Information Schema

The Information Schema is a special schema that contains virtual tables which are read-only and can be queried to get information about the state of the cluster.

Note

currently only the tables tables and columns are implemented.

Tables

The information schema contains a table called tables.

This table can be queried to get a list of all available tables and their settings like the number of shards or number of replicas:

cr> select * from information_schema.tables
... where table_name not like 'my_table%' order by schema_name asc, table_name asc
+--------------------+-------------------+------------------+--------------------+--------------+
| schema_name        | table_name        | number_of_shards | number_of_replicas | clustered_by |
+--------------------+-------------------+------------------+--------------------+--------------+
| doc                | documents         | 5                | 1                  | _id          |
| doc                | locations         | 2                | 0                  | id           |
| doc                | myblobs           | 3                | 1                  | _id          |
| doc                | quotes            | 2                | 0                  | id           |
| information_schema | columns           | 1                | 0                  | NULL         |
| information_schema | table_constraints | 1                | 0                  | NULL         |
| information_schema | tables            | 1                | 0                  | NULL         |
| sys                | cluster           | 1                | 0                  | NULL         |
| sys                | nodes             | 1                | 0                  | NULL         |
| sys                | shards            | 1                | 0                  | NULL         |
+--------------------+-------------------+------------------+--------------------+--------------+
SELECT 10 rows in set (... sec)

Columns

This table can be queried to get a list of all available columns of all tables and their definition like data type and ordinal position inside the table:

cr> select * from information_schema.columns
... where schema_name='doc' and table_name not like 'my_table%'
... order by table_name asc, column_name asc
+-------------+------------+------------------+------------------+-----------+
| schema_name | table_name | column_name      | ordinal_position | data_type |
+-------------+------------+------------------+------------------+-----------+
| doc         | documents  | body             | 1                | string    |
| doc         | documents  | title            | 2                | string    |
| doc         | locations  | date             | 1                | timestamp |
| doc         | locations  | description      | 2                | string    |
| doc         | locations  | id               | 3                | string    |
| doc         | locations  | kind             | 4                | string    |
| doc         | locations  | name             | 5                | string    |
| doc         | locations  | position         | 6                | integer   |
| doc         | locations  | race             | 7                | object    |
| doc         | locations  | race.description | 8                | string    |
| doc         | locations  | race.interests   | 9                | string    |
| doc         | locations  | race.name        | 10               | string    |
| doc         | quotes     | id               | 1                | integer   |
| doc         | quotes     | quote            | 2                | string    |
+-------------+------------+------------------+------------------+-----------+
SELECT 14 rows in set (... sec)

You can even query this tables’ own columns (attention: this might lead to infinite recursion of your mind, beware!):

cr> select column_name, data_type, ordinal_position from information_schema.columns
... where schema_name = 'information_schema' and table_name = 'columns' order by ordinal_position asc
+------------------+-----------+------------------+
| column_name      | data_type | ordinal_position |
+------------------+-----------+------------------+
| schema_name      | string    | 1                |
| table_name       | string    | 2                |
| column_name      | string    | 3                |
| ordinal_position | short     | 4                |
| data_type        | string    | 5                |
+------------------+-----------+------------------+
SELECT 5 rows in set (... sec)

Note

Columns at Crate are always sorted alphabetically in ascending order despite in which order they were defined on table creation. Thus the ordinal_position reflects the alphabetical position.

Table Constraints

This table can be queries to get a list of all defined table constraints, their type, name and which table they are defined in.

Note

Currently only PRIMARY_KEY constraints are supported.

cr> select * from information_schema.table_constraints
... where table_name not like 'my_table%'
... order by schema_name desc, table_name desc limit 10
+--------------------+------------+-------------------------------------------------+-----------------+
| schema_name        | table_name | constraint_name                                 | constraint_type |
+--------------------+------------+-------------------------------------------------+-----------------+
| sys                | shards     | [u'table_name', u'id']                          | PRIMARY_KEY     |
| sys                | nodes      | [u'id']                                         | PRIMARY_KEY     |
| information_schema | tables     | [u'schema_name', u'table_name']                 | PRIMARY_KEY     |
| information_schema | columns    | [u'schema_name', u'table_name', u'column_name'] | PRIMARY_KEY     |
| doc                | quotes     | [u'id']                                         | PRIMARY_KEY     |
| doc                | myblobs    | [u'_id']                                        | PRIMARY_KEY     |
| doc                | locations  | [u'id']                                         | PRIMARY_KEY     |
| doc                | documents  | [u'_id']                                        | PRIMARY_KEY     |
+--------------------+------------+-------------------------------------------------+-----------------+
SELECT 8 rows in set (... sec)

Indices

Note

currently not implemented

This table can be queried to get a list of all defined indices of all columns and their definition like index method, expression list and property list. Using a plain index for every column is the default behaviour at Crate, so almost all columns are listed as an index as well:

cr> select * from information_schema.indices
... where table_name not like 'my_table%' order by table_name asc, index_name asc  #doctest: +SKIP
+------------+---------------------+----------+---------------------------+------------------+
| table_name | index_name          | method   | columns                   | properties       |
+------------+---------------------+----------+---------------------------+------------------+
| documents  | body                | plain    | [u'body']                 |                  |
| documents  | title               | plain    | [u'title']                |                  |
| documents  | title_body_ft       | fulltext | [u'body', u'title']       | analyzer=english |
| locations  | date                | plain    | [u'date']                 |                  |
| locations  | description         | plain    | [u'description']          |                  |
| locations  | id                  | plain    | [u'id']                   |                  |
| locations  | kind                | plain    | [u'kind']                 |                  |
| locations  | name                | plain    | [u'name']                 |                  |
| locations  | name_description_ft | fulltext | [u'description', u'name'] | analyzer=english |
| locations  | position            | plain    | [u'position']             |                  |
| locations  | race                | plain    | [u'race']                 |                  |
| quotes     | id                  | plain    | [u'id']                   |                  |
| quotes     | quote               | plain    | [u'quote']                |                  |
+------------+---------------------+----------+---------------------------+------------------+
SELECT 13 rows in set (... sec)

Routines

Note

currently not implemented

The routines table contains all custom analyzers, tokenizers, token-filters and char-filters and all custom analyzers created by CREATE ANALYZER statements (see Create custom analyzer).

The column routine_definition contains the string BUILTIN for all builtin Routines. For custom analyzers it contains the SQL statements used for their creation.

cr> select * from information_schema.routines where routine_definition != 'BUILTIN' order by routine_name asc  #doctest: +SKIP
+-----------------+--------------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| routine_name    | routine_type | routine_definition                                                                                                                                                                                       |
+-----------------+--------------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| e2              | ANALYZER     | CREATE ANALYZER e2 EXTENDS myanalyzer WITH (TOKENIZER mypattern WITH ("pattern"='.*',"type"='pattern'))                                                                                                  |
| german_snowball | ANALYZER     | CREATE ANALYZER german_snowball EXTENDS snowball WITH ("language"='german')                                                                                                                              |
| myanalyzer      | ANALYZER     | CREATE ANALYZER myanalyzer WITH (TOKENIZER whitespace, TOKEN_FILTERS WITH (lowercase, kstem), CHAR_FILTERS WITH (html_strip, mymapping WITH ("mappings"=['ph=>f','qu=>q','foo=>bar'],"type"='mapping'))) |
+-----------------+--------------+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
SELECT 3 rows in set (... sec)

You can use this table to see e.g. what builtin tokenizers exist:

cr> select routine_name from information_schema.routines
... where routine_type='TOKENIZER' and  routine_definition='BUILTIN'
... order by routine_name asc  #doctest: +SKIP
+----------------+
| routine_name   |
+----------------+
| classic        |
| edgeNGram      |
| edge_ngram     |
| keyword        |
| letter         |
| lowercase      |
| nGram          |
| ngram          |
| path_hierarchy |
| pattern        |
| standard       |
| uax_url_email  |
| whitespace     |
+----------------+
SELECT 13 rows in set (... sec)

cr> select count(*), routine_type from information_schema.routines
... group by routine_type order by routine_type  #doctest: +SKIP
+----------+--------------+
| COUNT(*) | routine_type |
+----------+--------------+
| 45       | ANALYZER     |
| 4        | CHAR_FILTER  |
| 13       | TOKENIZER    |
| 44       | TOKEN_FILTER |
+----------+--------------+
SELECT 4 rows in set (... sec)

cr> select count(distinct routine_type) as distinct_routines from information_schema.routines  #doctest: +SKIP
+-------------------+
| distinct_routines |
+-------------------+
| 4                 |
+-------------------+
SELECT 1 row in set (... sec)