HiveQL DDL statements are documented here, including:
PARTITION statements are usually options of TABLE statements, except for SHOW PARTITIONS.
|added: ||added: |
REGEXP and RLIKE are non-reserved keywords prior to Hive 2.0.0 and reserved keywords starting in Hive 2.0.0 (HIVE-11703).
Reserved keywords are permitted as identifiers if you quote them as described in Supporting Quoted Identifiers in Column Names (version 0.13.0 and later, see HIVE-6013). Most of the keywords are reserved through HIVE-6617 in order to reduce the ambiguity in grammar (version 1.2.0 and later). There are two ways if the user still would like to use those reserved keywords as identifiers: (1) use quoted identifiers, (2) set hive.support.sql11.reserved.keywords=false. (version 2.1.0 and earlier)
CREATE (DATABASE|SCHEMA) [IF NOT EXISTS] database_name [COMMENT database_comment] [LOCATION hdfs_path] [WITH DBPROPERTIES (property_name=property_value, ...)];
The uses of SCHEMA and DATABASE are interchangeable – they mean the same thing. CREATE DATABASE was added in Hive 0.6 (HIVE-675). The WITH DBPROPERTIES clause was added in Hive 0.7 (HIVE-1836).
DROP (DATABASE|SCHEMA) [IF EXISTS] database_name [RESTRICT|CASCADE];
The uses of SCHEMA and DATABASE are interchangeable – they mean the same thing. DROP DATABASE was added in Hive 0.6 (HIVE-675). The default behavior is RESTRICT, where DROP DATABASE will fail if the database is not empty. To drop the tables in the database as well, use DROP DATABASE ... CASCADE. Support for RESTRICT and CASCADE was added in Hive 0.8 (HIVE-2090).
ALTER (DATABASE|SCHEMA) database_name SET DBPROPERTIES (property_name=property_value, ...); -- (Note: SCHEMA added in Hive 0.14.0) ALTER (DATABASE|SCHEMA) database_name SET OWNER [USER|ROLE] user_or_role; -- (Note: Hive 0.13.0 and later; SCHEMA added in Hive 0.14.0) ALTER (DATABASE|SCHEMA) database_name SET LOCATION hdfs_path; -- (Note: Hive 2.2.1, 2.4.0 and later)
The uses of SCHEMA and DATABASE are interchangeable – they mean the same thing. ALTER SCHEMA was added in Hive 0.14 (HIVE-6601).
The ALTER DATABASE ... SET LOCATION statement does not move the contents of the database's current directory to the newly specified location. It does not change the locations associated with any tables/partitions under the specified database. It only changes the default parent-directory where new tables will be added for this database. This behaviour is analogous to how changing a table-directory does not move existing partitions to a different location.
No other metadata about a database can be changed.
USE database_name; USE DEFAULT;
USE sets the current database for all subsequent HiveQL statements. To revert to the default database, use the keyword "
default" instead of a database name. To check which database is currently being used:
SELECT current_database() (as of Hive 0.13.0).
USE database_name was added in Hive 0.6 (HIVE-675).
CREATE TABLE creates a table with the given name. An error is thrown if a table or view with the same name already exists. You can use IF NOT EXISTS to skip the error.
See Alter Table below for more information about table comments, table properties, and SerDe properties.
Managed and External Tables
By default Hive creates managed tables, where files, metadata and statistics are managed by internal Hive processes. A managed table is stored under the hive.metastore.warehouse.dir path property, by default in a folder path similar to /apps/hive/warehouse/databasename.db/tablename/. The default location can be overridden by the location property during table creation. If a managed table or partition is dropped, the data and metadata associated with that table or partition are deleted. If the PURGE option is not specified, the data is moved to a trash folder for a defined duration.
Use managed tables when Hive should manage the lifecycle of the table, or when generating temporary tables.
An external table describes the metadata / schema on external files. External table files can be accessed and managed by processes outside of Hive. External tables can access data stored in sources such as Azure Storage Volumes (ASV) or remote HDFS locations. If the structure or partitioning of an external table is changed, an MSCK REPAIR TABLE table_name statement can be used to refresh metadata information.
Use external tables when files are already present or in remote locations, and the files should remain even if the table is dropped.
Managed or external tables can be identified using the DESCRIBE FORMATTED table_name command, which will display either MANAGED_TABLE or EXTERNAL_TABLE depending on table type.
Statistics can be managed on internal and external tables and partitions for query optimization.
Hive supports built-in and custom-developed file formats. See CompressedStorage for details on compressed table storage.
Stored as plain text file, translated by Regular Expression.
The following example defines a table in the default Apache Weblog format.
ROW FORMAT SERDE
STORED AS TEXTFILE
Stored as plain text file in JSON format.
The JsonSerDe for JSON files is available in Hive 0.12 and later.
In some distributions, a reference to hive-hcatalog-core.jar is required.
The JsonSerDe was moved to Hive from HCatalog and before it was in hive-contrib project. It was added to the Hive distribution by HIVE-4895.
An Amazon SerDe is available at
The JsonSerDe for JSON files is available in Hive 0.12 and later.
Starting in Hive 3.0.0, JsonSerDe is added to Hive Serde as "org.apache.hadoop.hive.serde2.JsonSerDe" (HIVE-19211).
ROW FORMAT SERDE
STORED AS TEXTFILE
Stored as plain text file in CSV / TSV format.
The CSVSerde is available in Hive 0.14 and greater.
The following example creates a TSV (Tab-separated) file.
Default properties for SerDe is Comma-Separated (CSV) file
This SerDe works for most CSV data, but does not handle embedded newlines. To use the SerDe, specify the fully qualified class name org.apache.hadoop.hive.serde2.OpenCSVSerde.
Documentation is based on original documentation at https://github.com/ogrodnek/csv-serde.
This SerDe treats all columns to be of type String. Even if you create a table with non-string column types using this SerDe, the DESCRIBE TABLE output would show string column type.
The type information is retrieved from the SerDe.
To convert columns to the desired type in a table, you can create a view over the table that does the CAST to the desired type.
The CSV SerDe is based on https://github.com/ogrodnek/csv-serde, and was added to the Hive distribution in HIVE-7777.The CSVSerde has been built and tested against Hive 0.14 and later, and uses Open-CSV 2.3 which is bundled with the Hive distribution.
For general information about SerDes, see Hive SerDe in the Developer Guide. Also see SerDe for details about input and output processing.
Partitioned tables can be created using the PARTITIONED BY clause. A table can have one or more partition columns and a separate data directory is created for each distinct value combination in the partition columns. Further, tables or partitions can be bucketed using CLUSTERED BY columns, and data can be sorted within that bucket via SORT BY columns. This can improve performance on certain kinds of queries.
If, when creating a partitioned table, you get this error: "FAILED: Error in semantic analysis: Column repeated in partitioning columns," it means you are trying to include the partitioned column in the data of the table itself. You probably really do have the column defined. However, the partition you create makes a pseudocolumn on which you can query, so you must rename your table column to something else (that users should not query on!).
For example, suppose your original unpartitioned table had three columns: id, date, and name.
id int, date date, name varchar
Now you want to partition on date. Your Hive definition could use "dtDontQuery" as a column name so that "date" can be used for partitioning (and querying).
create table table_name ( id int, dtDontQuery string, name string ) partitioned by (date string)
Now your users will still query on "
where date = '...'" but the second column dtDontQuery will hold the original values.
Here's an example statement to create a partitioned table:
CREATE TABLE page_view(viewTime INT, userid BIGINT, page_url STRING, referrer_url STRING, ip STRING COMMENT 'IP Address of the User') COMMENT 'This is the page view table' PARTITIONED BY(dt STRING, country STRING) STORED AS SEQUENCEFILE;
The statement above creates the page_view table with viewTime, userid, page_url, referrer_url, and ip columns (including comments). The table is also partitioned and data is stored in sequence files. The data format in the files is assumed to be field-delimited by ctrl-A and row-delimited by newline.
CREATE TABLE page_view(viewTime INT, userid BIGINT, page_url STRING, referrer_url STRING, ip STRING COMMENT 'IP Address of the User') COMMENT 'This is the page view table' PARTITIONED BY(dt STRING, country STRING) ROW FORMAT DELIMITED FIELDS TERMINATED BY '\001' STORED AS SEQUENCEFILE;
The above statement lets you create the same table as the previous table.
In the previous examples the data is stored in <hive.metastore.warehouse.dir>/page_view. Specify a value for the key
hive.metastore.warehouse.dir in the Hive config file hive-site.xml.
The EXTERNAL keyword lets you create a table and provide a LOCATION so that Hive does not use a default location for this table. This comes in handy if you already have data generated. When dropping an EXTERNAL table, data in the table is NOT deleted from the file system.
An EXTERNAL table points to any HDFS location for its storage, rather than being stored in a folder specified by the configuration property
CREATE EXTERNAL TABLE page_view(viewTime INT, userid BIGINT, page_url STRING, referrer_url STRING, ip STRING COMMENT 'IP Address of the User', country STRING COMMENT 'country of origination') COMMENT 'This is the staging page view table' ROW FORMAT DELIMITED FIELDS TERMINATED BY '\054' STORED AS TEXTFILE LOCATION '<hdfs_location>';
You can use the above statement to create a page_view table which points to any HDFS location for its storage. But you still have to make sure that the data is delimited as specified in the CREATE statement above.
For another example of creating an external table, see Loading Data in the Tutorial.
Tables can also be created and populated by the results of a query in one create-table-as-select (CTAS) statement. The table created by CTAS is atomic, meaning that the table is not seen by other users until all the query results are populated. So other users will either see the table with the complete results of the query or will not see the table at all.
There are two parts in CTAS, the SELECT part can be any SELECT statement supported by HiveQL. The CREATE part of the CTAS takes the resulting schema from the SELECT part and creates the target table with other table properties such as the SerDe and storage format.
CTAS has these restrictions:
CREATE TABLE new_key_value_store ROW FORMAT SERDE "org.apache.hadoop.hive.serde2.columnar.ColumnarSerDe" STORED AS RCFile AS SELECT (key % 1024) new_key, concat(key, value) key_value_pair FROM key_value_store SORT BY new_key, key_value_pair;
The above CTAS statement creates the target table new_key_value_store with the schema (new_key DOUBLE, key_value_pair STRING) derived from the results of the SELECT statement. If the SELECT statement does not specify column aliases, the column names will be automatically assigned to _col0, _col1, and _col2 etc. In addition, the new target table is created using a specific SerDe and a storage format independent of the source tables in the SELECT statement.
Being able to select data from one table to another is one of the most powerful features of Hive. Hive handles the conversion of the data from the source format to the destination format as the query is being executed.
The LIKE form of CREATE TABLE allows you to copy an existing table definition exactly (without copying its data). In contrast to CTAS, the statement below creates a new empty_key_value_store table whose definition exactly matches the existing key_value_store in all particulars other than table name. The new table contains no rows.
CREATE TABLE empty_key_value_store LIKE key_value_store [TBLPROPERTIES (property_name=property_value, ...)];
Before Hive 0.8.0, CREATE TABLE LIKE view_name would make a copy of the view. In Hive 0.8.0 and later releases, CREATE TABLE LIKE view_name creates a table by adopting the schema of view_name (fields and partition columns) using defaults for SerDe and file formats.
CREATE TABLE page_view(viewTime INT, userid BIGINT, page_url STRING, referrer_url STRING, ip STRING COMMENT 'IP Address of the User') COMMENT 'This is the page view table' PARTITIONED BY(dt STRING, country STRING) CLUSTERED BY(userid) SORTED BY(viewTime) INTO 32 BUCKETS ROW FORMAT DELIMITED FIELDS TERMINATED BY '\001' COLLECTION ITEMS TERMINATED BY '\002' MAP KEYS TERMINATED BY '\003' STORED AS SEQUENCEFILE;
In the example above, the page_view table is bucketed (clustered by) userid and within each bucket the data is sorted in increasing order of viewTime. Such an organization allows the user to do efficient sampling on the clustered column - in this case userid. The sorting property allows internal operators to take advantage of the better-known data structure while evaluating queries, also increasing efficiency. MAP KEYS and COLLECTION ITEMS keywords can be used if any of the columns are lists or maps.
The CLUSTERED BY and SORTED BY creation commands do not affect how data is inserted into a table – only how it is read. This means that users must be careful to insert data correctly by specifying the number of reducers to be equal to the number of buckets, and using CLUSTER BY and SORT BY commands in their query.
There is also an example of creating and populating bucketed tables.
This feature can be used to improve performance for tables where one or more columns have skewed values. By specifying the values that appear very often (heavy skew) Hive will split those out into separate files (or directories in case of list bucketing) automatically and take this fact into account during queries so that it can skip or include the whole file (or directory in case of list bucketing) if possible.
This can be specified on a per-table level during table creation.
The following example shows one column with three skewed values, optionally with the STORED AS DIRECTORIES clause which specifies list bucketing.
CREATE TABLE list_bucket_single (key STRING, value STRING) SKEWED BY (key) ON (1,5,6) [STORED AS DIRECTORIES];
And here is an example of a table with two skewed columns.
CREATE TABLE list_bucket_multiple (col1 STRING, col2 int, col3 STRING) SKEWED BY (col1, col2) ON (('s1',1), ('s3',3), ('s13',13), ('s78',78)) [STORED AS DIRECTORIES];
For corresponding ALTER TABLE statements, see Alter Table Skewed or Stored as Directories below.
As of Hive 0.14.0 (HIVE-7090).
A table that has been created as a temporary table will only be visible to the current session. Data will be stored in the user's scratch directory, and deleted at the end of the session.
If a temporary table is created with a database/table name of a permanent table which already exists in the database, then within that session any references to that table will resolve to the temporary table, rather than to the permanent table. The user will not be able to access the original table within that session without either dropping the temporary table, or renaming it to a non-conflicting name.
Temporary tables have the following limitations:
Starting in Hive 1.1.0 the storage policy for temporary tables can be set to
default with the hive.exec.temporary.table.storage configuration parameter (see HDFS Storage Types and Storage Policies).
As of Hive 2.1.0 (HIVE-13290).
Hive includes support for non-validated primary and foreign key constraints. Some SQL tools generate more efficient queries when constraints are present. Since these constraints are not validated, an upstream system needs to ensure data integrity before it is loaded into Hive.
create table pk(id1 integer, id2 integer, primary key(id1, id2) disable novalidate); create table fk(id1 integer, id2 integer, constraint c1 foreign key(id1, id2) references pk(id2, id1) disable novalidate);
DROP TABLE [IF EXISTS] table_name [PURGE]; -- (Note: PURGE available in Hive 0.14.0 and later)
DROP TABLE removes metadata and data for this table. The data is actually moved to the .Trash/Current directory if Trash is configured (and PURGE is not specified). The metadata is completely lost.
When dropping an EXTERNAL table, data in the table will NOT be deleted from the file system.
When dropping a table referenced by views, no warning is given (the views are left dangling as invalid and must be dropped or recreated by the user).
Otherwise, the table information is removed from the metastore and the raw data is removed as if by 'hadoop dfs -rm'. In many cases, this results in the table data being moved into the user's .Trash folder in their home directory; users who mistakenly DROP TABLEs may thus be able to recover their lost data by recreating a table with the same schema, recreating any necessary partitions, and then moving the data back into place manually using Hadoop. This solution is subject to change over time or across installations as it relies on the underlying implementation; users are strongly encouraged not to drop tables capriciously.
The PURGE option is added in version 0.14.0 by HIVE-7100.
If PURGE is specified, the table data does not go to the .Trash/Current directory and so cannot be retrieved in the event of a mistaken DROP. The purge option can also be specified with the table property auto.purge (see TBLPROPERTIES above).
In Hive 0.7.0 or later, DROP returns an error if the table doesn't exist, unless IF EXISTS is specified or the configuration variable hive.exec.drop.ignorenonexistent is set to true.
See the Alter Partition section below for how to drop partitions.
As of Hive 0.11.0 (HIVE-446).
TRUNCATE TABLE table_name [PARTITION partition_spec]; partition_spec: : (partition_column = partition_col_value, partition_column = partition_col_value, ...)
Removes all rows from a table or partition(s). The rows will be trashed if the filesystem Trash is enabled, otherwise they are deleted (as of Hive 2.2.0 with HIVE-14626). Currently the target table should be native/managed table or an exception will be thrown. User can specify partial partition_spec for truncating multiple partitions at once and omitting partition_spec will truncate all partitions in the table.
Starting with HIVE 2.3.0 (HIVE-15880) if the table property "auto.purge" (see TBLPROPERTIES above) is set to "true" the data of the table is not moved to Trash when a TRUNCATE TABLE command is issued against it and cannot be retrieved in the event of a mistaken TRUNCATE. This is applicable only for managed tables (see managed tables). This behavior can be turned off if the "auto.purge" property is unset or set to false for a managed table.
Alter table statements enable you to change the structure of an existing table. You can add columns/partitions, change SerDe, add table and SerDe properties, or rename the table itself. Similarly, alter table partition statements allow you change the properties of a specific partition in the named table.
This statement lets you change the name of a table to a different name.
As of version 0.6, a rename on a managed table moves its HDFS location. Rename has been changed as of version 2.2.0 (HIVE-14909) so that a managed table's HDFS location is moved only if the table is created without a LOCATION clause and under its database directory. Hive versions prior to 0.6 just renamed the table in the metastore without moving the HDFS location.
Alter Table Properties
You can use this statement to add your own metadata to the tables. Currently last_modified_user, last_modified_time properties are automatically added and managed by Hive. Users can add their own properties to this list. You can do DESCRIBE EXTENDED TABLE to get this information.
For more information, see the TBLPROPERTIES clause in Create Table above.
Alter Table Comment
To change the comment of a table you have to change the
Add SerDe Properties
These statements enable you to change a table's SerDe or add user-defined metadata to the table's SerDe object.
The SerDe properties are passed to the table's SerDe when it is being initialized by Hive to serialize and deserialize data. So users can store any information required for their custom SerDe here. Refer to the SerDe documentation and Hive SerDe in the Developer Guide for more information, and see Row Format, Storage Format, and SerDe above for details about setting a table's SerDe and SERDEPROPERTIES in a CREATE TABLE statement.
Note that both
Alter Table Storage Properties
These statements change the table's physical storage properties.
NOTE: These commands will only modify Hive's metadata, and will NOT reorganize or reformat existing data. Users should make sure the actual data layout conforms with the metadata definition.
Alter Table Skewed or Stored as Directories
A table's SKEWED and STORED AS DIRECTORIES options can be changed with ALTER TABLE statements. See Skewed Tables above for the corresponding CREATE TABLE syntax.
Alter Table Skewed
Alter Table Not Skewed
The NOT SKEWED option makes the table non-skewed and turns off the list bucketing feature (since a list-bucketing table is always skewed). This affects partitions created after the ALTER statement, but has no effect on partitions created before the ALTER statement.
Alter Table Not Stored as Directories
This turns off the list bucketing feature, although the table remains skewed.
Alter Table Set Skewed Location
This changes the location map for list bucketing.
Alter Table Constraints
Table constraints can be added or removed via ALTER TABLE statements.
Additional Alter Table Statements
See Alter Either Table or Partition below for more DDL statements that alter tables.
Partitions can be added, renamed, exchanged (moved), dropped, or (un)archived by using the PARTITION clause in an ALTER TABLE statement, as described below. To make the metastore aware of partitions that were added directly to HDFS, you can use the metastore check command (MSCK) or on Amazon EMR you can use the RECOVER PARTITIONS option of ALTER TABLE. See Alter Either Table or Partition below for more ways to alter partitions.
You can use ALTER TABLE ADD PARTITION to add partitions to a table. Partition values should be quoted only if they are strings. The location must be a directory inside of which data files reside. (ADD PARTITION changes the table metadata, but does not load data. If the data does not exist in the partition's location, queries will not return any results.) An error is thrown if the partition_spec for the table already exists. You can use IF NOT EXISTS to skip the error.
Specifically, the following example will FAIL silently and without error in Hive 0.7, and all queries will go only to dt='2008-08-08' partition, no matter which partition you specify.
In Hive 0.8 and later, you can add multiple partitions in a single ALTER TABLE statement as shown in the previous example.
In Hive 0.7, if you want to add many partitions you should use the following form:
Partitions can be added to a table dynamically, using a Hive INSERT statement (or a Pig STORE statement). See these documents for details and examples:
This statement lets you change the value of a partition column. One of use cases is that you can use this statement to normalize your legacy partition column value to conform to its type. In this case, the type conversion and normalization are not enabled for the column values in old partition_spec even with property hive.typecheck.on.insert set to true (default) which allows you to specify any legacy data in form of string in the old partition_spec.
Partitions can be exchanged (moved) between tables.
This statement lets you move the data in a partition from a table to another table that has the same schema and does not already have that partition.
Recover Partitions (MSCK REPAIR TABLE)
Hive stores a list of partitions for each table in its metastore. If, however, new partitions are directly added to HDFS (say by using
However, users can run a metastore check command with the repair table option:
which will add metadata about partitions to the Hive metastore for partitions for which such metadata doesn't already exist. In other words, it will add any partitions that exist on HDFS but not in metastore to the metastore. See HIVE-874 for more details. When there is a large number of untracked partitions, there is a provision to run MSCK REPAIR TABLE batch wise to avoid OOME (Out of Memory Error). By giving the configured batch size for the property hive.msck.repair.batch.size it can run in the batches internally. The default value of the property is zero, it means it will execute all the partitions at once.
The equivalent command on Amazon Elastic MapReduce (EMR)'s version of Hive is:
Starting with Hive 1.3, MSCK will throw exceptions if directories with disallowed characters in partition values are found on HDFS. Use hive.msck.path.validation setting on the client to alter this behavior; "skip" will simply skip the directories. "ignore" will try to create partitions anyway (old behavior). This may or may not work.
You can use ALTER TABLE DROP PARTITION to drop a partition for a table. This removes the data and metadata for this partition. The data is actually moved to the .Trash/Current directory if Trash is configured, unless PURGE is specified, but the metadata is completely lost (see Drop Table above).
For tables that are protected by NO_DROP CASCADE, you can use the predicate IGNORE PROTECTION to drop a specified partition or set of partitions (for example, when splitting a table between two Hadoop clusters):
The above command will drop that partition regardless of protection stats.
If PURGE is specified, the partition data does not go to the .Trash/Current directory and so cannot be retrieved in the event of a mistaken DROP:
The purge option can also be specified with the table property auto.purge (see TBLPROPERTIES above).
In Hive 0.7.0 or later, DROP returns an error if the partition doesn't exist, unless IF EXISTS is specified or the configuration variable hive.exec.drop.ignorenonexistent is set to true.
Archiving is a feature to moves a partition's files into a Hadoop Archive (HAR). Note that only the file count will be reduced; HAR does not provide any compression. See LanguageManual Archiving for more information
Alter Either Table or Partition
Alter Table/Partition File Format
This statement changes the table's (or partition's) file format. For available file_format options, see the section above on CREATE TABLE. The operation only changes the table metadata. Any conversion of existing data must be done outside of Hive.
Alter Table/Partition Location
Alter Table/Partition Touch
TOUCH reads the metadata, and writes it back. This has the effect of causing the pre/post execute hooks to fire. An example use case is if you have a hook that logs all the tables/partitions that were modified, along with an external script that alters the files on HDFS directly. Since the script modifies files outside of hive, the modification wouldn't be logged by the hook. The external script could call TOUCH to fire the hook and mark the said table or partition as modified.
Also, it may be useful later if we incorporate reliable last modified times. Then touch would update that time as well.
Note that TOUCH doesn't create a table or partition if it doesn't already exist. (See Create Table.)
Alter Table/Partition Protections
Protection on data can be set at either the table or partition level. Enabling NO_DROP prevents a table from being dropped. Enabling OFFLINE prevents the data in a table or partition from being queried, but the metadata can still be accessed.
If any partition in a table has NO_DROP enabled, the table cannot be dropped either. Conversely, if a table has NO_DROP enabled then partitions may be dropped, but with NO_DROP CASCADE partitions cannot be dropped either unless the drop partition command specifies IGNORE PROTECTION.
Alter Table/Partition Compact
In general you do not need to request compactions when Hive transactions are being used, because the system will detect the need for them and initiate the compaction. However, if compaction is turned off for a table or you want to compact the table at a time the system would not choose to, ALTER TABLE can initiate the compaction. By default the statement will enqueue a request for compaction and return. To watch the progress of the compaction, use SHOW COMPACTIONS. As of Hive 2.2.0 "AND WAIT" may be specified to have the operation block until compaction completes.
The compaction_type can be MAJOR or MINOR. See the Basic Design section in Hive Transactions for more information.
Alter Table/Partition Concatenate
If the table or partition contains many small RCFiles or ORC files, then the above command will merge them into larger files. In case of RCFile the merge happens at block level whereas for ORC files the merge happens at stripe level thereby avoiding the overhead of decompressing and decoding the data.
Alter Table/Partition Update columns
Tables that have serdes which self-describe the table schema may have different schemas in reality and the ones stored in Hive Metastore. For example when a user creates an Avro stored table using a schema url or schema literal, the schema will be inserted into HMS and then will never be changed in HMS regardless of url or literal changes within the serde. This can lead to problems especially when integrating with other Apache components.
The update columns feature provides a way for the user to let any schema changes made in the serde to be synced into HMS. It works on both the table and the partitions levels, and obviously only for tables whose schema is not tracked by HMS (see metastore.serdes.using.metastore.for.schema). Using the command on these latter serde types will result in error.
Rules for Column Names
Column names are case insensitive.
Change Column Name/Type/Position/Comment
This command will allow users to change a column's name, data type, comment, or position, or an arbitrary combination of them. The PARTITION clause is available in Hive 0.14.0 and later; see Upgrading Pre-Hive 0.13.0 Decimal Columns for usage. A patch for Hive 0.13 is also available (see HIVE-7971).
The CASCADE|RESTRICT clause is available in Hive 1.1.0. ALTER TABLE CHANGE COLUMN with CASCADE command changes the columns of a table's metadata, and cascades the same change to all the partition metadata. RESTRICT is the default, limiting column change only to table metadata.
ADD COLUMNS lets you add new columns to the end of the existing columns but before the partition columns. This is supported for Avro backed tables as well, for Hive 0.14 and later.
REPLACE COLUMNS removes all existing columns and adds the new set of columns. This can be done only for tables with a native SerDe (DynamicSerDe, MetadataTypedColumnsetSerDe, LazySimpleSerDe and ColumnarSerDe). Refer to Hive SerDe for more information. REPLACE COLUMNS can also be used to drop columns. For example, "
The CASCADE|RESTRICT clause is available in Hive 1.1.0. ALTER TABLE ADD|REPLACE COLUMNS with CASCADE command changes the columns of a table's metadata, and cascades the same change to all the partition metadata. RESTRICT is the default, limiting column changes only to table metadata.
Partial Partition Specification
As of Hive 0.14 (HIVE-8411), users are able to provide a partial partition spec for certain above alter column statements, similar to dynamic partitioning. So rather than having to issue an alter column statement for each partition that needs to be changed:
... you can change many existing partitions at once using a single ALTER statement with a partial partition specification:
Similar to dynamic partitioning, hive.exec.dynamic.partition must be set to true to enable use of partial partition specs during ALTER PARTITION. This is supported for the following operations:
CREATE VIEW creates a view with the given name. An error is thrown if a table or view with the same name already exists. You can use IF NOT EXISTS to skip the error.
If no column names are supplied, the names of the view's columns will be derived automatically from the defining SELECT expression. (If the SELECT contains unaliased scalar expressions such as x+y, the resulting view column names will be generated in the form _C0, _C1, etc.) When renaming columns, column comments can also optionally be supplied. (Comments are not automatically inherited from underlying columns.)
A CREATE VIEW statement will fail if the view's defining SELECT expression is invalid.
Note that a view is a purely logical object with no associated storage (except for materialized views, which Hive supports starting in release 2.3.0.) When a query references a view, the view's definition is evaluated in order to produce a set of rows for further processing by the query. (This is a conceptual description; in fact, as part of query optimization, Hive may combine the view's definition with the query's, e.g. pushing filters from the query down into the view.)
A view's schema is frozen at the time the view is created; subsequent changes to underlying tables (e.g. adding a column) will not be reflected in the view's schema. If an underlying table is dropped or changed in an incompatible fashion, subsequent attempts to query the invalid view will fail.
Views are read-only and may not be used as the target of LOAD/INSERT/ALTER. For changing metadata, see ALTER VIEW.
A view may contain ORDER BY and LIMIT clauses. If a referencing query also contains these clauses, the query-level clauses are evaluated after the view clauses (and after any other operations in the query). For example, if a view specifies LIMIT 5, and a referencing query is executed as (select * from v LIMIT 10), then at most 5 rows will be returned.
Starting with Hive 0.13.0, the view's select statement can include one or more common table expressions (CTEs) as shown in the SELECT syntax. For examples of CTEs in CREATE VIEW statements, see Common Table Expression.
DROP VIEW removes metadata for the specified view. (It is illegal to use DROP TABLE on a view.)
When dropping a view referenced by other views, no warning is given (the dependent views are left dangling as invalid and must be dropped or recreated by the user).
In Hive 0.7.0 or later, DROP returns an error if the view doesn't exist, unless IF EXISTS is specified or the configuration variable hive.exec.drop.ignorenonexistent is set to true.
Alter View Properties
As with ALTER TABLE, you can use this statement to add your own metadata to a view.
Alter View As Select
Alter View As Select changes the definition of a view, which must exist. The syntax is similar to that for CREATE VIEW and the effect is the same as for CREATE OR REPLACE VIEW.
Note: The view must already exist, and if the view has partitions, it could not be replaced by Alter View As Select.
As of Hive 0.7.
This section provides a brief introduction to Hive indexes, which are documented more fully here:
In Hive 0.12.0 and earlier releases, the index name is case-sensitive for CREATE INDEX and DROP INDEX statements. However, ALTER INDEX requires an index name that was created with lowercase letters (see HIVE-2752). This bug is fixed in Hive 0.13.0 by making index names case-insensitive for all HiveQL statements. For releases prior to 0.13.0, the best practice is to use lowercase letters for all index names.
CREATE INDEX index_name ON TABLE base_table_name (col_name, ...) AS index_type [WITH DEFERRED REBUILD] [IDXPROPERTIES (property_name=property_value, ...)] [IN TABLE index_table_name] [ [ ROW FORMAT ...] STORED AS ... | STORED BY ... ] [LOCATION hdfs_path] [TBLPROPERTIES (...)] [COMMENT "index comment"];
CREATE INDEX creates an index on a table using the given list of columns as keys. See CREATE INDEX in the Indexes design document.
DROP INDEX [IF EXISTS] index_name ON table_name;
DROP INDEX drops the index, as well as deleting the index table.
In Hive 0.7.0 or later, DROP returns an error if the index doesn't exist, unless IF EXISTS is specified or the configuration variable hive.exec.drop.ignorenonexistent is set to true.
ALTER INDEX index_name ON table_name [PARTITION partition_spec] REBUILD;
ALTER INDEX ... REBUILD builds an index that was created using the WITH DEFERRED REBUILD clause, or rebuilds a previously built index. If PARTITION is specified, only that partition is rebuilt.
As of Hive 0.12.0.
Hive 0.12.0 introduced macros to HiveQL, prior to which they could only be created in Java.
CREATE TEMPORARY MACRO macro_name([col_name col_type, ...]) expression;
CREATE TEMPORARY MACRO creates a macro using the given optional list of columns as inputs to the expression. Macros exist for the duration of the current session.
CREATE TEMPORARY MACRO fixed_number() 42; CREATE TEMPORARY MACRO string_len_plus_two(x string) length(x) + 2; CREATE TEMPORARY MACRO simple_add (x int, y int) x + y;
DROP TEMPORARY MACRO [IF EXISTS] macro_name;
DROP TEMPORARY MACRO returns an error if the function doesn't exist, unless IF EXISTS is specified.
CREATE TEMPORARY FUNCTION function_name AS class_name;
This statement lets you create a function that is implemented by the class_name. You can use this function in Hive queries as long as the session lasts. You can use any class that is in the class path of Hive. You can add jars to class path by executing 'ADD JAR' statements. Please refer to the CLI section Hive Interactive Shell Commands, including Hive Resources, for more information on how to add/delete files from the Hive classpath. Using this, you can register User Defined Functions (UDF's).
Also see Hive Plugins for general information about creating custom UDFs.
You can unregister a UDF as follows:
DROP TEMPORARY FUNCTION [IF EXISTS] function_name;
In Hive 0.7.0 or later, DROP returns an error if the function doesn't exist, unless IF EXISTS is specified or the configuration variable hive.exec.drop.ignorenonexistent is set to true.
In Hive 0.13 or later, functions can be registered to the metastore, so they can be referenced in a query without having to create a temporary function each session.
As of Hive 0.13.0 (HIVE-6047).
CREATE FUNCTION [db_name.]function_name AS class_name [USING JAR|FILE|ARCHIVE 'file_uri' [, JAR|FILE|ARCHIVE 'file_uri'] ];
This statement lets you create a function that is implemented by the class_name. Jars, files, or archives which need to be added to the environment can be specified with the USING clause; when the function is referenced for the first time by a Hive session, these resources will be added to the environment as if ADD JAR/FILE had been issued. If Hive is not in local mode, then the resource location must be a non-local URI such as an HDFS location.
The function will be added to the database specified, or to the current database at the time that the function was created. The function can be referenced by fully qualifying the function name (db_name.function_name), or can be referenced without qualification if the function is in the current database.
As of Hive 0.13.0 (HIVE-6047).
DROP FUNCTION [IF EXISTS] function_name;
DROP returns an error if the function doesn't exist, unless IF EXISTS is specified or the configuration variable hive.exec.drop.ignorenonexistent is set to true.
As of Hive 1.2.0 (HIVE-2573).
As of HIVE-2573, creating permanent functions in one Hive CLI session may not be reflected in HiveServer2 or other Hive CLI sessions, if they were started before the function was created. Issuing RELOAD FUNCTION within a HiveServer2 or HiveCLI session will allow it to pick up any changes to the permanent functions that may have been done by a different HiveCLI session.
Hive deprecated authorization mode / Legacy Mode has information about these DDL statements:
For SQL standard based authorization in Hive 0.13.0 and later releases, see these DDL statements:
These statements provide a way to query the Hive metastore for existing data and metadata accessible to this Hive system.
SHOW DATABASES or SHOW SCHEMAS lists all of the databases defined in the metastore. The uses of SCHEMAS and DATABASES are interchangeable – they mean the same thing.
The optional LIKE clause allows the list of databases to be filtered using a regular expression. Wildcards in the regular expression can only be '*' for any character(s) or '|' for a choice. Examples are 'employees', 'emp*', 'emp*|*ees', all of which will match the database named 'employees'.
SHOW PARTITIONS lists all the existing partitions for a given base table. Partitions are listed in alphabetical order.
It is also possible to specify parts of a partition specification to filter the resulting list.
Show Table/Partition Extended
SHOW TABLE EXTENDED will list information for all tables matching the given regular expression. Users cannot use regular expression for table name if a partition specification is present. This command's output includes basic table information and file system information like totalNumberFiles, totalFileSize, maxFileSize, minFileSize,lastAccessTime, and lastUpdateTime. If partition is present, it will output the given partition's file system information instead of table's file system information.
Show Table Properties
The first form lists all of the table properties for the table in question, one per row separated by tabs. The second form of the command prints only the value for the property that's being asked for.
For more information, see the TBLPROPERTIES clause in Create Table above.
Show Create Table
SHOW CREATE TABLE shows the CREATE TABLE statement that creates a given table, or the CREATE VIEW statement that creates a given view.
SHOW INDEXES shows all of the indexes on a certain column, as well as information about them: index name, table name, names of the columns used as keys, index table name, index type, and comment. If the FORMATTED keyword is used, then column titles are printed for each column.
SHOW COLUMNS shows all the columns in a table including partition columns.
SHOW FUNCTIONS lists all the user defined and builtin functions matching the regular expression. To get all functions use ".*"
Show Granted Roles and Privileges
Hive deprecated authorization mode / Legacy Mode has information about these SHOW statements:
In Hive 0.13.0 and later releases, SQL standard based authorization has these SHOW statements:
SHOW LOCKS displays the locks on a table or partition. See Hive Concurrency Model for information about locks.
SHOW CONF returns a description of the specified configuration property.
Note that SHOW CONF does not show the current value of a configuration property. For current property settings, use the "set" command in the CLI or a HiveQL script (see Commands) or in Beeline (see Beeline Hive Commands).
SHOW TRANSACTIONS is for use by administrators when Hive transactions are being used. It returns a list of all currently open and aborted transactions in the system, including this information:
Compactions are initiated automatically, but can also be initiated manually with an ALTER TABLE COMPACT statement.
DESCRIBE DATABASE shows the name of the database, its comment (if one has been set), and its root location on the filesystem. The uses of SCHEMA and DATABASE are interchangeable – they mean the same thing. DESCRIBE SCHEMA is added in Hive 1.1.0 (HIVE-8803).
EXTENDED also shows the database properties.
There are two formats for the describe table/view/column syntax, depending on whether or not the database is specified.
If the database is not specified, the optional column information is provided after a dot:
If the database is specified, the optional column information is provided after a space:
DESCRIBE shows the list of columns including partition columns for the given table. If the EXTENDED keyword is specified then it will show all the metadata for the table in Thrift serialized form. This is generally only useful for debugging and not for general use. If the FORMATTED keyword is specified, then it will show the metadata in a tabular format.
Note: DESCRIBE EXTENDED shows the number of rows only if statistics were gathered when the data was loaded (see Newly Created Tables), and if the Hive CLI is used instead of a Thrift client or Beeline. HIVE-6285 will address this issue. Although ANALYZE TABLE gathers statistics after the data has been loaded (see Existing Tables), it does not currently provide information about the number of rows.
If a table has a complex column then you can examine the attributes of this column by specifying table_name.complex_col_name (and field_name for an element of a struct, '$elem$' for array element, '$key$' for map key, and '$value$' for map value). You can specify this recursively to explore the complex column type.
For a view, DESCRIBE EXTENDED or FORMATTED can be used to retrieve the view's definition. Two relevant attributes are provided: both the original view definition as specified by the user, and an expanded definition used internally by Hive.
Display Column Statistics
ANALYZE TABLE table_name COMPUTE STATISTICS FOR COLUMNS will compute column statistics for all columns in the specified table (and for all partitions if the table is partitioned). To view the gathered column statistics, the following statements can be used:
See Statistics in Hive: Existing Tables for more information about the ANALYZE TABLE command.
There are two formats for the describe partition syntax, depending on whether or not the database is specified.
If the database is not specified, the optional column information is provided after a dot:
If the database is specified, the optional column information is provided after a space:
This statement lists metadata for a given partition. The output is similar to that of DESCRIBE table_name. Presently, the column information associated with a particular partition is not used while preparing plans. As of Hive 1.2 (HIVE-10307), the partition column values specified in partition_spec are type validated, converted and normalized to their column types when hive.typecheck.on.insert is set to true (default). These values can be number literals.
Hive 2.0+: Syntax Change
Warning: The new syntax could break current scripts.
ABORT TRANSACTIONS cleans up the specified transaction IDs from the Hive metastore so that users do not need to interact with the metastore directly in order to remove dangling or failed transactions. ABORT TRANSACTIONS is added in Hive 1.3.0 and 2.1.0 (HIVE-12634).
This command can be used together with SHOW TRANSACTIONS. The latter can help figure out the candidate transaction IDs to be cleaned up.
For information about DDL in HCatalog and WebHCat, see: