HBASE-11195 Potentially improve block locality during major compaction for old regions

Signed-off-by: Andrew Purtell <apurtell@apache.org>

hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RSRpcServices.java

@@ -23,6 +23,7 @@ import java.io.InterruptedIOException;
 import java.lang.annotation.Retention;
 import java.lang.annotation.RetentionPolicy;
 import java.net.InetSocketAddress;
+import java.net.UnknownHostException;
 import java.util.ArrayList;
 import java.util.Collections;
 import java.util.HashMap;
@@ -763,10 +764,7 @@ public class RSRpcServices implements HBaseRPCErrorHandler,
       throw new IllegalArgumentException(e);
     }
     // Server to handle client requests.
-    String hostname = rs.conf.get("hbase.regionserver.ipc.address",
-      Strings.domainNamePointerToHostName(DNS.getDefaultHost(
-        rs.conf.get("hbase.regionserver.dns.interface", "default"),
-        rs.conf.get("hbase.regionserver.dns.nameserver", "default"))));
+    String hostname = getHostname(rs.conf);
 
     boolean mode =
         rs.conf.getBoolean(HConstants.CLUSTER_DISTRIBUTED, HConstants.DEFAULT_CLUSTER_DISTRIBUTED);
@@ -807,6 +805,13 @@ public class RSRpcServices implements HBaseRPCErrorHandler,
     rs.setName(name);
   }
 
+  public static String getHostname(Configuration conf) throws UnknownHostException {
+    return conf.get("hbase.regionserver.ipc.address",
+        Strings.domainNamePointerToHostName(DNS.getDefaultHost(
+            conf.get("hbase.regionserver.dns.interface", "default"),
+            conf.get("hbase.regionserver.dns.nameserver", "default"))));
+  }
+
   RegionScanner getScanner(long scannerId) {
     String scannerIdString = Long.toString(scannerId);
     RegionScannerHolder scannerHolder = scanners.get(scannerIdString);

hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/compactions/CompactionConfiguration.java

@@ -36,6 +36,7 @@ import org.apache.hadoop.hbase.regionserver.StoreConfigInformation;
  * minFilesToCompact - lower bound on number of files in any minor compaction
  * maxFilesToCompact - upper bound on number of files in any minor compaction
  * compactionRatio - Ratio used for compaction
+ * minLocalityToForceCompact - Locality threshold for a store file to major compact (HBASE-11195)
  * <p/>
  * Set parameter as "hbase.hstore.compaction.<attribute>"
  */
@@ -56,6 +57,8 @@ public class CompactionConfiguration {
     "hbase.hstore.compaction.max.size";
   public static final String HBASE_HSTORE_OFFPEAK_END_HOUR = "hbase.offpeak.end.hour";
   public static final String HBASE_HSTORE_OFFPEAK_START_HOUR = "hbase.offpeak.start.hour";
+  public static final String HBASE_HSTORE_MIN_LOCALITY_TO_SKIP_MAJOR_COMPACT =
+      "hbase.hstore.min.locality.to.skip.major.compact";
 
   Configuration conf;
   StoreConfigInformation storeConfigInfo;
@@ -70,6 +73,7 @@ public class CompactionConfiguration {
   private final long throttlePoint;
   private final long majorCompactionPeriod;
   private final float majorCompactionJitter;
+  private final float minLocalityToForceCompact;
 
   CompactionConfiguration(Configuration conf, StoreConfigInformation storeConfigInfo) {
     this.conf = conf;
@@ -89,7 +93,7 @@ public class CompactionConfiguration {
     majorCompactionPeriod = conf.getLong(HConstants.MAJOR_COMPACTION_PERIOD, 1000*60*60*24*7);
     // Make it 0.5 so jitter has us fall evenly either side of when the compaction should run
     majorCompactionJitter = conf.getFloat("hbase.hregion.majorcompaction.jitter", 0.50F);
-
+    minLocalityToForceCompact = conf.getFloat(HBASE_HSTORE_MIN_LOCALITY_TO_SKIP_MAJOR_COMPACT, 0f);
     LOG.info(this);
   }
 
@@ -97,7 +101,7 @@ public class CompactionConfiguration {
   public String toString() {
     return String.format(
       "size [%d, %d); files [%d, %d); ratio %f; off-peak ratio %f; throttle point %d;"
-      + " major period %d, major jitter %f",
+      + " major period %d, major jitter %f, min locality to compact %f",
       minCompactSize,
       maxCompactSize,
       minFilesToCompact,
@@ -106,7 +110,8 @@ public class CompactionConfiguration {
       offPeakCompactionRatio,
       throttlePoint,
       majorCompactionPeriod,
-      majorCompactionJitter);
+      majorCompactionJitter,
+      minLocalityToForceCompact);
   }
 
   /**
@@ -173,4 +178,13 @@ public class CompactionConfiguration {
   public float getMajorCompactionJitter() {
     return majorCompactionJitter;
   }
+
+  /**
+   * @return Block locality ratio, the ratio at which we will include old regions with a single
+   * store file for major compaction.  Used to improve block locality for regions that
+   * haven't had writes in a while but are still being read.
+   */
+  public float getMinLocalityToForceCompact() {
+    return minLocalityToForceCompact;
+  }
 }

hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/compactions/CompactionPolicy.java

@@ -62,6 +62,8 @@ public abstract class CompactionPolicy {
     this.comConf = new CompactionConfiguration(conf, this.storeConfigInfo);
   }
 
+
+
   /**
    * @return The current compaction configuration settings.
    */

hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/compactions/RatioBasedCompactionPolicy.java

@@ -30,6 +30,7 @@ import org.apache.commons.logging.LogFactory;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.hbase.HConstants;
 import org.apache.hadoop.hbase.classification.InterfaceAudience;
+import org.apache.hadoop.hbase.regionserver.RSRpcServices;
 import org.apache.hadoop.hbase.regionserver.StoreConfigInformation;
 import org.apache.hadoop.hbase.regionserver.StoreFile;
 import org.apache.hadoop.hbase.regionserver.StoreUtils;
@@ -298,10 +299,25 @@ public class RatioBasedCompactionPolicy extends CompactionPolicy {
             : now - minTimestamp.longValue();
         if (sf.isMajorCompaction() &&
             (cfTtl == HConstants.FOREVER || oldest < cfTtl)) {
-          if (LOG.isDebugEnabled()) {
-            LOG.debug("Skipping major compaction of " + this +
-                " because one (major) compacted file only and oldestTime " +
-                oldest + "ms is < ttl=" + cfTtl);
+          float blockLocalityIndex = sf.getHDFSBlockDistribution().getBlockLocalityIndex(
+              RSRpcServices.getHostname(comConf.conf)
+          );
+          if (blockLocalityIndex < comConf.getMinLocalityToForceCompact()) {
+            if (LOG.isDebugEnabled()) {
+              LOG.debug("Major compaction triggered on only store " + this +
+                  "; to make hdfs blocks local, current blockLocalityIndex is " +
+                  blockLocalityIndex + " (min " + comConf.getMinLocalityToForceCompact() +
+                  ")");
+            }
+            result = true;
+          } else {
+            if (LOG.isDebugEnabled()) {
+              LOG.debug("Skipping major compaction of " + this +
+                  " because one (major) compacted file only, oldestTime " +
+                  oldest + "ms is < ttl=" + cfTtl + " and blockLocalityIndex is " +
+                  blockLocalityIndex + " (min " + comConf.getMinLocalityToForceCompact() +
+                  ")");
+            }
           }
         } else if (cfTtl != HConstants.FOREVER && oldest > cfTtl) {
           LOG.debug("Major compaction triggered on store " + this +