impala/docs/topics/impala_scan_bytes_limit.xml

<?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="scan_bytes_limit">

  <title>SCAN_BYTES_LIMIT Query Option (<keyword keyref="impala31"/> or higher only)</title>

  <titlealts audience="PDF">

    <navtitle>SCAN_BYTES_LIMIT</navtitle>

  </titlealts>

  <prolog>
    <metadata>
      <data name="Category" value="Impala"/>
      <data name="Category" value="Impala Query Options"/>
      <data name="Category" value="Scalability"/>
      <data name="Category" value="Memory"/>
      <data name="Category" value="Troubleshooting"/>
      <data name="Category" value="Developers"/>
      <data name="Category" value="Data Analysts"/>
    </metadata>
  </prolog>

  <conbody>

    <p>
      The <codeph>SCAN_BYTES_LIMIT</codeph> query option sets a limit on the bytes scanned by
      HDFS and HBase SCAN operations. If a query is still executing when the query’s
      coordinator detects that it has exceeded the limit, the query is terminated with an error.
      The option is intended to prevent runaway queries that scan more data than is intended.
    </p>

    <p>
      For example, an Impala administrator could set a default value of
      <codeph>SCAN_BYTES_LIMIT=100GB</codeph> for a resource pool to automatically kill queries
      that scan more than 100 GB of data (see
      <xref
        href="https://impala.apache.org/docs/build/html/topics/impala_admission.html"
        format="html" scope="external">Impala
      Admission Control and Query Queuing</xref> for information about default query options).
      If a user accidentally omits a partition filter in a <codeph>WHERE</codeph> clause and
      runs a large query that scans a lot of data, the query will be automatically terminated
      after it scans more data than the <codeph>SCAN_BYTES_LIMIT</codeph>.
    </p>

    <p>
      You can override the default value per-query or per-session, in the same way as other
      query options, if you do not want the default <codeph>SCAN_BYTES_LIMIT</codeph> value to
      apply to a specific query or session.
      <note>
        <ul>
          <li dir="ltr">
            <p dir="ltr">
              Only data actually read from the underlying storage layer is counted towards the
              limit. E.g. Impala’s Parquet scanner employs several techniques to skip over
              data in a file that is not relevant to a specific query, so often only a fraction
              of the file size is counted towards <codeph>SCAN_BYTES_LIMIT</codeph>.
            </p>
          </li>

          <li dir="ltr">
            <p dir="ltr">
              As of Impala 3.1, bytes scanned by Kudu tablet servers are not counted towards the
              limit.
            </p>
          </li>
        </ul>
      </note>
    </p>

    <p>
      Because the checks are done periodically, the query may scan over the limit at times.
    </p>

    <p>
      <b>Syntax:</b> <codeph>SET SCAN_BYTES_LIMIT=bytes;</codeph>
    </p>

    <p>
      <b>Type:</b> numeric
    </p>

    <p>
      <b>Units:</b>
      <ul>
        <li>
          A numeric argument represents memory size in bytes.
        </li>

        <li>
          Specify a suffix of <codeph>m</codeph> or <codeph>mb</codeph> for megabytes.
        </li>

        <li>
          Specify a suffix of <codeph>g</codeph> or <codeph>gb</codeph> for gigabytes.
        </li>

        <li>
          If you specify a suffix with unrecognized formats, subsequent queries fail with an
          error.
        </li>
      </ul>
    </p>

    <p>
      <b>Default:</b> <codeph>0</codeph> (no limit)
    </p>

    <p>
      <b>Added in:</b> <keyword keyref="impala31"/>
    </p>

  </conbody>

</concept>