4726a70ab0eab57c2d4480a04446a8dc
Parents: 219f4c1
Author: Andrew Wang <wang@apache.org>
Authored: Thu Jun 1 15:20:18 2017 -0700
Committer: Andrew Wang <wang@apache.org>
Committed: Thu Jun 1 15:20:18 2017 -0700

----------------------------------------------------------------------
 .../org/apache/hadoop/fs/BlockLocation.java     | 21 +++++------
 .../org/apache/hadoop/util/StringInterner.java  | 37 ++++++++++----------
 .../hadoop/hdfs/protocol/ExtendedBlock.java     | 16 +++++----
 3 files changed, 40 insertions(+), 34 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/hadoop/blob/7101477e/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/BlockLocation.java
----------------------------------------------------------------------
diff --git a/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/BlockLocation.java
b/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/BlockLocation.java
index b8cad3a..591febf 100644
--- a/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/BlockLocation.java
+++ b/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/BlockLocation.java
@@ -22,6 +22,7 @@ import java.io.Serializable;
 
 import org.apache.hadoop.classification.InterfaceAudience;
 import org.apache.hadoop.classification.InterfaceStability;
+import org.apache.hadoop.util.StringInterner;
 
 /**
  * Represents the network location of a block, information about the hosts
@@ -114,27 +115,27 @@ public class BlockLocation implements Serializable {
     if (names == null) {
       this.names = EMPTY_STR_ARRAY;
     } else {
-      this.names = names;
+      this.names = StringInterner.internStringsInArray(names);
     }
     if (hosts == null) {
       this.hosts = EMPTY_STR_ARRAY;
     } else {
-      this.hosts = hosts;
+      this.hosts = StringInterner.internStringsInArray(hosts);
     }
     if (cachedHosts == null) {
       this.cachedHosts = EMPTY_STR_ARRAY;
     } else {
-      this.cachedHosts = cachedHosts;
+      this.cachedHosts = StringInterner.internStringsInArray(cachedHosts);
     }
     if (topologyPaths == null) {
       this.topologyPaths = EMPTY_STR_ARRAY;
     } else {
-      this.topologyPaths = topologyPaths;
+      this.topologyPaths = StringInterner.internStringsInArray(topologyPaths);
     }
     if (storageIds == null) {
       this.storageIds = EMPTY_STR_ARRAY;
     } else {
-      this.storageIds = storageIds;
+      this.storageIds = StringInterner.internStringsInArray(storageIds);
     }
     if (storageTypes == null) {
       this.storageTypes = EMPTY_STORAGE_TYPE_ARRAY;
@@ -238,7 +239,7 @@ public class BlockLocation implements Serializable {
     if (hosts == null) {
       this.hosts = EMPTY_STR_ARRAY;
     } else {
-      this.hosts = hosts;
+      this.hosts = StringInterner.internStringsInArray(hosts);
     }
   }
 
@@ -249,7 +250,7 @@ public class BlockLocation implements Serializable {
     if (cachedHosts == null) {
       this.cachedHosts = EMPTY_STR_ARRAY;
     } else {
-      this.cachedHosts = cachedHosts;
+      this.cachedHosts = StringInterner.internStringsInArray(cachedHosts);
     }
   }
 
@@ -260,7 +261,7 @@ public class BlockLocation implements Serializable {
     if (names == null) {
       this.names = EMPTY_STR_ARRAY;
     } else {
-      this.names = names;
+      this.names = StringInterner.internStringsInArray(names);
     }
   }
 
@@ -271,7 +272,7 @@ public class BlockLocation implements Serializable {
     if (topologyPaths == null) {
       this.topologyPaths = EMPTY_STR_ARRAY;
     } else {
-      this.topologyPaths = topologyPaths;
+      this.topologyPaths = StringInterner.internStringsInArray(topologyPaths);
     }
   }
 
@@ -279,7 +280,7 @@ public class BlockLocation implements Serializable {
     if (storageIds == null) {
       this.storageIds = EMPTY_STR_ARRAY;
     } else {
-      this.storageIds = storageIds;
+      this.storageIds = StringInterner.internStringsInArray(storageIds);
     }
   }
 

http://git-wip-us.apache.org/repos/asf/hadoop/blob/7101477e/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/util/StringInterner.java
----------------------------------------------------------------------
diff --git a/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/util/StringInterner.java
b/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/util/StringInterner.java
index d74f810..028e49a 100644
--- a/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/util/StringInterner.java
+++ b/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/util/StringInterner.java
@@ -25,8 +25,9 @@ import com.google.common.collect.Interner;
 import com.google.common.collect.Interners;
 
 /**
- * Provides equivalent behavior to String.intern() to optimize performance, 
- * whereby does not consume memory in the permanent generation.
+ * Provides string interning utility methods. For weak interning,
+ * we use the standard String.intern() call, that performs very well
+ * (no problems with PermGen overflowing, etc.) starting from JDK 7.
  */
 @InterfaceAudience.Public
 @InterfaceStability.Stable
@@ -35,20 +36,9 @@ public class StringInterner {
   /**
    * Retains a strong reference to each string instance it has interned.
    */
-  private final static Interner<String> strongInterner;
-  
-  /**
-   * Retains a weak reference to each string instance it has interned. 
-   */
-  private final static Interner<String> weakInterner;
-  
-  
-  
-  static {
-    strongInterner = Interners.newStrongInterner();
-    weakInterner = Interners.newWeakInterner();
-  }
-  
+  private final static Interner<String> STRONG_INTERNER =
+      Interners.newStrongInterner();
+
   /**
    * Interns and returns a reference to the representative instance 
    * for any of a collection of string instances that are equal to each other.
@@ -62,7 +52,7 @@ public class StringInterner {
     if (sample == null) {
       return null;
     }
-    return strongInterner.intern(sample);
+    return STRONG_INTERNER.intern(sample);
   }
   
   /**
@@ -78,7 +68,18 @@ public class StringInterner {
     if (sample == null) {
       return null;
     }
-    return weakInterner.intern(sample);
+    return sample.intern();
+  }
+
+  /**
+   * Interns all the strings in the given array in place,
+   * returning the same array.
+   */
+  public static String[] internStringsInArray(String[] strings) {
+    for (int i = 0; i < strings.length; i++) {
+      strings[i] = weakIntern(strings[i]);
+    }
+    return strings;
   }
 
 }

http://git-wip-us.apache.org/repos/asf/hadoop/blob/7101477e/hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/protocol/ExtendedBlock.java
----------------------------------------------------------------------
diff --git a/hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/protocol/ExtendedBlock.java
b/hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/protocol/ExtendedBlock.java
index af24909..7939662 100644
--- a/hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/protocol/ExtendedBlock.java
+++ b/hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/protocol/ExtendedBlock.java
@@ -17,6 +17,7 @@
  */
 package org.apache.hadoop.hdfs.protocol;
 
+import org.apache.commons.lang.builder.HashCodeBuilder;
 import org.apache.hadoop.classification.InterfaceAudience;
 import org.apache.hadoop.classification.InterfaceStability;
 
@@ -42,13 +43,13 @@ public class ExtendedBlock {
   }
 
   public ExtendedBlock(String poolId, Block b) {
-    this.poolId = poolId;
+    this.poolId = poolId != null ? poolId.intern() : null;
     this.block = b;
   }
 
   public ExtendedBlock(final String poolId, final long blkid, final long len,
       final long genstamp) {
-    this.poolId = poolId;
+    this.poolId = poolId != null ? poolId.intern() : null;
     block = new Block(blkid, len, genstamp);
   }
 
@@ -86,7 +87,7 @@ public class ExtendedBlock {
   }
 
   public void set(String poolId, Block blk) {
-    this.poolId = poolId;
+    this.poolId = poolId != null ? poolId.intern() : null;
     this.block = blk;
   }
 
@@ -107,13 +108,16 @@ public class ExtendedBlock {
       return false;
     }
     ExtendedBlock b = (ExtendedBlock)o;
-    return b.block.equals(block) && b.poolId.equals(poolId);
+    return b.block.equals(block) &&
+        (b.poolId != null ? b.poolId.equals(poolId) : poolId == null);
   }
 
   @Override // Object
   public int hashCode() {
-    int result = 31 + poolId.hashCode();
-    return (31 * result + block.hashCode());
+    return new HashCodeBuilder(31, 17).
+        append(poolId).
+        append(block).
+        toHashCode();
   }
 
   @Override // Object


---------------------------------------------------------------------
To unsubscribe, e-mail: common-commits-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-commits-help@hadoop.apache.org


Mime
View raw message