org.lilyproject.hadooptestfw.fork.HBaseTestingUtility.java Source code

Java tutorial

Introduction

Here is the source code for org.lilyproject.hadooptestfw.fork.HBaseTestingUtility.java

Source

//
// LILY SPECIFIC NOTES:
//
//  This file was literally copied from HBase with the following changes applied:
//   - Fix ZooKeeper client port to 2181
//   - Fix Hadoop namenode port number to 8020
//   - Fix jobtracker port number to 8021 + set task tracker hostname to localhost (useful for links in web ui to work)
//   - Add getMRCluster()
//   - Allow to keep data from previous runs (optional) (doesn't format hdfs)
//   - Use custom version of MiniZooKeeperCluster that allows to keep data
//   - Disable shutdown as it often hangs for a long time (bad thread interruptibility)
//
//  All changes are commented with "Lily change"
//
//  In case of HBase upgrade, copy new version of this file and reapply the changes.

/**
 * Copyright 2009 The Apache Software Foundation
 *
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
package org.lilyproject.hadooptestfw.fork;

import static org.junit.Assert.assertTrue;

import java.io.File;
import java.io.IOException;
import java.io.OutputStream;
import java.lang.reflect.Field;
import java.net.InetAddress;
import java.net.ServerSocket;
import java.net.Socket;
import java.net.UnknownHostException;
import java.security.MessageDigest;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
import java.util.Collections;
import java.util.List;
import java.util.Map;
import java.util.NavigableSet;
import java.util.Random;

import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.commons.logging.impl.Jdk14Logger;
import org.apache.commons.logging.impl.Log4JLogger;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hbase.Abortable;
import org.apache.hadoop.hbase.DistributedHBaseCluster;
import org.apache.hadoop.hbase.EmptyWatcher;
import org.apache.hadoop.hbase.HBaseCluster;
import org.apache.hadoop.hbase.HBaseConfiguration;
import org.apache.hadoop.hbase.HColumnDescriptor;
import org.apache.hadoop.hbase.HConstants;
import org.apache.hadoop.hbase.HRegionInfo;
import org.apache.hadoop.hbase.HRegionLocation;
import org.apache.hadoop.hbase.HTableDescriptor;
import org.apache.hadoop.hbase.KeyValue;
import org.apache.hadoop.hbase.MasterNotRunningException;
import org.apache.hadoop.hbase.MiniHBaseCluster;
import org.apache.hadoop.hbase.ServerName;
import org.apache.hadoop.hbase.TableExistsException;
import org.apache.hadoop.hbase.TableNotEnabledException;
import org.apache.hadoop.hbase.ZooKeeperConnectionException;
import org.apache.hadoop.hbase.client.Delete;
import org.apache.hadoop.hbase.client.Get;
import org.apache.hadoop.hbase.client.HBaseAdmin;
import org.apache.hadoop.hbase.client.HConnection;
import org.apache.hadoop.hbase.client.HTable;
import org.apache.hadoop.hbase.client.Put;
import org.apache.hadoop.hbase.client.Result;
import org.apache.hadoop.hbase.client.ResultScanner;
import org.apache.hadoop.hbase.client.Scan;
import org.apache.hadoop.hbase.fs.HFileSystem;
import org.apache.hadoop.hbase.io.encoding.DataBlockEncoding;
import org.apache.hadoop.hbase.io.hfile.ChecksumUtil;
import org.apache.hadoop.hbase.io.hfile.Compression;
import org.apache.hadoop.hbase.io.hfile.Compression.Algorithm;
import org.apache.hadoop.hbase.io.hfile.HFile;
import org.apache.hadoop.hbase.master.HMaster;
import org.apache.hadoop.hbase.master.ServerManager;
import org.apache.hadoop.hbase.regionserver.HRegion;
import org.apache.hadoop.hbase.regionserver.HRegionServer;
import org.apache.hadoop.hbase.regionserver.InternalScanner;
import org.apache.hadoop.hbase.regionserver.MultiVersionConsistencyControl;
import org.apache.hadoop.hbase.regionserver.Store;
import org.apache.hadoop.hbase.regionserver.StoreFile;
import org.apache.hadoop.hbase.security.User;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.hadoop.hbase.util.FSUtils;
import org.apache.hadoop.hbase.util.JVMClusterUtil;
import org.apache.hadoop.hbase.util.JVMClusterUtil.MasterThread;
import org.apache.hadoop.hbase.util.RegionSplitter;
import org.apache.hadoop.hbase.util.Threads;
import org.apache.hadoop.hbase.util.Writables;
import org.apache.hadoop.hbase.zookeeper.ZKAssign;
import org.apache.hadoop.hbase.zookeeper.ZKConfig;
import org.apache.hadoop.hbase.zookeeper.ZooKeeperWatcher;
import org.apache.hadoop.hdfs.DFSClient;
import org.apache.hadoop.hdfs.DistributedFileSystem;
import org.apache.hadoop.hdfs.MiniDFSCluster;
import org.apache.hadoop.mapred.JobConf;
import org.apache.hadoop.mapred.MiniMRCluster;
import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.KeeperException.NodeExistsException;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.ZooKeeper;
import org.lilyproject.hadooptestfw.HBaseTestingUtilityFactory;

//Lily change: comment out import in order to use our own fork
//import org.apache.hadoop.hbase.zookeeper.MiniZooKeeperCluster;

/**
 * Facility for testing HBase. Replacement for
 * old HBaseTestCase and HBaseClusterTestCase functionality.
 * Create an instance and keep it around testing HBase.  This class is
 * meant to be your one-stop shop for anything you might need testing.  Manages
 * one cluster at a time only. Managed cluster can be an in-process
 * {@link MiniHBaseCluster}, or a deployed cluster of type {@link DistributedHBaseCluster}.
 * Not all methods work with the real cluster.
 * Depends on log4j being on classpath and
 * hbase-site.xml for logging and test-run configuration.  It does not set
 * logging levels nor make changes to configuration parameters.
 */
public class HBaseTestingUtility {
    private static final Log LOG = LogFactory.getLog(HBaseTestingUtility.class);
    private Configuration conf;
    private MiniZooKeeperCluster zkCluster = null;

    /**
     * The default number of regions per regionserver when creating a pre-split
     * table.
     */
    private static int DEFAULT_REGIONS_PER_SERVER = 5;

    /**
     * Set if we were passed a zkCluster.  If so, we won't shutdown zk as
     * part of general shutdown.
     */
    private boolean passedZkCluster = false;
    private MiniDFSCluster dfsCluster = null;

    private MiniHBaseCluster hbaseCluster = null;
    private MiniMRCluster mrCluster = null;

    // Directory where we put the data for this instance of HBaseTestingUtility
    private File dataTestDir = null;

    // Directory (usually a subdirectory of dataTestDir) used by the dfs cluster
    //  if any
    private File clusterTestDir = null;

    // Lily change: add the clearData boolean
    private boolean clearData;

    /**
     * System property key to get test directory value.
     * Name is as it is because mini dfs has hard-codings to put test data here.
     * It should NOT be used directly in HBase, as it's a property used in
     * mini dfs.
     *
     * @deprecated can be used only with mini dfs
     */
    @Deprecated
    // Lily change: made public
    public static final String TEST_DIRECTORY_KEY = "test.build.data";

    /**
     * System property key to get base test directory value
     */
    public static final String BASE_TEST_DIRECTORY_KEY = "test.build.data.basedirectory";

    /**
     * Default base directory for test output.
     */
    public static final String DEFAULT_BASE_TEST_DIRECTORY = "target/test-data";

    /**
     * Compression algorithms to use in parameterized JUnit 4 tests
     */
    public static final List<Object[]> COMPRESSION_ALGORITHMS_PARAMETERIZED = Arrays
            .asList(new Object[][] { { Compression.Algorithm.NONE }, { Compression.Algorithm.GZ } });

    /**
     * This is for unit tests parameterized with a single boolean.
     */
    public static final List<Object[]> BOOLEAN_PARAMETERIZED = Arrays
            .asList(new Object[][] { { new Boolean(false) }, { new Boolean(true) } });

    /**
     * Compression algorithms to use in testing
     */
    public static final Compression.Algorithm[] COMPRESSION_ALGORITHMS = { Compression.Algorithm.NONE,
            Compression.Algorithm.GZ };

    /**
     * Create all combinations of Bloom filters and compression algorithms for
     * testing.
     */
    private static List<Object[]> bloomAndCompressionCombinations() {
        List<Object[]> configurations = new ArrayList<Object[]>();
        for (Compression.Algorithm comprAlgo : HBaseTestingUtility.COMPRESSION_ALGORITHMS) {
            for (StoreFile.BloomType bloomType : StoreFile.BloomType.values()) {
                configurations.add(new Object[] { comprAlgo, bloomType });
            }
        }
        return Collections.unmodifiableList(configurations);
    }

    public static final Collection<Object[]> BLOOM_AND_COMPRESSION_COMBINATIONS = bloomAndCompressionCombinations();

    public HBaseTestingUtility() {
        this(HBaseConfiguration.create());
    }

    public HBaseTestingUtility(Configuration conf) {
        this.conf = conf;

        // a hbase checksum verification failure will cause unit tests to fail
        ChecksumUtil.generateExceptionForChecksumFailureForTest(true);
        setHDFSClientRetryProperty();
    }

    private void setHDFSClientRetryProperty() {
        this.conf.setInt("hdfs.client.retries.number", 1);
        // Lily change: we can't call this protected method in our fork
        // HBaseFileSystem.setRetryCounts(conf);
    }

    // Lily change: added this constructor
    public HBaseTestingUtility(Configuration conf, boolean clearData) {
        this(conf);
        this.clearData = clearData;
    }

    /**
     * Returns this classes's instance of {@link Configuration}.  Be careful how
     * you use the returned Configuration since {@link HConnection} instances
     * can be shared.  The Map of HConnections is keyed by the Configuration.  If
     * say, a Connection was being used against a cluster that had been shutdown,
     * see {@link #shutdownMiniCluster()}, then the Connection will no longer
     * be wholesome.  Rather than use the return direct, its usually best to
     * make a copy and use that.  Do
     * <code>Configuration c = new Configuration(INSTANCE.getConfiguration());</code>
     *
     * @return Instance of Configuration.
     */
    public Configuration getConfiguration() {
        return this.conf;
    }

    /**
     * @return Where to write test data on local filesystem; usually
     *         {@link #DEFAULT_BASE_TEST_DIRECTORY}
     *         Should not be used by the unit tests, hence its's private.
     *         Unit test will use a subdirectory of this directory.
     * @see #setupDataTestDir()
     * @see #getTestFileSystem()
     */
    private Path getBaseTestDir() {
        String PathName = System.getProperty(BASE_TEST_DIRECTORY_KEY, DEFAULT_BASE_TEST_DIRECTORY);

        return new Path(PathName);
    }

    /**
     * @return Where to write test data on local filesystem, specific to
     *         the test.  Useful for tests that do not use a cluster.
     *         Creates it if it does not exist already.
     * @see #getTestFileSystem()
     */
    public Path getDataTestDir() {
        if (dataTestDir == null) {
            setupDataTestDir();
        }
        return new Path(dataTestDir.getAbsolutePath());
    }

    /**
     * @return Where the DFS cluster will write data on the local subsystem.
     *         Creates it if it does not exist already.
     * @see #getTestFileSystem()
     */
    public Path getClusterTestDir() {
        if (clusterTestDir == null) {
            setupClusterTestDir();
        }
        return new Path(clusterTestDir.getAbsolutePath());
    }

    /**
     * @return Path to a subdirectory named <code>subdirName</code> under
     *         {@link #getDataTestDir()}.
     *         Does *NOT* create it if it does not exist.
     */
    public Path getDataTestDir(final String subdirName) {
        return new Path(getDataTestDir(), subdirName);
    }

    /**
     * Home our data in a dir under {@link #DEFAULT_BASE_TEST_DIRECTORY}.
     * Give it a random name so can have many concurrent tests running if
     * we need to.  It needs to amend the {@link #TEST_DIRECTORY_KEY}
     * System property, as it's what minidfscluster bases
     * it data dir on.  Moding a System property is not the way to do concurrent
     * instances -- another instance could grab the temporary
     * value unintentionally -- but not anything can do about it at moment;
     * single instance only is how the minidfscluster works.
     *
     * We also create the underlying directory for
     * hadoop.log.dir, mapred.local.dir and hadoop.tmp.dir, and set the values
     * in the conf, and as a system property for hadoop.tmp.dir
     *
     * @return The calculated data test build directory.
     */
    private void setupDataTestDir() {
        if (dataTestDir != null) {
            LOG.warn("Data test dir already setup in " + dataTestDir.getAbsolutePath());
            return;
        }

        // Lily change: don't use random dir name + don't delete on exit
        // This parameter was provided to us by HBaseTestingUtilityFactory
        String dataTestDirName = conf.get(HBaseTestingUtilityFactory.TEST_DIR_KEY);
        if (dataTestDirName == null) {
            throw new IllegalStateException("No value provided for " + HBaseTestingUtilityFactory.TEST_DIR_KEY);
        }
        Path testPath = new Path(dataTestDirName);
        dataTestDir = new File(testPath.toString()).getAbsoluteFile();
        dataTestDir.mkdirs();

        /*
        String randomStr = UUID.randomUUID().toString();
        Path testPath= new Path(getBaseTestDir(), randomStr);
            
        dataTestDir = new File(testPath.toString()).getAbsoluteFile();
        dataTestDir.deleteOnExit();
        */

        createSubDirAndSystemProperty("hadoop.log.dir", testPath, "hadoop-log-dir");

        // This is defaulted in core-default.xml to /tmp/hadoop-${user.name}, but
        //  we want our own value to ensure uniqueness on the same machine
        createSubDirAndSystemProperty("hadoop.tmp.dir", testPath, "hadoop-tmp-dir");

        // Read and modified in org.apache.hadoop.mapred.MiniMRCluster
        createSubDir("mapred.local.dir", testPath, "mapred-local-dir");

        createSubDirAndSystemProperty("mapred.working.dir", testPath, "mapred-working-dir");

        createSubDir("hbase.local.dir", testPath, "hbase-local-dir");
    }

    private void createSubDir(String propertyName, Path parent, String subDirName) {
        Path newPath = new Path(parent, subDirName);
        File newDir = new File(newPath.toString()).getAbsoluteFile();

        // Lily change: don't delete on exit
        // newDir.deleteOnExit();
        conf.set(propertyName, newDir.getAbsolutePath());
    }

    private void createSubDirAndSystemProperty(String propertyName, Path parent, String subDirName) {

        String sysValue = System.getProperty(propertyName);

        if (sysValue != null) {
            // There is already a value set. So we do nothing but hope
            //  that there will be no conflicts
            LOG.info("System.getProperty(\"" + propertyName + "\") already set to: " + sysValue
                    + " so I do NOT create it in " + dataTestDir.getAbsolutePath());
            String confValue = conf.get(propertyName);
            if (confValue != null && !confValue.endsWith(sysValue)) {
                LOG.warn(propertyName + " property value differs in configuration and system: " + "Configuration="
                        + confValue + " while System=" + sysValue
                        + " Erasing configuration value by system value.");
            }
            conf.set(propertyName, sysValue);
        } else {
            // Ok, it's not set, so we create it as a subdirectory
            createSubDir(propertyName, parent, subDirName);
            System.setProperty(propertyName, conf.get(propertyName));
        }
    }

    /**
     * Creates a directory for the DFS cluster, under the test data
     */
    private void setupClusterTestDir() {
        if (clusterTestDir != null) {
            LOG.warn("Cluster test dir already setup in " + clusterTestDir.getAbsolutePath());
            return;
        }

        // Lily change: don't use random dir name + don't delete on exit
        Path testDir = getDataTestDir("dfscluster");
        clusterTestDir = new File(testDir.toString()).getAbsoluteFile();
        clusterTestDir.mkdirs();

        /*
        // Using randomUUID ensures that multiple clusters can be launched by
        //  a same test, if it stops & starts them
        Path testDir = getDataTestDir("dfscluster_" + UUID.randomUUID().toString());
        clusterTestDir = new File(testDir.toString()).getAbsoluteFile();
        // Have it cleaned up on exit
        clusterTestDir.deleteOnExit();
        */
    }

    /**
     * @throws IOException If a cluster -- zk, dfs, or hbase -- already running.
     */
    public void isRunningCluster() throws IOException {
        if (dfsCluster == null)
            return;
        throw new IOException("Cluster already running at " + this.clusterTestDir);
    }

    /**
     * Start a minidfscluster.
     *
     * @param servers How many DNs to start.
     * @return The mini dfs cluster created.
     * @see {@link #shutdownMiniDFSCluster()}
     */
    public MiniDFSCluster startMiniDFSCluster(int servers) throws Exception {
        return startMiniDFSCluster(servers, null);
    }

    /**
     * Start a minidfscluster.
     * This is useful if you want to run datanode on distinct hosts for things
     * like HDFS block location verification.
     * If you start MiniDFSCluster without host names, all instances of the
     * datanodes will have the same host name.
     *
     * @param hosts hostnames DNs to run on.
     * @return The mini dfs cluster created.
     * @see {@link #shutdownMiniDFSCluster()}
     */
    public MiniDFSCluster startMiniDFSCluster(final String hosts[]) throws Exception {
        if (hosts != null && hosts.length != 0) {
            return startMiniDFSCluster(hosts.length, hosts);
        } else {
            return startMiniDFSCluster(1, null);
        }
    }

    /**
     * Start a minidfscluster.
     * Can only create one.
     *
     * @param servers How many DNs to start.
     * @param hosts   hostnames DNs to run on.
     * @return The mini dfs cluster created.
     * @see {@link #shutdownMiniDFSCluster()}
     */
    public MiniDFSCluster startMiniDFSCluster(int servers, final String hosts[]) throws Exception {

        // Check that there is not already a cluster running
        isRunningCluster();

        // Initialize the local directory used by the MiniDFS
        if (clusterTestDir == null) {
            setupClusterTestDir();
        }

        // We have to set this property as it is used by MiniCluster
        System.setProperty(TEST_DIRECTORY_KEY, this.clusterTestDir.toString());

        // Some tests also do this:
        //  System.getProperty("test.cache.data", "build/test/cache");
        // It's also deprecated
        System.setProperty("test.cache.data", this.clusterTestDir.toString());

        // Ok, now we can start

        // Lily change: first argument changed from 0 to 8020
        // Lily change: let the formatting of NameNode and DataNodes depend on whether the dir is empty
        boolean format = this.clusterTestDir.list().length == 0;

        this.dfsCluster = new MiniDFSCluster(8020 /* Lily change */, this.conf, servers, format /* Lily change */,
                true, true, null, null, hosts, null);

        // Set this just-started cluster as our filesystem.
        FileSystem fs = this.dfsCluster.getFileSystem();
        this.conf.set("fs.defaultFS", fs.getUri().toString());
        // Do old style too just to be safe.
        // Lily change: commented this out to avoid warnings
        // this.conf.set("fs.default.name", fs.getUri().toString());

        // Wait for the cluster to be totally up
        this.dfsCluster.waitClusterUp();

        return this.dfsCluster;
    }

    /**
     * Shuts down instance created by call to {@link #startMiniDFSCluster(int)}
     * or does nothing.
     */
    public void shutdownMiniDFSCluster() throws Exception {
        if (this.dfsCluster != null) {
            // The below throws an exception per dn, AsynchronousCloseException.
            this.dfsCluster.shutdown();
            dfsCluster = null;
        }

    }

    /**
     * Call this if you only want a zk cluster.
     *
     * @return zk cluster started.
     * @see #startMiniZKCluster() if you want zk + dfs + hbase mini cluster.
     * @see #shutdownMiniZKCluster()
     */
    public MiniZooKeeperCluster startMiniZKCluster() throws Exception {
        return startMiniZKCluster(1);
    }

    /**
     * Call this if you only want a zk cluster.
     *
     * @return zk cluster started.
     * @see #startMiniZKCluster() if you want zk + dfs + hbase mini cluster.
     * @see #shutdownMiniZKCluster()
     */
    public MiniZooKeeperCluster startMiniZKCluster(int zooKeeperServerNum) throws Exception {
        File zkClusterFile = new File(getClusterTestDir().toString());
        return startMiniZKCluster(zkClusterFile, zooKeeperServerNum);
    }

    private MiniZooKeeperCluster startMiniZKCluster(final File dir) throws Exception {
        return startMiniZKCluster(dir, 1);
    }

    private MiniZooKeeperCluster startMiniZKCluster(final File dir, int zooKeeperServerNum) throws Exception {
        if (this.zkCluster != null) {
            throw new IOException("Cluster already running at " + dir);
        }
        this.passedZkCluster = false;
        this.zkCluster = new MiniZooKeeperCluster(this.getConfiguration());
        // Lily change: set client port
        this.zkCluster.setDefaultClientPort(2181);
        int clientPort = this.zkCluster.startup(dir, zooKeeperServerNum);
        this.conf.set(HConstants.ZOOKEEPER_CLIENT_PORT, Integer.toString(clientPort));
        return this.zkCluster;
    }

    /**
     * Shuts down zk cluster created by call to {@link #startMiniZKCluster(File)}
     * or does nothing.
     *
     * @see #startMiniZKCluster()
     */
    public void shutdownMiniZKCluster() throws IOException {
        if (this.zkCluster != null) {
            this.zkCluster.shutdown();
            this.zkCluster = null;
        }
    }

    /**
     * Start up a minicluster of hbase, dfs, and zookeeper.
     *
     * @return Mini hbase cluster instance created.
     * @see {@link #shutdownMiniDFSCluster()}
     */
    public MiniHBaseCluster startMiniCluster() throws Exception {
        return startMiniCluster(1, 1);
    }

    /**
     * Start up a minicluster of hbase, optionally dfs, and zookeeper.
     * Modifies Configuration.  Homes the cluster data directory under a random
     * subdirectory in a directory under System property test.build.data.
     * Directory is cleaned up on exit.
     *
     * @param numSlaves Number of slaves to start up.  We'll start this many
     *                  datanodes and regionservers.  If numSlaves is > 1, then make sure
     *                  hbase.regionserver.info.port is -1 (i.e. no ui per regionserver) otherwise
     *                  bind errors.
     * @return Mini hbase cluster instance created.
     * @see {@link #shutdownMiniCluster()}
     */
    public MiniHBaseCluster startMiniCluster(final int numSlaves) throws Exception {
        return startMiniCluster(1, numSlaves);
    }

    /**
     * start minicluster
     *
     * @return Mini hbase cluster instance created.
     * @see {@link #shutdownMiniCluster()}
     */
    public MiniHBaseCluster startMiniCluster(final int numMasters, final int numSlaves) throws Exception {
        return startMiniCluster(numMasters, numSlaves, null);
    }

    /**
     * Start up a minicluster of hbase, optionally dfs, and zookeeper.
     * Modifies Configuration.  Homes the cluster data directory under a random
     * subdirectory in a directory under System property test.build.data.
     * Directory is cleaned up on exit.
     *
     * @param numMasters    Number of masters to start up.  We'll start this many
     *                      hbase masters.  If numMasters > 1, you can find the active/primary master
     *                      with {@link MiniHBaseCluster#getMaster()}.
     * @param numSlaves     Number of slaves to start up.  We'll start this many
     *                      regionservers. If dataNodeHosts == null, this also indicates the number of
     *                      datanodes to start. If dataNodeHosts != null, the number of datanodes is
     *                      based on dataNodeHosts.length.
     *                      If numSlaves is > 1, then make sure
     *                      hbase.regionserver.info.port is -1 (i.e. no ui per regionserver) otherwise
     *                      bind errors.
     * @param dataNodeHosts hostnames DNs to run on.
     *                      This is useful if you want to run datanode on distinct hosts for things
     *                      like HDFS block location verification.
     *                      If you start MiniDFSCluster without host names,
     *                      all instances of the datanodes will have the same host name.
     * @return Mini hbase cluster instance created.
     * @see {@link #shutdownMiniCluster()}
     */
    public MiniHBaseCluster startMiniCluster(final int numMasters, final int numSlaves,
            final String[] dataNodeHosts) throws Exception {
        int numDataNodes = numSlaves;
        if (dataNodeHosts != null && dataNodeHosts.length != 0) {
            numDataNodes = dataNodeHosts.length;
        }

        LOG.info("Starting up minicluster with " + numMasters + " master(s) and " + numSlaves
                + " regionserver(s) and " + numDataNodes + " datanode(s)");

        // If we already put up a cluster, fail.
        isRunningCluster();

        // Bring up mini dfs cluster. This spews a bunch of warnings about missing
        // scheme. Complaints are 'Scheme is undefined for build/test/data/dfs/name1'.
        startMiniDFSCluster(numDataNodes, dataNodeHosts);

        // Start up a zk cluster.
        if (this.zkCluster == null) {
            startMiniZKCluster(clusterTestDir);
        }

        // Start the MiniHBaseCluster
        return startMiniHBaseCluster(numMasters, numSlaves);
    }

    /**
     * Starts up mini hbase cluster.  Usually used after call to
     * {@link #startMiniCluster(int, int)} when doing stepped startup of clusters.
     * Usually you won't want this.  You'll usually want {@link #startMiniCluster()}.
     *
     * @return Reference to the hbase mini hbase cluster.
     * @see {@link #startMiniCluster()}
     */
    public MiniHBaseCluster startMiniHBaseCluster(final int numMasters, final int numSlaves)
            throws IOException, InterruptedException {
        // Now do the mini hbase cluster.  Set the hbase.rootdir in config.
        createRootDir();

        // These settings will make the server waits until this exact number of
        // regions servers are connected.
        if (conf.getInt(ServerManager.WAIT_ON_REGIONSERVERS_MINTOSTART, -1) == -1) {
            conf.setInt(ServerManager.WAIT_ON_REGIONSERVERS_MINTOSTART, numSlaves);
        }
        if (conf.getInt(ServerManager.WAIT_ON_REGIONSERVERS_MAXTOSTART, -1) == -1) {
            conf.setInt(ServerManager.WAIT_ON_REGIONSERVERS_MAXTOSTART, numSlaves);
        }

        Configuration c = new Configuration(this.conf);
        this.hbaseCluster = new MiniHBaseCluster(c, numMasters, numSlaves);
        // Don't leave here till we've done a successful scan of the .META.
        HTable t = new HTable(c, HConstants.META_TABLE_NAME);
        ResultScanner s = t.getScanner(new Scan());
        while (s.next() != null) {
            continue;
        }
        s.close();
        t.close();

        // Lily change: fix wrong master filesystem
        // TODO Remove this once HBASE-6441 is available in the currently-used version of HBase
        Configuration masterConf = hbaseCluster.getMaster().getConfiguration();
        FileSystem fs = FileSystem.get(masterConf);
        fs.setConf(masterConf);

        getHBaseAdmin(); // create immediately the hbaseAdmin
        LOG.info("Minicluster is up");
        return (MiniHBaseCluster) this.hbaseCluster;
    }

    /**
     * Starts the hbase cluster up again after shutting it down previously in a
     * test.  Use this if you want to keep dfs/zk up and just stop/start hbase.
     *
     * @param servers number of region servers
     */
    public void restartHBaseCluster(int servers) throws IOException, InterruptedException {
        this.hbaseCluster = new MiniHBaseCluster(this.conf, servers);
        // Don't leave here till we've done a successful scan of the .META.
        HTable t = new HTable(new Configuration(this.conf), HConstants.META_TABLE_NAME);
        ResultScanner s = t.getScanner(new Scan());
        while (s.next() != null) {
            // do nothing
        }
        LOG.info("HBase has been restarted");
        s.close();
        t.close();
    }

    /**
     * @return Current mini hbase cluster. Only has something in it after a call
     *         to {@link #startMiniCluster()}.
     * @see #startMiniCluster()
     */
    public MiniHBaseCluster getMiniHBaseCluster() {
        if (this.hbaseCluster instanceof MiniHBaseCluster) {
            return (MiniHBaseCluster) this.hbaseCluster;
        }
        throw new RuntimeException(hbaseCluster + " not an instance of " + MiniHBaseCluster.class.getName());
    }

    /**
     * Stops mini hbase, zk, and hdfs clusters.
     *
     * @see {@link #startMiniCluster(int)}
     */
    public void shutdownMiniCluster() throws Exception {
        LOG.info("Shutting down minicluster");
        shutdownMiniHBaseCluster();
        if (!this.passedZkCluster) {
            shutdownMiniZKCluster();
        }
        shutdownMiniDFSCluster();

        // Lily Change: if clearData is true, delete the clusterTestBuildDir
        if (clearData) {
            // Clean up our directory.
            if (this.clusterTestDir != null && this.clusterTestDir.exists()) {
                // Need to use deleteDirectory because File.delete required dir is empty.
                if (!FSUtils.deleteDirectory(FileSystem.getLocal(this.conf),
                        new Path(this.clusterTestDir.toString()))) {
                    LOG.warn("Failed delete of " + this.clusterTestDir.toString());
                }
                this.clusterTestDir = null;
            }
        }
        LOG.info("Minicluster is down");
    }

    /**
     * Shutdown HBase mini cluster.  Does not shutdown zk or dfs if running.
     */
    public void shutdownMiniHBaseCluster() throws IOException {
        if (hbaseAdmin != null) {
            hbaseAdmin.close();
            hbaseAdmin = null;
        }
        // unset the configuration for MIN and MAX RS to start
        conf.setInt(ServerManager.WAIT_ON_REGIONSERVERS_MINTOSTART, -1);
        conf.setInt(ServerManager.WAIT_ON_REGIONSERVERS_MAXTOSTART, -1);
        if (this.hbaseCluster != null) {
            this.hbaseCluster.shutdown();
            // Wait till hbase is down before going on to shutdown zk.
            this.hbaseCluster.waitUntilShutDown();
            this.hbaseCluster = null;
        }
    }

    /**
     * Returns the path to the default root dir the minicluster uses.
     * Note: this does not cause the root dir to be created.
     *
     * @return Fully qualified path for the default hbase root dir
     */
    public Path getDefaultRootDirPath() throws IOException {
        FileSystem fs = FileSystem.get(this.conf);
        return new Path(fs.makeQualified(fs.getHomeDirectory()), "hbase");
    }

    /**
     * Creates an hbase rootdir in user home directory.  Also creates hbase
     * version file.  Normally you won't make use of this method.  Root hbasedir
     * is created for you as part of mini cluster startup.  You'd only use this
     * method if you were doing manual operation.
     *
     * @return Fully qualified path to hbase root dir
     */
    public Path createRootDir() throws IOException {
        FileSystem fs = FileSystem.get(this.conf);
        // Lily change: create "hbase" subdirectory under home directory
        // to serve as hbaseRootdir. The home directory can contain other
        // directories and files, which are not necessarily hbase tables.
        // For instance a 'target' dir created by MiniMRCluster.
        // Cfr. HBASE-5317 and HBASE-4025
        Path hbaseRootdir = fs.makeQualified(new Path(fs.getHomeDirectory(), "hbase"));
        this.conf.set(HConstants.HBASE_DIR, hbaseRootdir.toString());
        fs.mkdirs(hbaseRootdir);
        FSUtils.setVersion(fs, hbaseRootdir);
        return hbaseRootdir;
    }

    /**
     * Flushes all caches in the mini hbase cluster
     */
    public void flush() throws IOException {
        getMiniHBaseCluster().flushcache();
    }

    /**
     * Flushes all caches in the mini hbase cluster
     */
    public void flush(byte[] tableName) throws IOException {
        getMiniHBaseCluster().flushcache(tableName);
    }

    /**
     * Compact all regions in the mini hbase cluster
     */
    public void compact(boolean major) throws IOException {
        getMiniHBaseCluster().compact(major);
    }

    /**
     * Compact all of a table's reagion in the mini hbase cluster
     */
    public void compact(byte[] tableName, boolean major) throws IOException {
        getMiniHBaseCluster().compact(tableName, major);
    }

    /**
     * Create a table.
     *
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[] family) throws IOException {
        return createTable(tableName, new byte[][] { family });
    }

    /**
     * Create a table.
     *
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[][] families) throws IOException {
        return createTable(tableName, families, new Configuration(getConfiguration()));
    }

    public HTable createTable(byte[] tableName, byte[][] families, int numVersions, byte[] startKey, byte[] endKey,
            int numRegions) throws IOException {
        HTableDescriptor desc = new HTableDescriptor(tableName);
        for (byte[] family : families) {
            HColumnDescriptor hcd = new HColumnDescriptor(family).setMaxVersions(numVersions);
            desc.addFamily(hcd);
        }
        getHBaseAdmin().createTable(desc, startKey, endKey, numRegions);
        return new HTable(getConfiguration(), tableName);
    }

    /**
     * Create a table.
     *
     * @param c Configuration to use
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[][] families, final Configuration c) throws IOException {
        HTableDescriptor desc = new HTableDescriptor(tableName);
        for (byte[] family : families) {
            desc.addFamily(new HColumnDescriptor(family));
        }
        getHBaseAdmin().createTable(desc);
        return new HTable(c, tableName);
    }

    /**
     * Create a table.
     *
     * @param c Configuration to use
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[][] families, final Configuration c, int numVersions)
            throws IOException {
        HTableDescriptor desc = new HTableDescriptor(tableName);
        for (byte[] family : families) {
            HColumnDescriptor hcd = new HColumnDescriptor(family).setMaxVersions(numVersions);
            desc.addFamily(hcd);
        }
        getHBaseAdmin().createTable(desc);
        return new HTable(c, tableName);
    }

    /**
     * Create a table.
     *
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[] family, int numVersions) throws IOException {
        return createTable(tableName, new byte[][] { family }, numVersions);
    }

    /**
     * Create a table.
     *
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[][] families, int numVersions) throws IOException {
        HTableDescriptor desc = new HTableDescriptor(tableName);
        for (byte[] family : families) {
            HColumnDescriptor hcd = new HColumnDescriptor(family).setMaxVersions(numVersions);
            desc.addFamily(hcd);
        }
        getHBaseAdmin().createTable(desc);
        return new HTable(new Configuration(getConfiguration()), tableName);
    }

    /**
     * Create a table.
     *
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[][] families, int numVersions, int blockSize)
            throws IOException {
        HTableDescriptor desc = new HTableDescriptor(tableName);
        for (byte[] family : families) {
            HColumnDescriptor hcd = new HColumnDescriptor(family).setMaxVersions(numVersions)
                    .setBlocksize(blockSize);
            desc.addFamily(hcd);
        }
        getHBaseAdmin().createTable(desc);
        return new HTable(new Configuration(getConfiguration()), tableName);
    }

    /**
     * Create a table.
     *
     * @return An HTable instance for the created table.
     */
    public HTable createTable(byte[] tableName, byte[][] families, int[] numVersions) throws IOException {
        HTableDescriptor desc = new HTableDescriptor(tableName);
        int i = 0;
        for (byte[] family : families) {
            HColumnDescriptor hcd = new HColumnDescriptor(family).setMaxVersions(numVersions[i]);
            desc.addFamily(hcd);
            i++;
        }
        getHBaseAdmin().createTable(desc);
        return new HTable(new Configuration(getConfiguration()), tableName);
    }

    /**
     * Drop an existing table
     *
     * @param tableName existing table
     */
    public void deleteTable(byte[] tableName) throws IOException {
        try {
            getHBaseAdmin().disableTable(tableName);
        } catch (TableNotEnabledException e) {
            LOG.debug("Table: " + Bytes.toString(tableName) + " already disabled, so just deleting it.");
        }
        getHBaseAdmin().deleteTable(tableName);
    }

    /**
     * Provide an existing table name to truncate
     *
     * @param tableName existing table
     * @return HTable to that new table
     */
    public HTable truncateTable(byte[] tableName) throws IOException {
        HTable table = new HTable(getConfiguration(), tableName);
        Scan scan = new Scan();
        ResultScanner resScan = table.getScanner(scan);
        for (Result res : resScan) {
            Delete del = new Delete(res.getRow());
            table.delete(del);
        }
        resScan = table.getScanner(scan);
        resScan.close();
        return table;
    }

    /**
     * Load table with rows from 'aaa' to 'zzz'.
     *
     * @param t Table
     * @param f Family
     * @return Count of rows loaded.
     */
    public int loadTable(final HTable t, final byte[] f) throws IOException {
        t.setAutoFlush(false);
        byte[] k = new byte[3];
        int rowCount = 0;
        for (byte b1 = 'a'; b1 <= 'z'; b1++) {
            for (byte b2 = 'a'; b2 <= 'z'; b2++) {
                for (byte b3 = 'a'; b3 <= 'z'; b3++) {
                    k[0] = b1;
                    k[1] = b2;
                    k[2] = b3;
                    Put put = new Put(k);
                    put.add(f, null, k);
                    t.put(put);
                    rowCount++;
                }
            }
        }
        t.flushCommits();
        return rowCount;
    }

    /**
     * Load table of multiple column families with rows from 'aaa' to 'zzz'.
     *
     * @param t Table
     * @param f Array of Families to load
     * @return Count of rows loaded.
     */
    public int loadTable(final HTable t, final byte[][] f) throws IOException {
        t.setAutoFlush(false);
        byte[] k = new byte[3];
        int rowCount = 0;
        for (byte b1 = 'a'; b1 <= 'z'; b1++) {
            for (byte b2 = 'a'; b2 <= 'z'; b2++) {
                for (byte b3 = 'a'; b3 <= 'z'; b3++) {
                    k[0] = b1;
                    k[1] = b2;
                    k[2] = b3;
                    Put put = new Put(k);
                    for (int i = 0; i < f.length; i++) {
                        put.add(f[i], null, k);
                    }
                    t.put(put);
                    rowCount++;
                }
            }
        }
        t.flushCommits();
        return rowCount;
    }

    /**
     * Load region with rows from 'aaa' to 'zzz'.
     *
     * @param r Region
     * @param f Family
     * @return Count of rows loaded.
     */
    public int loadRegion(final HRegion r, final byte[] f) throws IOException {
        return loadRegion(r, f, false);
    }

    /**
     * Load region with rows from 'aaa' to 'zzz'.
     *
     * @param r     Region
     * @param f     Family
     * @param flush flush the cache if true
     * @return Count of rows loaded.
     */
    public int loadRegion(final HRegion r, final byte[] f, final boolean flush) throws IOException {
        byte[] k = new byte[3];
        int rowCount = 0;
        for (byte b1 = 'a'; b1 <= 'z'; b1++) {
            for (byte b2 = 'a'; b2 <= 'z'; b2++) {
                for (byte b3 = 'a'; b3 <= 'z'; b3++) {
                    k[0] = b1;
                    k[1] = b2;
                    k[2] = b3;
                    Put put = new Put(k);
                    put.add(f, null, k);
                    if (r.getLog() == null)
                        put.setWriteToWAL(false);
                    r.put(put);
                    rowCount++;
                }
            }
            if (flush) {
                r.flushcache();
            }
        }
        return rowCount;
    }

    /**
     * Return the number of rows in the given table.
     */
    public int countRows(final HTable table) throws IOException {
        Scan scan = new Scan();
        ResultScanner results = table.getScanner(scan);
        int count = 0;
        for (@SuppressWarnings("unused")
        Result res : results) {
            count++;
        }
        results.close();
        return count;
    }

    public int countRows(final HTable table, final byte[]... families) throws IOException {
        Scan scan = new Scan();
        for (byte[] family : families) {
            scan.addFamily(family);
        }
        ResultScanner results = table.getScanner(scan);
        int count = 0;
        for (@SuppressWarnings("unused")
        Result res : results) {
            count++;
        }
        results.close();
        return count;
    }

    /**
     * Return an md5 digest of the entire contents of a table.
     */
    public String checksumRows(final HTable table) throws Exception {
        Scan scan = new Scan();
        ResultScanner results = table.getScanner(scan);
        MessageDigest digest = MessageDigest.getInstance("MD5");
        for (Result res : results) {
            digest.update(res.getRow());
        }
        results.close();
        return digest.toString();
    }

    /**
     * Creates many regions names "aaa" to "zzz".
     *
     * @param table        The table to use for the data.
     * @param columnFamily The family to insert the data into.
     * @return count of regions created.
     * @throws IOException When creating the regions fails.
     */
    public int createMultiRegions(HTable table, byte[] columnFamily) throws IOException {
        return createMultiRegions(table, columnFamily, true);
    }

    public static final byte[][] KEYS = { HConstants.EMPTY_BYTE_ARRAY, Bytes.toBytes("bbb"), Bytes.toBytes("ccc"),
            Bytes.toBytes("ddd"), Bytes.toBytes("eee"), Bytes.toBytes("fff"), Bytes.toBytes("ggg"),
            Bytes.toBytes("hhh"), Bytes.toBytes("iii"), Bytes.toBytes("jjj"), Bytes.toBytes("kkk"),
            Bytes.toBytes("lll"), Bytes.toBytes("mmm"), Bytes.toBytes("nnn"), Bytes.toBytes("ooo"),
            Bytes.toBytes("ppp"), Bytes.toBytes("qqq"), Bytes.toBytes("rrr"), Bytes.toBytes("sss"),
            Bytes.toBytes("ttt"), Bytes.toBytes("uuu"), Bytes.toBytes("vvv"), Bytes.toBytes("www"),
            Bytes.toBytes("xxx"), Bytes.toBytes("yyy") };

    public static final byte[][] KEYS_FOR_HBA_CREATE_TABLE = { Bytes.toBytes("bbb"), Bytes.toBytes("ccc"),
            Bytes.toBytes("ddd"), Bytes.toBytes("eee"), Bytes.toBytes("fff"), Bytes.toBytes("ggg"),
            Bytes.toBytes("hhh"), Bytes.toBytes("iii"), Bytes.toBytes("jjj"), Bytes.toBytes("kkk"),
            Bytes.toBytes("lll"), Bytes.toBytes("mmm"), Bytes.toBytes("nnn"), Bytes.toBytes("ooo"),
            Bytes.toBytes("ppp"), Bytes.toBytes("qqq"), Bytes.toBytes("rrr"), Bytes.toBytes("sss"),
            Bytes.toBytes("ttt"), Bytes.toBytes("uuu"), Bytes.toBytes("vvv"), Bytes.toBytes("www"),
            Bytes.toBytes("xxx"), Bytes.toBytes("yyy"), Bytes.toBytes("zzz") };

    /**
     * Creates many regions names "aaa" to "zzz".
     *
     * @param table        The table to use for the data.
     * @param columnFamily The family to insert the data into.
     * @param cleanupFS    True if a previous region should be remove from the FS
     * @return count of regions created.
     * @throws IOException When creating the regions fails.
     */
    public int createMultiRegions(HTable table, byte[] columnFamily, boolean cleanupFS) throws IOException {
        return createMultiRegions(getConfiguration(), table, columnFamily, KEYS, cleanupFS);
    }

    /**
     * Creates the specified number of regions in the specified table.
     */
    public int createMultiRegions(final Configuration c, final HTable table, final byte[] family, int numRegions)
            throws IOException {
        if (numRegions < 3)
            throw new IOException("Must create at least 3 regions");
        byte[] startKey = Bytes.toBytes("aaaaa");
        byte[] endKey = Bytes.toBytes("zzzzz");
        byte[][] splitKeys = Bytes.split(startKey, endKey, numRegions - 3);
        byte[][] regionStartKeys = new byte[splitKeys.length + 1][];
        for (int i = 0; i < splitKeys.length; i++) {
            regionStartKeys[i + 1] = splitKeys[i];
        }
        regionStartKeys[0] = HConstants.EMPTY_BYTE_ARRAY;
        return createMultiRegions(c, table, family, regionStartKeys);
    }

    public int createMultiRegions(final Configuration c, final HTable table, final byte[] columnFamily,
            byte[][] startKeys) throws IOException {
        return createMultiRegions(c, table, columnFamily, startKeys, true);
    }

    public int createMultiRegions(final Configuration c, final HTable table, final byte[] columnFamily,
            byte[][] startKeys, boolean cleanupFS) throws IOException {
        Arrays.sort(startKeys, Bytes.BYTES_COMPARATOR);
        HTable meta = new HTable(c, HConstants.META_TABLE_NAME);
        HTableDescriptor htd = table.getTableDescriptor();
        if (!htd.hasFamily(columnFamily)) {
            HColumnDescriptor hcd = new HColumnDescriptor(columnFamily);
            htd.addFamily(hcd);
        }
        // remove empty region - this is tricky as the mini cluster during the test
        // setup already has the "<tablename>,,123456789" row with an empty start
        // and end key. Adding the custom regions below adds those blindly,
        // including the new start region from empty to "bbb". lg
        List<byte[]> rows = getMetaTableRows(htd.getName());
        String regionToDeleteInFS = table.getRegionsInRange(Bytes.toBytes(""), Bytes.toBytes("")).get(0)
                .getRegionInfo().getEncodedName();
        List<HRegionInfo> newRegions = new ArrayList<HRegionInfo>(startKeys.length);
        // add custom ones
        int count = 0;
        for (int i = 0; i < startKeys.length; i++) {
            int j = (i + 1) % startKeys.length;
            HRegionInfo hri = new HRegionInfo(table.getTableName(), startKeys[i], startKeys[j]);
            Put put = new Put(hri.getRegionName());
            put.add(HConstants.CATALOG_FAMILY, HConstants.REGIONINFO_QUALIFIER, Writables.getBytes(hri));
            meta.put(put);
            LOG.info("createMultiRegions: inserted " + hri.toString());
            newRegions.add(hri);
            count++;
        }
        // see comment above, remove "old" (or previous) single region
        for (byte[] row : rows) {
            LOG.info("createMultiRegions: deleting meta row -> " + Bytes.toStringBinary(row));
            meta.delete(new Delete(row));
        }
        if (cleanupFS) {
            // see HBASE-7417 - this confused TestReplication
            // remove the "old" region from FS
            Path tableDir = new Path(getDefaultRootDirPath().toString() + System.getProperty("file.separator")
                    + htd.getNameAsString() + System.getProperty("file.separator") + regionToDeleteInFS);
            getDFSCluster().getFileSystem().delete(tableDir);
        }
        // flush cache of regions
        HConnection conn = table.getConnection();
        conn.clearRegionCache();
        // assign all the new regions IF table is enabled.
        HBaseAdmin admin = getHBaseAdmin();
        if (admin.isTableEnabled(table.getTableName())) {
            for (HRegionInfo hri : newRegions) {
                admin.assign(hri.getRegionName());
            }
        }

        meta.close();

        return count;
    }

    /**
     * Create rows in META for regions of the specified table with the specified
     * start keys.  The first startKey should be a 0 length byte array if you
     * want to form a proper range of regions.
     *
     * @return list of region info for regions added to meta
     */
    public List<HRegionInfo> createMultiRegionsInMeta(final Configuration conf, final HTableDescriptor htd,
            byte[][] startKeys) throws IOException {
        HTable meta = new HTable(conf, HConstants.META_TABLE_NAME);
        Arrays.sort(startKeys, Bytes.BYTES_COMPARATOR);
        List<HRegionInfo> newRegions = new ArrayList<HRegionInfo>(startKeys.length);
        // add custom ones
        for (int i = 0; i < startKeys.length; i++) {
            int j = (i + 1) % startKeys.length;
            HRegionInfo hri = new HRegionInfo(htd.getName(), startKeys[i], startKeys[j]);
            Put put = new Put(hri.getRegionName());
            put.add(HConstants.CATALOG_FAMILY, HConstants.REGIONINFO_QUALIFIER, Writables.getBytes(hri));
            meta.put(put);
            LOG.info("createMultiRegionsInMeta: inserted " + hri.toString());
            newRegions.add(hri);
        }

        meta.close();
        return newRegions;
    }

    /**
     * Returns all rows from the .META. table.
     *
     * @throws IOException When reading the rows fails.
     */
    public List<byte[]> getMetaTableRows() throws IOException {
        // TODO: Redo using MetaReader class
        HTable t = new HTable(new Configuration(this.conf), HConstants.META_TABLE_NAME);
        List<byte[]> rows = new ArrayList<byte[]>();
        ResultScanner s = t.getScanner(new Scan());
        for (Result result : s) {
            LOG.info("getMetaTableRows: row -> " + Bytes.toStringBinary(result.getRow()));
            rows.add(result.getRow());
        }
        s.close();
        t.close();
        return rows;
    }

    /**
     * Returns all rows from the .META. table for a given user table
     *
     * @throws IOException When reading the rows fails.
     */
    public List<byte[]> getMetaTableRows(byte[] tableName) throws IOException {
        // TODO: Redo using MetaReader.
        HTable t = new HTable(new Configuration(this.conf), HConstants.META_TABLE_NAME);
        List<byte[]> rows = new ArrayList<byte[]>();
        ResultScanner s = t.getScanner(new Scan());
        for (Result result : s) {
            byte[] val = result.getValue(HConstants.CATALOG_FAMILY, HConstants.REGIONINFO_QUALIFIER);
            if (val == null) {
                LOG.error("No region info for row " + Bytes.toString(result.getRow()));
                // TODO figure out what to do for this new hosed case.
                continue;
            }
            HRegionInfo info = Writables.getHRegionInfo(val);
            if (Bytes.compareTo(info.getTableName(), tableName) == 0) {
                LOG.info("getMetaTableRows: row -> " + Bytes.toStringBinary(result.getRow()) + info);
                rows.add(result.getRow());
            }
        }
        s.close();
        t.close();
        return rows;
    }

    /**
     * Tool to get the reference to the region server object that holds the
     * region of the specified user table.
     * It first searches for the meta rows that contain the region of the
     * specified table, then gets the index of that RS, and finally retrieves
     * the RS's reference.
     *
     * @param tableName user table to lookup in .META.
     * @return region server that holds it, null if the row doesn't exist
     */
    public HRegionServer getRSForFirstRegionInTable(byte[] tableName) throws IOException {
        List<byte[]> metaRows = getMetaTableRows(tableName);
        if (metaRows == null || metaRows.isEmpty()) {
            return null;
        }
        LOG.debug("Found " + metaRows.size() + " rows for table " + Bytes.toString(tableName));
        byte[] firstrow = metaRows.get(0);
        LOG.debug("FirstRow=" + Bytes.toString(firstrow));
        int index = getMiniHBaseCluster().getServerWith(firstrow);
        return getMiniHBaseCluster().getRegionServerThreads().get(index).getRegionServer();
    }

    /**
     * Starts a <code>MiniMRCluster</code> with a default number of
     * <code>TaskTracker</code>'s.
     *
     * @throws IOException When starting the cluster fails.
     */
    public void startMiniMapReduceCluster() throws IOException {
        startMiniMapReduceCluster(2);
    }

    /**
     * Starts a <code>MiniMRCluster</code>.
     *
     * @param servers The number of <code>TaskTracker</code>'s to start.
     * @throws IOException When starting the cluster fails.
     */
    public void startMiniMapReduceCluster(final int servers) throws IOException {
        LOG.info("Starting mini mapreduce cluster...");
        // These are needed for the new and improved Map/Reduce framework
        Configuration c = getConfiguration();
        String logDir = c.get("hadoop.log.dir");
        String tmpDir = c.get("hadoop.tmp.dir");
        if (logDir == null) {
            logDir = tmpDir;
        }
        System.setProperty("hadoop.log.dir", logDir);
        c.set("mapred.output.dir", tmpDir);

        // Lily change: fix port number + set host name (+ fix # of servers to 1) + use custom jobconf so that
        // hadoop dir properties get passed on.
        JobConf jobConf = new JobConf(conf);
        mrCluster = new MiniMRCluster(8021, 0, 1, FileSystem.get(conf).getUri().toString(), 1, null,
                new String[] { "localhost" }, null, jobConf);
        // End Lily change

        LOG.info("Mini mapreduce cluster started");
        JobConf mrClusterJobConf = mrCluster.createJobConf();

        // In hadoop2, YARN/MR2 starts a mini cluster with its own conf instance and updates settings.
        // Our HBase MR jobs need several of these settings in order to properly run.  So we copy the
        // necessary config properties here.  YARN-129 required adding a few properties.
        c.set("mapred.job.tracker", mrClusterJobConf.get("mapred.job.tracker"));
        /* this for mrv2 support */
        conf.set("mapreduce.framework.name", "yarn");
        conf.setBoolean("yarn.is.minicluster", true);
        String rmAdress = mrClusterJobConf.get("yarn.resourcemanager.address");
        if (rmAdress != null) {
            conf.set("yarn.resourcemanager.address", rmAdress);
        }
        String schedulerAdress = mrClusterJobConf.get("yarn.resourcemanager.scheduler.address");
        if (schedulerAdress != null) {
            conf.set("yarn.resourcemanager.scheduler.address", schedulerAdress);
        }
    }

    /**
     * Stops the previously started <code>MiniMRCluster</code>.
     */
    public void shutdownMiniMapReduceCluster() {
        LOG.info("Stopping mini mapreduce cluster...");
        if (mrCluster != null) {
            // Lily change
            // Disable the shutdown as it sometimes hangs for a longer while, until a timeout passed.
            // This is not nice for launch-test-lily users.
            // At the time of CDH4u0, this was the related stack:
            //     at java.lang.Object.wait(Native Method)
            //     - waiting on <0x00000007dddc1b90> (a org.apache.hadoop.ipc.Client$Call)
            //     at java.lang.Object.wait(Object.java:485)
            //     at org.apache.hadoop.ipc.Client.call(Client.java:1147)
            //     - locked <0x00000007dddc1b90> (a org.apache.hadoop.ipc.Client$Call)
            //     at org.apache.hadoop.ipc.WritableRpcEngine$Invoker.invoke(WritableRpcEngine.java:224)
            //     at org.apache.hadoop.mapred.$Proxy21.heartbeat(Unknown Source)
            //     at org.apache.hadoop.mapred.TaskTracker.transmitHeartBeat(TaskTracker.java:2030)
            //     at org.apache.hadoop.mapred.TaskTracker.offerService(TaskTracker.java:1824)
            //     at org.apache.hadoop.mapred.TaskTracker.run(TaskTracker.java:2698)
            //     at org.apache.hadoop.mapred.MiniMRCluster$TaskTrackerRunner.run(MiniMRCluster.java:220)
            //     at java.lang.Thread.run(Thread.java:662)

            // mrCluster.shutdown();
            mrCluster = null;
        }
        // Restore configuration to point to local jobtracker
        conf.set("mapred.job.tracker", "local");
        LOG.info("Mini mapreduce cluster stopped");
    }

    /**
     * Switches the logger for the given class to DEBUG level.
     *
     * @param clazz The class for which to switch to debug logging.
     */
    public void enableDebug(Class<?> clazz) {
        Log l = LogFactory.getLog(clazz);
        if (l instanceof Log4JLogger) {
            ((Log4JLogger) l).getLogger().setLevel(org.apache.log4j.Level.DEBUG);
        } else if (l instanceof Jdk14Logger) {
            ((Jdk14Logger) l).getLogger().setLevel(java.util.logging.Level.ALL);
        }
    }

    /**
     * Expire the Master's session
     */
    public void expireMasterSession() throws Exception {
        HMaster master = getMiniHBaseCluster().getMaster();
        expireSession(master.getZooKeeper(), false);
    }

    /**
     * Expire a region server's session
     *
     * @param index which RS
     */
    public void expireRegionServerSession(int index) throws Exception {
        HRegionServer rs = getMiniHBaseCluster().getRegionServer(index);
        expireSession(rs.getZooKeeper(), false);
        decrementMinRegionServerCount();
    }

    private void decrementMinRegionServerCount() {
        // decrement the count for this.conf, for newly spwaned master
        // this.hbaseCluster shares this configuration too
        decrementMinRegionServerCount(getConfiguration());

        // each master thread keeps a copy of configuration
        for (MasterThread master : getHBaseCluster().getMasterThreads()) {
            decrementMinRegionServerCount(master.getMaster().getConfiguration());
        }
    }

    private void decrementMinRegionServerCount(Configuration conf) {
        int currentCount = conf.getInt(ServerManager.WAIT_ON_REGIONSERVERS_MINTOSTART, -1);
        if (currentCount != -1) {
            conf.setInt(ServerManager.WAIT_ON_REGIONSERVERS_MINTOSTART, Math.max(currentCount - 1, 1));
        }
    }

    /**
     * Expire a ZooKeeper session as recommended in ZooKeeper documentation
     * http://wiki.apache.org/hadoop/ZooKeeper/FAQ#A4
     * There are issues when doing this:
     * [1] http://www.mail-archive.com/dev@zookeeper.apache.org/msg01942.html
     * [2] https://issues.apache.org/jira/browse/ZOOKEEPER-1105
     *
     * @param nodeZK      - the ZK to make expiry
     * @param checkStatus - true to check if the we can create a HTable with the
     *                    current configuration.
     */
    public void expireSession(ZooKeeperWatcher nodeZK, boolean checkStatus) throws Exception {
        Configuration c = new Configuration(this.conf);
        String quorumServers = ZKConfig.getZKQuorumServersString(c);
        int sessionTimeout = 500;
        ZooKeeper zk = nodeZK.getRecoverableZooKeeper().getZooKeeper();
        byte[] password = zk.getSessionPasswd();
        long sessionID = zk.getSessionId();

        // Expiry seems to be asynchronous (see comment from P. Hunt in [1]),
        //  so we create a first watcher to be sure that the
        //  event was sent. We expect that if our watcher receives the event
        //  other watchers on the same machine will get is as well.
        // When we ask to close the connection, ZK does not close it before
        //  we receive all the events, so don't have to capture the event, just
        //  closing the connection should be enough.
        ZooKeeper monitor = new ZooKeeper(quorumServers, 1000, new org.apache.zookeeper.Watcher() {
            @Override
            public void process(WatchedEvent watchedEvent) {
                LOG.info("Monitor ZKW received event=" + watchedEvent);
            }
        }, sessionID, password);

        // Making it expire
        ZooKeeper newZK = new ZooKeeper(quorumServers, sessionTimeout, EmptyWatcher.instance, sessionID, password);
        newZK.close();
        LOG.info("ZK Closed Session 0x" + Long.toHexString(sessionID));

        // Now closing & waiting to be sure that the clients get it.
        monitor.close();

        if (checkStatus) {
            new HTable(new Configuration(conf), HConstants.META_TABLE_NAME).close();
        }
    }

    /**
     * Get the Mini HBase cluster.
     *
     * @return hbase cluster
     * @see #getHBaseClusterInterface()
     */
    public MiniHBaseCluster getHBaseCluster() {
        return getMiniHBaseCluster();
    }

    /**
     * Returns the HBaseCluster instance.
     * <p>Returned object can be any of the subclasses of HBaseCluster, and the
     * tests referring this should not assume that the cluster is a mini cluster or a
     * distributed one. If the test only works on a mini cluster, then specific
     * method {@link #getMiniHBaseCluster()} can be used instead w/o the
     * need to type-cast.
     */
    public HBaseCluster getHBaseClusterInterface() {
        //implementation note: we should rename this method as #getHBaseCluster(),
        //but this would require refactoring 90+ calls.
        return hbaseCluster;
    }

    /**
     * Returns a HBaseAdmin instance.
     * This instance is shared between HBaseTestingUtility intance users.
     * Don't close it, it will be closed automatically when the
     * cluster shutdowns
     *
     * @return The HBaseAdmin instance.
     */
    public synchronized HBaseAdmin getHBaseAdmin() throws IOException {
        if (hbaseAdmin == null) {
            hbaseAdmin = new HBaseAdmin(new Configuration(getConfiguration()));
        }
        return hbaseAdmin;
    }

    private HBaseAdmin hbaseAdmin = null;

    /**
     * Closes the named region.
     *
     * @param regionName The region to close.
     */
    public void closeRegion(String regionName) throws IOException {
        closeRegion(Bytes.toBytes(regionName));
    }

    /**
     * Closes the named region.
     *
     * @param regionName The region to close.
     */
    public void closeRegion(byte[] regionName) throws IOException {
        getHBaseAdmin().closeRegion(regionName, null);
    }

    /**
     * Closes the region containing the given row.
     *
     * @param row   The row to find the containing region.
     * @param table The table to find the region.
     */
    public void closeRegionByRow(String row, HTable table) throws IOException {
        closeRegionByRow(Bytes.toBytes(row), table);
    }

    /**
     * Closes the region containing the given row.
     *
     * @param row   The row to find the containing region.
     * @param table The table to find the region.
     */
    public void closeRegionByRow(byte[] row, HTable table) throws IOException {
        HRegionLocation hrl = table.getRegionLocation(row);
        closeRegion(hrl.getRegionInfo().getRegionName());
    }

    public MiniZooKeeperCluster getZkCluster() {
        return zkCluster;
    }

    public void setZkCluster(MiniZooKeeperCluster zkCluster) {
        this.passedZkCluster = true;
        this.zkCluster = zkCluster;
        conf.setInt(HConstants.ZOOKEEPER_CLIENT_PORT, zkCluster.getClientPort());
    }

    public MiniDFSCluster getDFSCluster() {
        return dfsCluster;
    }

    public void setDFSCluster(MiniDFSCluster cluster) throws IOException {
        if (dfsCluster != null && dfsCluster.isClusterUp()) {
            throw new IOException("DFSCluster is already running! Shut it down first.");
        }
        this.dfsCluster = cluster;
    }

    public FileSystem getTestFileSystem() throws IOException {
        return HFileSystem.get(conf);
    }

    /**
     * @return True if we removed the test dir
     */
    public boolean cleanupTestDir() throws IOException {
        if (dataTestDir == null) {
            return false;
        } else {
            boolean ret = deleteDir(getDataTestDir());
            dataTestDir = null;
            return ret;
        }
    }

    /**
     * @param subdir Test subdir name.
     * @return True if we removed the test dir
     */
    public boolean cleanupTestDir(final String subdir) throws IOException {
        if (dataTestDir == null) {
            return false;
        }
        return deleteDir(getDataTestDir(subdir));
    }

    /**
     * @param dir Directory to delete
     * @return True if we deleted it.
     */
    public boolean deleteDir(final Path dir) throws IOException {
        FileSystem fs = getTestFileSystem();
        if (fs.exists(dir)) {
            return fs.delete(getDataTestDir(), true);
        }
        return false;
    }

    public void waitTableAvailable(byte[] table, long timeoutMillis) throws InterruptedException, IOException {
        long startWait = System.currentTimeMillis();
        while (!getHBaseAdmin().isTableAvailable(table)) {
            assertTrue("Timed out waiting for table to become available " + Bytes.toStringBinary(table),
                    System.currentTimeMillis() - startWait < timeoutMillis);
            Thread.sleep(200);
        }
    }

    public void waitTableEnabled(byte[] table, long timeoutMillis) throws InterruptedException, IOException {
        long startWait = System.currentTimeMillis();
        while (!getHBaseAdmin().isTableAvailable(table) && !getHBaseAdmin().isTableEnabled(table)) {
            assertTrue("Timed out waiting for table to become available and enabled " + Bytes.toStringBinary(table),
                    System.currentTimeMillis() - startWait < timeoutMillis);
            Thread.sleep(200);
        }
    }

    /**
     * Make sure that at least the specified number of region servers
     * are running
     *
     * @param num minimum number of region servers that should be running
     * @return true if we started some servers
     */
    public boolean ensureSomeRegionServersAvailable(final int num) throws IOException {
        boolean startedServer = false;
        MiniHBaseCluster hbaseCluster = getMiniHBaseCluster();
        for (int i = hbaseCluster.getLiveRegionServerThreads().size(); i < num; ++i) {
            LOG.info("Started new server=" + hbaseCluster.startRegionServer());
            startedServer = true;
        }

        return startedServer;
    }

    /**
     * Make sure that at least the specified number of region servers
     * are running. We don't count the ones that are currently stopping or are
     * stopped.
     *
     * @param num minimum number of region servers that should be running
     * @return true if we started some servers
     */
    public boolean ensureSomeNonStoppedRegionServersAvailable(final int num) throws IOException {
        boolean startedServer = ensureSomeRegionServersAvailable(num);

        int nonStoppedServers = 0;
        for (JVMClusterUtil.RegionServerThread rst : getMiniHBaseCluster().getRegionServerThreads()) {

            HRegionServer hrs = rst.getRegionServer();
            if (hrs.isStopping() || hrs.isStopped()) {
                LOG.info("A region server is stopped or stopping:" + hrs);
            } else {
                nonStoppedServers++;
            }
        }
        for (int i = nonStoppedServers; i < num; ++i) {
            LOG.info("Started new server=" + getMiniHBaseCluster().startRegionServer());
            startedServer = true;
        }
        return startedServer;
    }

    /**
     * This method clones the passed <code>c</code> configuration setting a new
     * user into the clone.  Use it getting new instances of FileSystem.  Only
     * works for DistributedFileSystem.
     *
     * @param c                     Initial configuration
     * @param differentiatingSuffix Suffix to differentiate this user from others.
     * @return A new configuration instance with a different user set into it.
     */
    public static User getDifferentUser(final Configuration c, final String differentiatingSuffix)
            throws IOException {
        FileSystem currentfs = FileSystem.get(c);
        if (!(currentfs instanceof DistributedFileSystem)) {
            return User.getCurrent();
        }
        // Else distributed filesystem.  Make a new instance per daemon.  Below
        // code is taken from the AppendTestUtil over in hdfs.
        String username = User.getCurrent().getName() + differentiatingSuffix;
        User user = User.createUserForTesting(c, username, new String[] { "supergroup" });
        return user;
    }

    /**
     * Set maxRecoveryErrorCount in DFSClient.  In 0.20 pre-append its hard-coded to 5 and
     * makes tests linger.  Here is the exception you'll see:
     * <pre>
     * 2010-06-15 11:52:28,511 WARN  [DataStreamer for file /hbase/.logs/hlog.1276627923013 block blk_928005470262850423_1021] hdfs.DFSClient$DFSOutputStream(2657): Error Recovery for block blk_928005470262850423_1021 failed  because recovery from primary datanode 127.0.0.1:53683 failed 4 times.  Pipeline was 127.0.0.1:53687, 127.0.0.1:53683. Will retry...
     * </pre>
     *
     * @param stream A DFSClient.DFSOutputStream.
     */
    public static void setMaxRecoveryErrorCount(final OutputStream stream, final int max) {
        try {
            Class<?>[] clazzes = DFSClient.class.getDeclaredClasses();
            for (Class<?> clazz : clazzes) {
                String className = clazz.getSimpleName();
                if (className.equals("DFSOutputStream")) {
                    if (clazz.isInstance(stream)) {
                        Field maxRecoveryErrorCountField = stream.getClass()
                                .getDeclaredField("maxRecoveryErrorCount");
                        maxRecoveryErrorCountField.setAccessible(true);
                        maxRecoveryErrorCountField.setInt(stream, max);
                        break;
                    }
                }
            }
        } catch (Exception e) {
            LOG.info("Could not set max recovery field", e);
        }
    }

    /**
     * Wait until <code>countOfRegion</code> in .META. have a non-empty
     * info:server.  This means all regions have been deployed, master has been
     * informed and updated .META. with the regions deployed server.
     *
     * @param countOfRegions How many regions in .META.
     */
    public void waitUntilAllRegionsAssigned(final int countOfRegions) throws IOException {
        HTable meta = new HTable(getConfiguration(), HConstants.META_TABLE_NAME);
        while (true) {
            int rows = 0;
            Scan scan = new Scan();
            scan.addColumn(HConstants.CATALOG_FAMILY, HConstants.SERVER_QUALIFIER);
            ResultScanner s = meta.getScanner(scan);
            for (Result r = null; (r = s.next()) != null;) {
                byte[] b = r.getValue(HConstants.CATALOG_FAMILY, HConstants.SERVER_QUALIFIER);
                if (b == null || b.length <= 0) {
                    break;
                }
                rows++;
            }
            s.close();
            // If I get to here and all rows have a Server, then all have been assigned.
            if (rows == countOfRegions) {
                break;
            }
            LOG.info("Found=" + rows);
            Threads.sleep(200);
        }
    }

    /**
     * Do a small get/scan against one store. This is required because store
     * has no actual methods of querying itself, and relies on StoreScanner.
     */
    public static List<KeyValue> getFromStoreFile(Store store, Get get) throws IOException {
        MultiVersionConsistencyControl.resetThreadReadPoint();
        Scan scan = new Scan(get);
        InternalScanner scanner = (InternalScanner) store.getScanner(scan,
                scan.getFamilyMap().get(store.getFamily().getName()));

        List<KeyValue> result = new ArrayList<KeyValue>();
        scanner.next(result);
        if (!result.isEmpty()) {
            // verify that we are on the row we want:
            KeyValue kv = result.get(0);
            if (!Bytes.equals(kv.getRow(), get.getRow())) {
                result.clear();
            }
        }
        scanner.close();
        return result;
    }

    /**
     * Do a small get/scan against one store. This is required because store
     * has no actual methods of querying itself, and relies on StoreScanner.
     */
    public static List<KeyValue> getFromStoreFile(Store store, byte[] row, NavigableSet<byte[]> columns)
            throws IOException {
        Get get = new Get(row);
        Map<byte[], NavigableSet<byte[]>> s = get.getFamilyMap();
        s.put(store.getFamily().getName(), columns);

        return getFromStoreFile(store, get);
    }

    /**
     * Gets a ZooKeeperWatcher.
     */
    public static ZooKeeperWatcher getZooKeeperWatcher(HBaseTestingUtility TEST_UTIL)
            throws ZooKeeperConnectionException, IOException {
        ZooKeeperWatcher zkw = new ZooKeeperWatcher(TEST_UTIL.getConfiguration(), "unittest", new Abortable() {
            boolean aborted = false;

            @Override
            public void abort(String why, Throwable e) {
                aborted = true;
                throw new RuntimeException("Fatal ZK error, why=" + why, e);
            }

            @Override
            public boolean isAborted() {
                return aborted;
            }
        });
        return zkw;
    }

    /**
     * Creates a znode with OPENED state.
     */
    public static ZooKeeperWatcher createAndForceNodeToOpenedState(HBaseTestingUtility TEST_UTIL, HRegion region,
            ServerName serverName)
            throws ZooKeeperConnectionException, IOException, KeeperException, NodeExistsException {
        ZooKeeperWatcher zkw = getZooKeeperWatcher(TEST_UTIL);
        ZKAssign.createNodeOffline(zkw, region.getRegionInfo(), serverName);
        int version = ZKAssign.transitionNodeOpening(zkw, region.getRegionInfo(), serverName);
        ZKAssign.transitionNodeOpened(zkw, region.getRegionInfo(), serverName, version);
        return zkw;
    }

    public static void assertKVListsEqual(String additionalMsg, final List<KeyValue> expected,
            final List<KeyValue> actual) {
        final int eLen = expected.size();
        final int aLen = actual.size();
        final int minLen = Math.min(eLen, aLen);

        int i;
        for (i = 0; i < minLen && KeyValue.COMPARATOR.compare(expected.get(i), actual.get(i)) == 0; ++i) {
        }

        if (additionalMsg == null) {
            additionalMsg = "";
        }
        if (!additionalMsg.isEmpty()) {
            additionalMsg = ". " + additionalMsg;
        }

        if (eLen != aLen || i != minLen) {
            throw new AssertionError("Expected and actual KV arrays differ at position " + i + ": "
                    + safeGetAsStr(expected, i) + " (length " + eLen + ") vs. " + safeGetAsStr(actual, i)
                    + " (length " + aLen + ")" + additionalMsg);
        }
    }

    private static <T> String safeGetAsStr(List<T> lst, int i) {
        if (0 <= i && i < lst.size()) {
            return lst.get(i).toString();
        } else {
            return "<out_of_range>";
        }
    }

    public String getClusterKey() {
        return conf.get(HConstants.ZOOKEEPER_QUORUM) + ":" + conf.get(HConstants.ZOOKEEPER_CLIENT_PORT) + ":"
                + conf.get(HConstants.ZOOKEEPER_ZNODE_PARENT, HConstants.DEFAULT_ZOOKEEPER_ZNODE_PARENT);
    }

    /**
     * Creates a random table with the given parameters
     */
    public HTable createRandomTable(String tableName, final Collection<String> families, final int maxVersions,
            final int numColsPerRow, final int numFlushes, final int numRegions, final int numRowsPerFlush)
            throws IOException, InterruptedException {

        LOG.info("\n\nCreating random table " + tableName + " with " + numRegions + " regions, " + numFlushes
                + " storefiles per region, " + numRowsPerFlush + " rows per flush, maxVersions=" + maxVersions
                + "\n");

        final Random rand = new Random(tableName.hashCode() * 17L + 12938197137L);
        final int numCF = families.size();
        final byte[][] cfBytes = new byte[numCF][];
        final byte[] tableNameBytes = Bytes.toBytes(tableName);

        {
            int cfIndex = 0;
            for (String cf : families) {
                cfBytes[cfIndex++] = Bytes.toBytes(cf);
            }
        }

        final int actualStartKey = 0;
        final int actualEndKey = Integer.MAX_VALUE;
        final int keysPerRegion = (actualEndKey - actualStartKey) / numRegions;
        final int splitStartKey = actualStartKey + keysPerRegion;
        final int splitEndKey = actualEndKey - keysPerRegion;
        final String keyFormat = "%08x";
        final HTable table = createTable(tableNameBytes, cfBytes, maxVersions,
                Bytes.toBytes(String.format(keyFormat, splitStartKey)),
                Bytes.toBytes(String.format(keyFormat, splitEndKey)), numRegions);
        if (hbaseCluster != null) {
            getMiniHBaseCluster().flushcache(HConstants.META_TABLE_NAME);
        }

        for (int iFlush = 0; iFlush < numFlushes; ++iFlush) {
            for (int iRow = 0; iRow < numRowsPerFlush; ++iRow) {
                final byte[] row = Bytes.toBytes(
                        String.format(keyFormat, actualStartKey + rand.nextInt(actualEndKey - actualStartKey)));

                Put put = new Put(row);
                Delete del = new Delete(row);
                for (int iCol = 0; iCol < numColsPerRow; ++iCol) {
                    final byte[] cf = cfBytes[rand.nextInt(numCF)];
                    final long ts = rand.nextInt();
                    final byte[] qual = Bytes.toBytes("col" + iCol);
                    if (rand.nextBoolean()) {
                        final byte[] value = Bytes
                                .toBytes("value_for_row_" + iRow + "_cf_" + Bytes.toStringBinary(cf) + "_col_"
                                        + iCol + "_ts_" + ts + "_random_" + rand.nextLong());
                        put.add(cf, qual, ts, value);
                    } else if (rand.nextDouble() < 0.8) {
                        del.deleteColumn(cf, qual, ts);
                    } else {
                        del.deleteColumns(cf, qual, ts);
                    }
                }

                if (!put.isEmpty()) {
                    table.put(put);
                }

                if (!del.isEmpty()) {
                    table.delete(del);
                }
            }
            LOG.info("Initiating flush #" + iFlush + " for table " + tableName);
            table.flushCommits();
            if (hbaseCluster != null) {
                getMiniHBaseCluster().flushcache(tableNameBytes);
            }
        }

        return table;
    }

    private static final int MIN_RANDOM_PORT = 0xc000;
    private static final int MAX_RANDOM_PORT = 0xfffe;

    /**
     * Returns a random port. These ports cannot be registered with IANA and are
     * intended for dynamic allocation (see http://bit.ly/dynports).
     */
    public static int randomPort() {
        return MIN_RANDOM_PORT + new Random().nextInt(MAX_RANDOM_PORT - MIN_RANDOM_PORT);
    }

    public static int randomFreePort() {
        int port = 0;
        do {
            port = randomPort();
            try {
                ServerSocket sock = new ServerSocket(port);
                sock.close();
            } catch (IOException ex) {
                port = 0;
            }
        } while (port == 0);
        return port;
    }

    public static void waitForHostPort(String host, int port) throws IOException {
        final int maxTimeMs = 10000;
        final int maxNumAttempts = maxTimeMs / HConstants.SOCKET_RETRY_WAIT_MS;
        IOException savedException = null;
        LOG.info("Waiting for server at " + host + ":" + port);
        for (int attempt = 0; attempt < maxNumAttempts; ++attempt) {
            try {
                Socket sock = new Socket(InetAddress.getByName(host), port);
                sock.close();
                savedException = null;
                LOG.info("Server at " + host + ":" + port + " is available");
                break;
            } catch (UnknownHostException e) {
                throw new IOException("Failed to look up " + host, e);
            } catch (IOException e) {
                savedException = e;
            }
            Threads.sleepWithoutInterrupt(HConstants.SOCKET_RETRY_WAIT_MS);
        }

        if (savedException != null) {
            throw savedException;
        }
    }

    /**
     * Creates a pre-split table for load testing. If the table already exists,
     * logs a warning and continues.
     *
     * @return the number of regions the table was split into
     */
    public static int createPreSplitLoadTestTable(Configuration conf, byte[] tableName, byte[] columnFamily,
            Algorithm compression, DataBlockEncoding dataBlockEncoding) throws IOException {
        HTableDescriptor desc = new HTableDescriptor(tableName);
        HColumnDescriptor hcd = new HColumnDescriptor(columnFamily);
        hcd.setDataBlockEncoding(dataBlockEncoding);
        hcd.setCompressionType(compression);
        desc.addFamily(hcd);

        int totalNumberOfRegions = 0;
        try {
            HBaseAdmin admin = new HBaseAdmin(conf);

            // create a table a pre-splits regions.
            // The number of splits is set as:
            //    region servers * regions per region server).
            int numberOfServers = admin.getClusterStatus().getServers().size();
            if (numberOfServers == 0) {
                throw new IllegalStateException("No live regionservers");
            }

            totalNumberOfRegions = numberOfServers * DEFAULT_REGIONS_PER_SERVER;
            LOG.info("Number of live regionservers: " + numberOfServers + ", " + "pre-splitting table into "
                    + totalNumberOfRegions + " regions " + "(default regions per server: "
                    + DEFAULT_REGIONS_PER_SERVER + ")");

            byte[][] splits = new RegionSplitter.HexStringSplit().split(totalNumberOfRegions);

            admin.createTable(desc, splits);
            admin.close();
        } catch (MasterNotRunningException e) {
            LOG.error("Master not running", e);
            throw new IOException(e);
        } catch (TableExistsException e) {
            LOG.warn("Table " + Bytes.toStringBinary(tableName) + " already exists, continuing");
        }
        return totalNumberOfRegions;
    }

    public static int getMetaRSPort(Configuration conf) throws IOException {
        HTable table = new HTable(conf, HConstants.META_TABLE_NAME);
        HRegionLocation hloc = table.getRegionLocation(Bytes.toBytes(""));
        table.close();
        return hloc.getPort();
    }

    public HRegion createTestRegion(String tableName, HColumnDescriptor hcd) throws IOException {
        HTableDescriptor htd = new HTableDescriptor(tableName);
        htd.addFamily(hcd);
        HRegionInfo info = new HRegionInfo(Bytes.toBytes(tableName), null, null, false);
        HRegion region = HRegion.createHRegion(info, getDataTestDir(), getConfiguration(), htd);
        return region;
    }

    // Lily change: add getter for MRCluster
    public MiniMRCluster getMRCluster() {
        return mrCluster;
    }

    /**
     * Create a set of column descriptors with the combination of compression,
     * encoding, bloom codecs available.
     *
     * @param prefix family names prefix
     * @return the list of column descriptors
     */
    public static List<HColumnDescriptor> generateColumnDescriptors(final String prefix) {
        List<HColumnDescriptor> htds = new ArrayList<HColumnDescriptor>();
        long familyId = 0;
        for (Compression.Algorithm compressionType : getSupportedCompressionAlgorithms()) {
            for (DataBlockEncoding encodingType : DataBlockEncoding.values()) {
                for (StoreFile.BloomType bloomType : StoreFile.BloomType.values()) {
                    String name = String.format("%s-cf-!@#&-%d!@#", prefix, familyId);
                    HColumnDescriptor htd = new HColumnDescriptor(name);
                    htd.setCompressionType(compressionType);
                    htd.setDataBlockEncoding(encodingType);
                    htd.setBloomFilterType(bloomType);
                    htds.add(htd);
                    familyId++;
                }
            }
        }
        return htds;
    }

    /**
     * Get supported compression algorithms.
     *
     * @return supported compression algorithms.
     */
    public static Compression.Algorithm[] getSupportedCompressionAlgorithms() {
        String[] allAlgos = HFile.getSupportedCompressionAlgorithms();
        List<Compression.Algorithm> supportedAlgos = new ArrayList<Compression.Algorithm>();
        for (String algoName : allAlgos) {
            try {
                Compression.Algorithm algo = Compression.getCompressionAlgorithmByName(algoName);
                algo.getCompressor();
                supportedAlgos.add(algo);
            } catch (Throwable t) {
                // this algo is not available
            }
        }
        return supportedAlgos.toArray(new Compression.Algorithm[0]);
    }
}