Indexes
| Table of Contents |
|---|
Introduction
This document explains the proposed design for adding index support to Hive (HIVE-417). Indexing is a standard database technique, but with many possible variations. Rather than trying to provide a "one-size-fits-all" index implementation, the approach we are taking is to define indexing in a pluggable manner (related to StorageHandlers) and provide one concrete indexing implementation as a reference, leaving it open for contributors to plug in other indexing schemes as time goes by. No index support will be available until Hive 0.7.
...
This document currently only covers index creation and maintenance. A follow-on will explain how indexes are used to optimize queries (building on FilterPushdownDev).
CREATE INDEX
| Code Block |
|---|
CREATE INDEX index_name
ON TABLE base_table_name (col_name, ...)
AS 'index.handler.class.name'
[WITH DEFERRED REBUILD]
[IDXPROPERTIES (property_name=property_value, ...)]
[IN TABLE index_table_name]
[PARTITIONED BY (col_name, ...)]
[
[ ROW FORMAT ...] STORED AS ...
| STORED BY ...
]
[LOCATION hdfs_path]
[TBLPROPERTIES (...)]
[COMMENT "index comment"]
|
...
So, given the following DDL:
| Code Block |
|---|
CREATE TABLE t(i int, j int);
CREATE INDEX x ON TABLE t(j)
AS 'org.apache.hadoop.hive.ql.index.compact.CompactIndexHandler';
|
...
To avoid the generated name, a user-specified name such as t_x can be supplied instead:
| Code Block |
|---|
CREATE INDEX x ON TABLE t(j)
AS 'org.apache.hadoop.hive.ql.index.compact.CompactIndexHandler'
IN TABLE t_x;
|
...
- change IDX_TYPE to IDX_HANDLER
- what does LAST_ACCESS_TIME mean? last time the optimizer used this index?
- need LAST_REBUILD_TIME? how do we track it at partition-level? it should be in the metastore (not just HDFS)
- in the case where the index partitioning is a subset of the base table partitioning, we need a way to model this in the metastore
Metastore Upgrades
Here are the MySQL metastore upgrade statements.
| Code Block |
|---|
DROP TABLE IF EXISTS {{IDXS}};
CREATE TABLE {{IDXS}} (
{{INDEX_ID}} bigint(20) NOT NULL,
{{CREATE_TIME}} int(11) NOT NULL,
{{DEFERRED_REBUILD}} bit(1) NOT NULL,
{{INDEX_HANDLER_CLASS}} varchar(256) DEFAULT NULL,
{{INDEX_NAME}} varchar(128) DEFAULT NULL,
{{INDEX_TBL_ID}} bigint(20) DEFAULT NULL,
{{LAST_ACCESS_TIME}} int(11) NOT NULL,
{{ORIG_TBL_ID}} bigint(20) DEFAULT NULL,
{{SD_ID}} bigint(20) DEFAULT NULL,
PRIMARY KEY ({{INDEX_ID}}),
UNIQUE KEY {{UNIQUEINDEX}} ({{INDEX_NAME}},{{ORIG_TBL_ID}}),
KEY {{IDXS_FK1}} ({{SD_ID}}),
KEY {{IDXS_FK2}} ({{INDEX_TBL_ID}}),
KEY {{IDXS_FK3}} ({{ORIG_TBL_ID}}),
CONSTRAINT {{IDXS_FK3}} FOREIGN KEY ({{ORIG_TBL_ID}}) REFERENCES {{TBLS}} ({{TBL_ID}}),
CONSTRAINT {{IDXS_FK1}} FOREIGN KEY ({{SD_ID}}) REFERENCES {{SDS}} ({{SD_ID}}),
CONSTRAINT {{IDXS_FK2}} FOREIGN KEY ({{INDEX_TBL_ID}}) REFERENCES {{TBLS}} ({{TBL_ID}})
) ENGINE=InnoDB DEFAULT CHARSET=latin1;
--
-- Table structure for table {{INDEX_PARAMS}}
--
DROP TABLE IF EXISTS {{INDEX_PARAMS}};
CREATE TABLE {{INDEX_PARAMS}} (
{{INDEX_ID}} bigint(20) NOT NULL,
{{PARAM_KEY}} varchar(256) NOT NULL,
{{PARAM_VALUE}} varchar(767) DEFAULT NULL,
PRIMARY KEY ({{INDEX_ID}},{{PARAM_KEY}}),
CONSTRAINT {{INDEX_PARAMS_FK1}} FOREIGN KEY ({{INDEX_ID}}) REFERENCES {{IDXS}} ({{INDEX_ID}})
) ENGINE=InnoDB DEFAULT CHARSET=latin1;
|
REBUILD
| Code Block |
|---|
ALTER INDEX index_name ON table_name [PARTITION (...)] REBUILD
|
...
If data in the base table changes, then the REBUILD command must be used to bring the index up to date. This is an atomic operation, so if the table was previously indexed, and a rebuild fails, then the stale index remains untouched.
DROP INDEX
| Code Block |
|---|
DROP INDEX index_name ON table_name
|
...
The corresponding Java inerface is defined below, together with a companion abstract base class which handlers should extend.
| Code Block |
|---|
package org.apache.hadoop.hive.ql.metadata;
import java.util.List;
import org.apache.hadoop.conf.Configurable;
import org.apache.hadoop.hive.ql.plan.api.Task;
/**
* HiveIndexHandler defines a pluggable interface for adding new
* index handlers to Hive.
*/
public interface HiveIndexHandler extends Configurable
{
/**
* Determines whether this handler implements indexes by creating
* an index table.
*
* @return true if index creation implies creation of an index table in Hive;
* false if the index representation is not stored in a Hive table
*/
boolean usesIndexTable();
/**
* Requests that the handler validate an index definition and
* fill in additional information about its stored representation.
*
* @param baseTable the definition of the table being indexed
*
* @param index the definition of the index being created
*
* @param indexTable a partial definition of the index table to be used for
* storing the index representation, or null if usesIndexTable() returns
* false; the handler can augment the index's storage descriptor
* (e.g. with information about input/output format)
* and/or the index table's definition (typically with additional
* columns containing the index representation, e.g. pointers into HDFS)
*
* @throw HiveException if the index definition is invalid with
* respect to either the base table or the supplied index table definition
*/
void analyzeIndexDefinition(
org.apache.hadoop.hive.metastore.api.Table baseTable,
org.apache.hadoop.hive.metastore.api.Index index,
org.apache.hadoop.hive.metastore.api.Table indexTable)
throws HiveException;
/**
* Requests that the handler generate a plan for building the index;
* the plan should read the base table and write out the index representation.
*
* @param baseTable the definition of the table being indexed
*
* @param index the definition of the index
*
* @param partitions a list of specific partitions of the base
* table for which the index should be built, or null if
* an index for the entire table should be rebuilt
*
* @param indexTable the definition of the index table, or
* null if usesIndexTable() returns null
*
* @return list of tasks to be executed in parallel for building
* the index
*
* @throw HiveException if plan generation fails
*/
List<Task<?>> generateIndexBuildTaskList(
org.apache.hadoop.hive.metastore.api.Table baseTable,
org.apache.hadoop.hive.metastore.api.Index index,
List<org.apache.hadoop.hive.metastore.api.Partition> partitions,
org.apache.hadoop.hive.metastore.api.Table indexTable)
throws HiveException;
}
/**
* Abstract base class for index handlers. This is provided as insulation
* so that as HiveIndexHandler evolves, default implementations of new
* methods can be added here in order to avoid breaking existing
* plugin implementations.
*/
public abstract class AbstractIndexHandler implements HiveIndexHandler
{
}
|
...
The reference implementation can be plugged in with
| Code Block |
|---|
ADD JAR /path/to/hive_idx-compact.jar;
CREATE INDEX ...
AS 'org.apache.hadoop.hive.ql.index.compact.CompactIndexHandler';
|
...
- specs for SHOW/DESCRIBE INDEX (HIVE-1497)
- ALTER INDEX DROP PARTITION?
- ALTER INDEX SET IDXPROPERTIES, change tableformat, etc
- what happens when the structure of a table or partition changes after it has already been indexed
- automatic indexing as part of INSERT when WITH DEFERRED REBUILD is not specified
- prevent creation of an index on an index table?
- metastore upgrade script
- stats collection for index tables
- intersection with new Hive concurrency control feature (what locks do we take for various index operations?)
Current Status (JIRA)
| Jira Issues | ||
|---|---|---|
|