<?xml version="1.0" encoding="UTF-8"?>
<!--
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.
-->
<!DOCTYPE concept PUBLIC "-//OASIS//DTD DITA Concept//EN" "concept.dtd">
<concept id="impala_hbase">

  <title id="hbase">Using Impala to Query HBase Tables</title>
  <titlealts audience="PDF"><navtitle>HBase Tables</navtitle></titlealts>
  <prolog>
    <metadata>
      <data name="Category" value="Impala"/>
      <data name="Category" value="HBase"/>
      <data name="Category" value="Querying"/>
      <data name="Category" value="Data Analysts"/>
      <data name="Category" value="Developers"/>
      <data name="Category" value="Tables"/>
    </metadata>
  </prolog>

  <conbody>

    <p>
      <indexterm audience="hidden">HBase</indexterm>
      You can use Impala to query HBase tables. This capability allows convenient access to a storage system that
      is tuned for different kinds of workloads than the default with Impala. The default Impala tables use data
      files stored on HDFS, which are ideal for bulk loads and queries using full-table scans. In contrast, HBase
      can do efficient queries for data organized for OLTP-style workloads, with lookups of individual rows or
      ranges of values.
    </p>

    <p>
      From the perspective of an Impala user, coming from an RDBMS background, HBase is a kind of key-value store
      where the value consists of multiple fields. The key is mapped to one column in the Impala table, and the
      various fields of the value are mapped to the other columns in the Impala table.
    </p>

    <p>
      For background information on HBase, see <xref keyref="upstream_hbase_docs"/>.
    </p>

    <p outputclass="toc inpage"/>
  </conbody>

  <concept id="hbase_using">

    <title>Overview of Using HBase with Impala</title>
  <prolog>
    <metadata>
      <data name="Category" value="Concepts"/>
    </metadata>
  </prolog>

    <conbody>

      <p>
        When you use Impala with HBase:
      </p>

      <ul>
        <li>
          You create the tables on the Impala side using the Hive shell, because the Impala <codeph>CREATE
          TABLE</codeph> statement currently does not support custom SerDes and some other syntax needed for these
          tables:
          <ul>
            <li>
              You designate it as an HBase table using the <codeph>STORED BY
              'org.apache.hadoop.hive.hbase.HBaseStorageHandler'</codeph> clause on the Hive <codeph>CREATE
              TABLE</codeph> statement.
            </li>

            <li>
              You map these specially created tables to corresponding tables that exist in HBase, with the clause
              <codeph>TBLPROPERTIES("hbase.table.name" = "<varname>table_name_in_hbase</varname>")</codeph> on the
              Hive <codeph>CREATE TABLE</codeph> statement.
            </li>

            <li>
              See <xref href="#hbase_queries"/> for a full example.
            </li>
          </ul>
        </li>

        <li>
          You define the column corresponding to the HBase row key as a string with the <codeph>#string</codeph>
          keyword, or map it to a <codeph>STRING</codeph> column.
        </li>

        <li>
          Because Impala and Hive share the same metastore database, once you create the table in Hive, you can
          query or insert into it through Impala. (After creating a new table through Hive, issue the
          <codeph>INVALIDATE METADATA</codeph> statement in <cmdname>impala-shell</cmdname> to make Impala aware of
          the new table.)
        </li>

        <li>
          You issue queries against the Impala tables. For efficient queries, use <codeph>WHERE</codeph> clauses to
          find a single key value or a range of key values wherever practical, by testing the Impala column
          corresponding to the HBase row key. Avoid queries that do full-table scans, which are efficient for
          regular Impala tables but inefficient in HBase.
        </li>
      </ul>

      <p>
        To work with an HBase table from Impala, ensure that the <codeph>impala</codeph> user has read/write
        privileges for the HBase table, using the <codeph>GRANT</codeph> command in the HBase shell. For details
        about HBase security, see <xref keyref="upstream_hbase_security_docs"/>.
      </p>
    </conbody>
  </concept>

  <concept id="hbase_config">

    <title>Configuring HBase for Use with Impala</title>
  <prolog>
    <metadata>
      <data name="Category" value="Configuring"/>
    </metadata>
  </prolog>

    <conbody>

      <p>
        HBase works out of the box with Impala. There is no mandatory configuration needed to use these two
        components together.
      </p>

      <p>
        To avoid delays if HBase is unavailable during Impala startup or after an <codeph>INVALIDATE
        METADATA</codeph> statement, set timeout values similar to the following in
        <filepath>/etc/impala/conf/hbase-site.xml</filepath>:
      </p>

<codeblock>&lt;property&gt;
  &lt;name&gt;hbase.client.retries.number&lt;/name&gt;
  &lt;value&gt;3&lt;/value&gt;
&lt;/property&gt;
&lt;property&gt;
  &lt;name&gt;hbase.rpc.timeout&lt;/name&gt;
  &lt;value&gt;3000&lt;/value&gt;
&lt;/property&gt;
</codeblock>

    </conbody>
  </concept>

  <concept id="hbase_types">

    <title>Supported Data Types for HBase Columns</title>

    <conbody>

      <p>
        To understand how Impala column data types are mapped to fields in HBase, you should have some background
        knowledge about HBase first. You set up the mapping by running the <codeph>CREATE TABLE</codeph> statement
        in the Hive shell. See
        <xref href="https://cwiki.apache.org/confluence/display/Hive/HBaseIntegration" scope="external" format="html">the
        Hive wiki</xref> for a starting point, and <xref href="#hbase_queries"/> for examples.
      </p>

      <p>
        HBase works as a kind of <q>bit bucket</q>, in the sense that HBase does not enforce any typing for the
        key or value fields. All the type enforcement is done on the Impala side.
      </p>

      <p>
        For best performance of Impala queries against HBase tables, most queries will perform comparisons in the
        <codeph>WHERE</codeph> against the column that corresponds to the HBase row key. When creating the table
        through the Hive shell, use the <codeph>STRING</codeph> data type for the column that corresponds to the
        HBase row key. Impala can translate conditional tests (through operators such as <codeph>=</codeph>,
        <codeph>&lt;</codeph>, <codeph>BETWEEN</codeph>, and <codeph>IN</codeph>) against this column into fast
        lookups in HBase, but this optimization (<q>predicate pushdown</q>) only works when that column is
        defined as <codeph>STRING</codeph>.
      </p>

      <p>
        Starting in Impala 1.1, Impala also supports reading and writing to columns that are defined in the Hive
        <codeph>CREATE TABLE</codeph> statement using binary data types, represented in the Hive table definition
        using the <codeph>#binary</codeph> keyword, often abbreviated as <codeph>#b</codeph>. Defining numeric
        columns as binary can reduce the overall data volume in the HBase tables. You should still define the
        column that corresponds to the HBase row key as a <codeph>STRING</codeph>, to allow fast lookups using
        those columns.
      </p>
    </conbody>
  </concept>

  <concept id="hbase_performance">

    <title>Performance Considerations for the Impala-HBase Integration</title>
  <prolog>
    <metadata>
      <data name="Category" value="Performance"/>
    </metadata>
  </prolog>

    <conbody>

      <p>
        To understand the performance characteristics of SQL queries against data stored in HBase, you should have
        some background knowledge about how HBase interacts with SQL-oriented systems first. See
        <xref href="https://cwiki.apache.org/confluence/display/Hive/HBaseIntegration" scope="external" format="html">the
        Hive wiki</xref> for a starting point; because Impala shares the same metastore database as Hive, the
        information about mapping columns from Hive tables to HBase tables is generally applicable to Impala too.
      </p>

      <p>
        Impala uses the HBase client API via Java Native Interface (JNI) to query data stored in HBase. This
        querying does not read HFiles directly. The extra communication overhead makes it important to choose what
        data to store in HBase or in HDFS, and construct efficient queries that can retrieve the HBase data
        efficiently:
      </p>

      <ul>
        <li>
          Use HBase table for queries that return a single row or a range of rows, not queries that scan the entire
          table. (If a query has no <codeph>WHERE</codeph> clause, that is a strong indicator that it is an
          inefficient query for an HBase table.)
        </li>

        <li>
          If you have join queries that do aggregation operations on large fact tables and join the results against
          small dimension tables, consider using Impala for the fact tables and HBase for the dimension tables.
          (Because Impala does a full scan on the HBase table in this case, rather than doing single-row HBase
          lookups based on the join column, only use this technique where the HBase table is small enough that
          doing a full table scan does not cause a performance bottleneck for the query.)
        </li>
      </ul>

      <p>
        Query predicates are applied to row keys as start and stop keys, thereby limiting the scope of a particular
        lookup. If row keys are not mapped to string columns, then ordering is typically incorrect and comparison
        operations do not work. For example, if row keys are not mapped to string columns, evaluating for greater
        than (&gt;) or less than (&lt;) cannot be completed.
      </p>

      <p>
        Predicates on non-key columns can be sent to HBase to scan as <codeph>SingleColumnValueFilters</codeph>,
        providing some performance gains. In such a case, HBase returns fewer rows than if those same predicates
        were applied using Impala. While there is some improvement, it is not as great when start and stop rows are
        used. This is because the number of rows that HBase must examine is not limited as it is when start and
        stop rows are used. As long as the row key predicate only applies to a single row, HBase will locate and
        return that row. Conversely, if a non-key predicate is used, even if it only applies to a single row, HBase
        must still scan the entire table to find the correct result.
      </p>

      <example>

        <title>Interpreting EXPLAIN Output for HBase Queries</title>

        <p>
          For example, here are some queries against the following Impala table, which is mapped to an HBase table.
          The examples show excerpts from the output of the <codeph>EXPLAIN</codeph> statement, demonstrating what
          things to look for to indicate an efficient or inefficient query against an HBase table.
        </p>

        <p>
          The first column (<codeph>cust_id</codeph>) was specified as the key column in the <codeph>CREATE
          EXTERNAL TABLE</codeph> statement; for performance, it is important to declare this column as
          <codeph>STRING</codeph>. Other columns, such as <codeph>BIRTH_YEAR</codeph> and
          <codeph>NEVER_LOGGED_ON</codeph>, are also declared as <codeph>STRING</codeph>, rather than their
          <q>natural</q> types of <codeph>INT</codeph> or <codeph>BOOLEAN</codeph>, because Impala can optimize
          those types more effectively in HBase tables. For comparison, we leave one column,
          <codeph>YEAR_REGISTERED</codeph>, as <codeph>INT</codeph> to show that filtering on this column is
          inefficient.
        </p>

<codeblock>describe hbase_table;
Query: describe hbase_table
+-----------------------+--------+---------+
| name                  | type   | comment |
+-----------------------+--------+---------+
| cust_id               | <b>string</b> |         |
| birth_year            | <b>string</b> |         |
| never_logged_on       | <b>string</b> |         |
| private_email_address | string |         |
| year_registered       | <b>int</b>    |         |
+-----------------------+--------+---------+
</codeblock>

        <p>
          The best case for performance involves a single row lookup using an equality comparison on the column
          defined as the row key:
        </p>

<codeblock>explain select count(*) from hbase_table where cust_id = 'some_user@example.com';
+------------------------------------------------------------------------------------+
| Explain String                                                                     |
+------------------------------------------------------------------------------------+
| Estimated Per-Host Requirements: Memory=1.01GB VCores=1                            |
| WARNING: The following tables are missing relevant table and/or column statistics. |
| hbase.hbase_table                                                                  |
|                                                                                    |
| 03:AGGREGATE [MERGE FINALIZE]                                                      |
| |  output: sum(count(*))                                                           |
| |                                                                                  |
| 02:EXCHANGE [PARTITION=UNPARTITIONED]                                              |
| |                                                                                  |
| 01:AGGREGATE                                                                       |
| |  output: count(*)                                                                |
| |                                                                                  |
<b>| 00:SCAN HBASE [hbase.hbase_table]                                                  |</b>
<b>|    start key: some_user@example.com                                                |</b>
<b>|    stop key: some_user@example.com\0                                               |</b>
+------------------------------------------------------------------------------------+
</codeblock>

        <p>
          Another type of efficient query involves a range lookup on the row key column, using SQL operators such
          as greater than (or equal), less than (or equal), or <codeph>BETWEEN</codeph>. This example also includes
          an equality test on a non-key column; because that column is a <codeph>STRING</codeph>, Impala can let
          HBase perform that test, indicated by the <codeph>hbase filters:</codeph> line in the
          <codeph>EXPLAIN</codeph> output. Doing the filtering within HBase is more efficient than transmitting all
          the data to Impala and doing the filtering on the Impala side.
        </p>

<codeblock>explain select count(*) from hbase_table where cust_id between 'a' and 'b'
  and never_logged_on = 'true';
+------------------------------------------------------------------------------------+
| Explain String                                                                     |
+------------------------------------------------------------------------------------+
...
<!--| Estimated Per-Host Requirements: Memory=1.01GB VCores=1                            |
| WARNING: The following tables are missing relevant table and/or column statistics. |
| hbase.hbase_table                                                                  |
|                                                                                    |
| 03:AGGREGATE [MERGE FINALIZE]                                                      |
| |  output: sum(count(*))                                                           |
| |                                                                                  |
| 02:EXCHANGE [PARTITION=UNPARTITIONED]                                              |
| |                                                                                  |-->
| 01:AGGREGATE                                                                       |
| |  output: count(*)                                                                |
| |                                                                                  |
<b>| 00:SCAN HBASE [hbase.hbase_table]                                                  |</b>
<b>|    start key: a                                                                    |</b>
<b>|    stop key: b\0                                                                   |</b>
<b>|    hbase filters: cols:never_logged_on EQUAL 'true'                                |</b>
+------------------------------------------------------------------------------------+
</codeblock>

        <p>
          The query is less efficient if Impala has to evaluate any of the predicates, because Impala must scan the
          entire HBase table. Impala can only push down predicates to HBase for columns declared as
          <codeph>STRING</codeph>. This example tests a column declared as <codeph>INT</codeph>, and the
          <codeph>predicates:</codeph> line in the <codeph>EXPLAIN</codeph> output indicates that the test is
          performed after the data is transmitted to Impala.
        </p>

<codeblock>explain select count(*) from hbase_table where year_registered = 2010;
+------------------------------------------------------------------------------------+
| Explain String                                                                     |
+------------------------------------------------------------------------------------+
...
<!--| Estimated Per-Host Requirements: Memory=1.01GB VCores=1                            |
| WARNING: The following tables are missing relevant table and/or column statistics. |
| hbase.hbase_table                                                                  |
|                                                                                    |
| 03:AGGREGATE [MERGE FINALIZE]                                                      |
| |  output: sum(count(*))                                                           |
| |                                                                                  |
| 02:EXCHANGE [PARTITION=UNPARTITIONED]                                              |
| |                                                                                  |-->
| 01:AGGREGATE                                                                       |
| |  output: count(*)                                                                |
| |                                                                                  |
<b>| 00:SCAN HBASE [hbase.hbase_table]                                                  |</b>
<b>|    predicates: year_registered = 2010                                              |</b>
+------------------------------------------------------------------------------------+
</codeblock>

        <p>
          The same inefficiency applies if the key column is compared to any non-constant value. Here, even though
          the key column is a <codeph>STRING</codeph>, and is tested using an equality operator, Impala must scan
          the entire HBase table because the key column is compared to another column value rather than a constant.
        </p>

<codeblock>explain select count(*) from hbase_table where cust_id = private_email_address;
+------------------------------------------------------------------------------------+
| Explain String                                                                     |
+------------------------------------------------------------------------------------+
...
<!--| Estimated Per-Host Requirements: Memory=1.01GB VCores=1                            |
| WARNING: The following tables are missing relevant table and/or column statistics. |
| hbase.hbase_table                                                                  |
|                                                                                    |
| 03:AGGREGATE [MERGE FINALIZE]                                                      |
| |  output: sum(count(*))                                                           |
| |                                                                                  |
| 02:EXCHANGE [PARTITION=UNPARTITIONED]                                              |
| |                                                                                  |-->
| 01:AGGREGATE                                                                       |
| |  output: count(*)                                                                |
| |                                                                                  |
<b>| 00:SCAN HBASE [hbase.hbase_table]                                                  |</b>
<b>|    predicates: cust_id = private_email_address                                    |</b>
+------------------------------------------------------------------------------------+
</codeblock>

        <p>
          Currently, tests on the row key using <codeph>OR</codeph> or <codeph>IN</codeph> clauses are not
          optimized into direct lookups either. Such limitations might be lifted in the future, so always check the
          <codeph>EXPLAIN</codeph> output to be sure whether a particular SQL construct results in an efficient
          query or not for HBase tables.
        </p>

<codeblock>explain select count(*) from hbase_table where
  cust_id = 'some_user@example.com' or cust_id = 'other_user@example.com';
+----------------------------------------------------------------------------------------+
| Explain String                                                                         |
+----------------------------------------------------------------------------------------+
...
<!--| Estimated Per-Host Requirements: Memory=1.01GB VCores=1                                |
| WARNING: The following tables are missing relevant table and/or column statistics.     |
| hbase.hbase_table                                                                      |
|                                                                                        |
| 03:AGGREGATE [MERGE FINALIZE]                                                          |
| |  output: sum(count(*))                                                               |
| |                                                                                      |
| 02:EXCHANGE [PARTITION=UNPARTITIONED]                                                  |
| |                                                                                      |-->
| 01:AGGREGATE                                                                           |
| |  output: count(*)                                                                    |
| |                                                                                      |
<b>| 00:SCAN HBASE [hbase.hbase_table]                                                      |</b>
<b>|    predicates: cust_id = 'some_user@example.com' OR cust_id = 'other_user@example.com' |</b>
+----------------------------------------------------------------------------------------+

explain select count(*) from hbase_table where
  cust_id in ('some_user@example.com', 'other_user@example.com');
+------------------------------------------------------------------------------------+
| Explain String                                                                     |
+------------------------------------------------------------------------------------+
...
<!--| Estimated Per-Host Requirements: Memory=1.01GB VCores=1                            |
| WARNING: The following tables are missing relevant table and/or column statistics. |
| hbase.hbase_table                                                                  |
|                                                                                    |
| 03:AGGREGATE [MERGE FINALIZE]                                                      |
| |  output: sum(count(*))                                                           |
| |                                                                                  |
| 02:EXCHANGE [PARTITION=UNPARTITIONED]                                              |
| |                                                                                  |-->
| 01:AGGREGATE                                                                       |
| |  output: count(*)                                                                |
| |                                                                                  |
<b>| 00:SCAN HBASE [hbase.hbase_table]                                                  |</b>
<b>|    predicates: cust_id IN ('some_user@example.com', 'other_user@example.com')      |</b>
+------------------------------------------------------------------------------------+
</codeblock>

        <p>
          Either rewrite into separate queries for each value and combine the results in the application, or
          combine the single-row queries using UNION ALL:
        </p>

<codeblock>select count(*) from hbase_table where cust_id = 'some_user@example.com';
select count(*) from hbase_table where cust_id = 'other_user@example.com';

explain
  select count(*) from hbase_table where cust_id = 'some_user@example.com'
  union all
  select count(*) from hbase_table where cust_id = 'other_user@example.com';
+------------------------------------------------------------------------------------+
| Explain String                                                                     |
+------------------------------------------------------------------------------------+
...
<!--| Estimated Per-Host Requirements: Memory=1.01GB VCores=1                            |
| WARNING: The following tables are missing relevant table and/or column statistics. |
| hbase.hbase_table                                                                  |
|                                                                                    |
| 09:EXCHANGE [PARTITION=UNPARTITIONED]                                              |
| |                                                                                  |
| |&minus;&minus;11:MERGE                                                                        |
| |  |                                                                               |
| |  08:AGGREGATE [MERGE FINALIZE]                                                   |
| |  |  output: sum(count(*))                                                        |
| |  |                                                                               |
| |  07:EXCHANGE [PARTITION=UNPARTITIONED]                                           |
| |  |                                                                               |-->
| |  04:AGGREGATE                                                                    |
| |  |  output: count(*)                                                             |
| |  |                                                                               |
<b>| |  03:SCAN HBASE [hbase.hbase_table]                                               |</b>
<b>| |     start key: other_user@example.com                                            |</b>
<b>| |     stop key: other_user@example.com\0                                           |</b>
| |                                                                                  |
| 10:MERGE                                                                           |
...
<!--| |                                                                                  |
| 06:AGGREGATE [MERGE FINALIZE]                                                      |
| |  output: sum(count(*))                                                           |
| |                                                                                  |
| 05:EXCHANGE [PARTITION=UNPARTITIONED]                                              |
| |                                                                                  |-->
| 02:AGGREGATE                                                                       |
| |  output: count(*)                                                                |
| |                                                                                  |
<b>| 01:SCAN HBASE [hbase.hbase_table]                                                  |</b>
<b>|    start key: some_user@example.com                                                |</b>
<b>|    stop key: some_user@example.com\0                                               |</b>
+------------------------------------------------------------------------------------+
</codeblock>

      </example>

      <example>

        <title>Configuration Options for Java HBase Applications</title>

        <p> If you have an HBase Java application that calls the
            <codeph>setCacheBlocks</codeph> or <codeph>setCaching</codeph>
          methods of the class <xref
            href="http://hbase.apache.org/apidocs/org/apache/hadoop/hbase/client/Scan.html"
            scope="external" format="html"
            >org.apache.hadoop.hbase.client.Scan</xref>, you can set these same
          caching behaviors through Impala query options, to control the memory
          pressure on the HBase RegionServer. For example, when doing queries in
          HBase that result in full-table scans (which by default are
          inefficient for HBase), you can reduce memory usage and speed up the
          queries by turning off the <codeph>HBASE_CACHE_BLOCKS</codeph> setting
          and specifying a large number for the <codeph>HBASE_CACHING</codeph>
          setting.
        </p>

        <p>
          To set these options, issue commands like the following in <cmdname>impala-shell</cmdname>:
        </p>

<codeblock>-- Same as calling setCacheBlocks(true) or setCacheBlocks(false).
set hbase_cache_blocks=true;
set hbase_cache_blocks=false;

-- Same as calling setCaching(rows).
set hbase_caching=1000;
</codeblock>

        <p>
          Or update the <cmdname>impalad</cmdname> defaults file <filepath>/etc/default/impala</filepath> and
          include settings for <codeph>HBASE_CACHE_BLOCKS</codeph> and/or <codeph>HBASE_CACHING</codeph> in the
          <codeph>-default_query_options</codeph> setting for <codeph>IMPALA_SERVER_ARGS</codeph>. See
          <xref href="impala_config_options.xml#config_options"/> for details.
        </p>

        <note>
          In Impala 2.0 and later, these options are settable through the JDBC or ODBC interfaces using the
          <codeph>SET</codeph> statement.
        </note>

      </example>
    </conbody>
  </concept>

  <concept id="hbase_scenarios">

    <title>Use Cases for Querying HBase through Impala</title>
    <prolog>
      <metadata>
        <data name="Category" value="Use Cases"/>
      </metadata>
    </prolog>

    <conbody>

      <p>
        The following are popular use cases for using Impala to query HBase tables:
      </p>

      <ul>
        <li>
          Keeping large fact tables in Impala, and smaller dimension tables in HBase. The fact tables use Parquet
          or other binary file format optimized for scan operations. Join queries scan through the large Impala
          fact tables, and cross-reference the dimension tables using efficient single-row lookups in HBase.
        </li>

        <li>
          Using HBase to store rapidly incrementing counters, such as how many times a web page has been viewed, or
          on a social network, how many connections a user has or how many votes a post received. HBase is
          efficient for capturing such changeable data: the append-only storage mechanism is efficient for writing
          each change to disk, and a query always returns the latest value. An application could query specific
          totals like these from HBase, and combine the results with a broader set of data queried from Impala.
        </li>

        <li>
          <p>
            Storing very wide tables in HBase. Wide tables have many columns, possibly thousands, typically
            recording many attributes for an important subject such as a user of an online service. These tables
            are also often sparse, that is, most of the columns values are <codeph>NULL</codeph>, 0,
            <codeph>false</codeph>, empty string, or other blank or placeholder value. (For example, any particular
            web site user might have never used some site feature, filled in a certain field in their profile,
            visited a particular part of the site, and so on.) A typical query against this kind of table is to
            look up a single row to retrieve all the information about a specific subject, rather than summing,
            averaging, or filtering millions of rows as in typical Impala-managed tables.
          </p>
          <p>
            Or the HBase table could be joined with a larger Impala-managed table. For example, analyze the large
            Impala table representing web traffic for a site and pick out 50 users who view the most pages. Join
            that result with the wide user table in HBase to look up attributes of those users. The HBase side of
            the join would result in 50 efficient single-row lookups in HBase, rather than scanning the entire user
            table.
          </p>
        </li>
      </ul>
    </conbody>
  </concept>

  <concept audience="hidden" id="hbase_create_new">

    <title>Creating a New HBase Table for Impala to Use</title>

    <conbody>

      <p>
        You can create an HBase-backed table through a <codeph>CREATE TABLE</codeph> statement in the Hive shell,
        without going into the HBase shell at all:
      </p>

      <!-- To do:
        Add example. (Not critical because this subtopic is currently hidden.)
      -->
    </conbody>
  </concept>

  <concept audience="hidden" id="hbase_reuse_existing">

    <title>Associate Impala with an Existing HBase Table</title>

    <conbody>

      <p>
        If you already have some HBase tables created through the HBase shell, you can make them accessible to
        Impala through a <codeph>CREATE TABLE</codeph> statement in the Hive shell:
      </p>

      <!-- To do:
        Add example. (Not critical because this subtopic is currently hidden.)
      -->
    </conbody>
  </concept>

  <concept audience="hidden" id="hbase_column_families">

    <title>Map HBase Columns and Column Families to Impala Columns</title>

    <conbody>

      <p/>
    </conbody>
  </concept>

  <concept id="hbase_loading">

    <title>Loading Data into an HBase Table</title>
  <prolog>
    <metadata>
      <data name="Category" value="ETL"/>
      <data name="Category" value="Ingest"/>
    </metadata>
  </prolog>

    <conbody>

      <p>
        The Impala <codeph>INSERT</codeph> statement works for HBase tables. The <codeph>INSERT ... VALUES</codeph>
        syntax is ideally suited to HBase tables, because inserting a single row is an efficient operation for an
        HBase table. (For regular Impala tables, with data files in HDFS, the tiny data files produced by
        <codeph>INSERT ... VALUES</codeph> are extremely inefficient, so you would not use that technique with
        tables containing any significant data volume.)
      </p>

      <!-- To do:
        Add examples throughout this section.
      -->

      <p>
        When you use the <codeph>INSERT ... SELECT</codeph> syntax, the result in the HBase table could be fewer
        rows than you expect. HBase only stores the most recent version of each unique row key, so if an
        <codeph>INSERT ... SELECT</codeph> statement copies over multiple rows containing the same value for the
        key column, subsequent queries will only return one row with each key column value:
      </p>

      <p>
        Although Impala does not have an <codeph>UPDATE</codeph> statement, you can achieve the same effect by
        doing successive <codeph>INSERT</codeph> statements using the same value for the key column each time:
      </p>

    </conbody>
  </concept>

  <concept id="hbase_limitations">

    <title>Limitations and Restrictions of the Impala and HBase Integration</title>

    <conbody>

      <p>
        The Impala integration with HBase has the following limitations and restrictions, some inherited from the
        integration between HBase and Hive, and some unique to Impala:
      </p>

      <ul>
        <li>
          <p>
            If you issue a <codeph>DROP TABLE</codeph> for an internal (Impala-managed) table that is mapped to an
            HBase table, the underlying table is not removed in HBase. The Hive <codeph>DROP TABLE</codeph>
            statement also removes the HBase table in this case.
          </p>
        </li>

        <li>
          <p>
            The <codeph>INSERT OVERWRITE</codeph> statement is not available for HBase tables. You can insert new
            data, or modify an existing row by inserting a new row with the same key value, but not replace the
            entire contents of the table. You can do an <codeph>INSERT OVERWRITE</codeph> in Hive if you need this
            capability.
          </p>
        </li>

        <li>
          <p>
            If you issue a <codeph>CREATE TABLE LIKE</codeph> statement for a table mapped to an HBase table, the
            new table is also an HBase table, but inherits the same underlying HBase table name as the original.
            The new table is effectively an alias for the old one, not a new table with identical column structure.
            Avoid using <codeph>CREATE TABLE LIKE</codeph> for HBase tables, to avoid any confusion.
          </p>
        </li>

        <li>
          <p>
            Copying data into an HBase table using the Impala <codeph>INSERT ... SELECT</codeph> syntax might
            produce fewer new rows than are in the query result set. If the result set contains multiple rows with
            the same value for the key column, each row supercedes any previous rows with the same key value.
            Because the order of the inserted rows is unpredictable, you cannot rely on this technique to preserve
            the <q>latest</q> version of a particular key value.
          </p>
        </li>
        <li rev="2.3.0">
          <p>
            Because the complex data types (<codeph>ARRAY</codeph>, <codeph>STRUCT</codeph>, and <codeph>MAP</codeph>)
            available in <keyword keyref="impala23_full"/> and higher are currently only supported in Parquet tables, you cannot
            use these types in HBase tables that are queried through Impala.
          </p>
        </li>
        <li>
          <p conref="../shared/impala_common.xml#common/hbase_no_load_data"/>
        </li>
        <li>
          <p conref="../shared/impala_common.xml#common/tablesample_caveat"/>
        </li>
      </ul>
    </conbody>
  </concept>

  <concept id="hbase_queries">

    <title>Examples of Querying HBase Tables from Impala</title>

    <conbody>

      <p>
        The following examples create an HBase table with four column families,
        create a corresponding table through Hive,
        then insert and query the table through Impala.
      </p>
      <p>
        In HBase shell, the table
        name is quoted in <codeph>CREATE</codeph> and <codeph>DROP</codeph> statements. Tables created in HBase
        begin in <q>enabled</q> state; before dropping them through the HBase shell, you must issue a
        <codeph>disable '<varname>table_name</varname>'</codeph> statement.
      </p>

<codeblock>$ hbase shell
15/02/10 16:07:45
HBase Shell; enter 'help&lt;RETURN>' for list of supported commands.
Type "exit&lt;RETURN>" to leave the HBase Shell
...

hbase(main):001:0> create 'hbasealltypessmall', 'boolsCF', 'intsCF', 'floatsCF', 'stringsCF'
0 row(s) in 4.6520 seconds

=> Hbase::Table - hbasealltypessmall
hbase(main):006:0> quit
</codeblock>

        <p>
          Issue the following <codeph>CREATE TABLE</codeph> statement in the Hive shell. (The Impala <codeph>CREATE
          TABLE</codeph> statement currently does not support the <codeph>STORED BY</codeph> clause, so you switch into Hive to
          create the table, then back to Impala and the <cmdname>impala-shell</cmdname> interpreter to issue the
          queries.)
        </p>

        <p>
          This example creates an external table mapped to the HBase table, usable by both Impala and Hive. It is
          defined as an external table so that when dropped by Impala or Hive, the original HBase table is not touched at all.
        </p>

        <p>
          The <codeph>WITH SERDEPROPERTIES</codeph> clause
          specifies that the first column (<codeph>ID</codeph>) represents the row key, and maps the remaining
          columns of the SQL table to HBase column families. The mapping relies on the ordinal order of the
          columns in the table, not the column names in the <codeph>CREATE TABLE</codeph> statement.
          The first column is defined to be the lookup key; the
          <codeph>STRING</codeph> data type produces the fastest key-based lookups for HBase tables.
        </p>

        <note>
          For Impala with HBase tables, the most important aspect to ensure good performance is to use a
          <codeph>STRING</codeph> column as the row key, as shown in this example.
        </note>

<codeblock>$ hive
...
hive> use hbase;
OK
Time taken: 4.095 seconds
hive> CREATE EXTERNAL TABLE hbasestringids (
    >   id string,
    >   bool_col boolean,
    >   tinyint_col tinyint,
    >   smallint_col smallint,
    >   int_col int,
    >   bigint_col bigint,
    >   float_col float,
    >   double_col double,
    >   date_string_col string,
    >   string_col string,
    >   timestamp_col timestamp)
    > STORED BY 'org.apache.hadoop.hive.hbase.HBaseStorageHandler'
    > WITH SERDEPROPERTIES (
    >   "hbase.columns.mapping" =
    >   ":key,boolsCF:bool_col,intsCF:tinyint_col,intsCF:smallint_col,intsCF:int_col,intsCF:\
    >   bigint_col,floatsCF:float_col,floatsCF:double_col,stringsCF:date_string_col,\
    >   stringsCF:string_col,stringsCF:timestamp_col"
    > )
    > TBLPROPERTIES("hbase.table.name" = "hbasealltypessmall");
OK
Time taken: 2.879 seconds
hive> quit;
</codeblock>

        <p>
          Once you have established the mapping to an HBase table, you can issue DML statements and queries
          from Impala. The following example shows a series of <codeph>INSERT</codeph>
          statements followed by a query.
          The ideal kind of query from a performance standpoint
          retrieves a row from the table based on a row key
          mapped to a string column.
          An initial <codeph>INVALIDATE METADATA <varname>table_name</varname></codeph>
          statement makes the table created through Hive visible to Impala.
        </p>

<codeblock>$ impala-shell -i localhost -d hbase
Starting Impala Shell without Kerberos authentication
Connected to localhost:21000
...
Query: use `hbase`
[localhost:21000] > invalidate metadata hbasestringids;
Fetched 0 row(s) in 0.09s
[localhost:21000] > desc hbasestringids;
+-----------------+-----------+---------+
| name            | type      | comment |
+-----------------+-----------+---------+
| id              | string    |         |
| bool_col        | boolean   |         |
| double_col      | double    |         |
| float_col       | float     |         |
| bigint_col      | bigint    |         |
| int_col         | int       |         |
| smallint_col    | smallint  |         |
| tinyint_col     | tinyint   |         |
| date_string_col | string    |         |
| string_col      | string    |         |
| timestamp_col   | timestamp |         |
+-----------------+-----------+---------+
Fetched 11 row(s) in 0.02s
[localhost:21000] > insert into hbasestringids values ('0001',true,3.141,9.94,1234567,32768,4000,76,'2014-12-31','Hello world',now());
Inserted 1 row(s) in 0.26s
[localhost:21000] > insert into hbasestringids values ('0002',false,2.004,6.196,1500,8000,129,127,'2014-01-01','Foo bar',now());
Inserted 1 row(s) in 0.12s
[localhost:21000] > select * from hbasestringids where id = '0001';
+------+----------+------------+-------------------+------------+---------+--------------+-------------+-----------------+-------------+-------------------------------+
| id   | bool_col | double_col | float_col         | bigint_col | int_col | smallint_col | tinyint_col | date_string_col | string_col  | timestamp_col                 |
+------+----------+------------+-------------------+------------+---------+--------------+-------------+-----------------+-------------+-------------------------------+
| 0001 | true     | 3.141      | 9.939999580383301 | 1234567    | 32768   | 4000         | 76          | 2014-12-31      | Hello world | 2015-02-10 16:36:59.764838000 |
+------+----------+------------+-------------------+------------+---------+--------------+-------------+-----------------+-------------+-------------------------------+
Fetched 1 row(s) in 0.54s
</codeblock>

        <note conref="../shared/impala_common.xml#common/invalidate_metadata_hbase"/>
    </conbody>
  </concept>
</concept>