Apache Cassandra Query Language (CQL)

SPECIFICATION 3.1.0

REFERENCE GUIDE - P.1

ALTER KEYSPACE

ALTER ( KEYSPACE | SCHEMA TM

) keyspace_name
WITH REPLICATION = map
| ( WITH DURABLE_WRITES = ( true | false ))
AND ( DURABLE_WRITES = ( true | false ))

CREATE KEYSPACE

CREATE ( KEYSPACE | SCHEMA ) IF NOT EXISTS keyspace_name

WITH REPLICATION = map
AND DURABLE_WRITES = ( true | false )
map is described in ALTER KEYSPACE.

map is a map collection, a JSON-style array of literals:

{ literal : literal, literal : literal ... }
CREATE TABLE
ALTER TABLE

ALTER TABLE keyspace_name.table_name instruction

instruction is:
ALTER column_name TYPE cql_type
| ( ADD column_name cql_type )
| ( DROP column_name )
| ( RENAME column_name TO column_name )
| ( WITH property AND property ... )

column_definition is:
column_name cql_type PRIMARY KEY
| ( PRIMARY KEY ( partition_key ) )
Restrictions:
There should always be exactly one primary key denition.
cql_type of the primary key must be a custom type or a CQL type (see last page),
other than a collection type.
cql_type of a collection uses this syntax:

cql_type is a CQL type, other than a collection orcounter.

Exceptions: ADD supports a collection type and also, if the table
is a counter, a counter type.

LIST<cql_type>
| SET<cql_type>
| MAP<cql_type, cql_type>

property is a CQL 3 table storage property (p. 3) and value,

such as caching = all.
ALTER USER

ALTER USER user_name WITH PASSWORD 'password'

( NOSUPERUSER | SUPERUSER )

BATCH

BEGIN BATCH ( UNLOGGED | COUNTER )

USING TIMESTAMP timestamp
dml_statement;
dml_statement;
...
APPLY BATCH;

PRIMARY KEY is:

column_name
| ( column_name1, column_name2, column_name3 ...)
| ((column_name4, column_name5), column_name6,
column_name7 ...)
column_name1 is the partition key.
column_name2, column_name3 ... are clustering keys.
column_name4, column_name5 are partitioning keys.
column_name6, column_name7 ... are clustering keys.

dml_statement is:
INSERT
UPDATE
DELETE
CREATE INDEX

CREATE TABLE IF NOT EXISTS keyspace_name.table_name

( column_definition, column_definition, ...)
WITH property AND property ...

property is a CQL table storage property or one of these directives:

COMPACT STORAGE
| ( CLUSTERING ORDER BY (clustering_key ( ASC | DESC ), ...)

CREATE CUSTOM INDEX IF NOT EXISTS index_name

ON keyspace_name.table_name (column_name)
USING class_name

CREATE TRIGGER

CREATE TRIGGER trigger_name ON table_name

USING java_class

Restrictions:
USING class_name is only allowed if CUSTOM is used and
class_name is a string literal containing a java class name.

CREATE USER

CREATE USER user_name WITH PASSWORD 'password'

NOSUPERUSER | SUPERUSER

index_name is an identier, enclosed or not enclosed in double

quotation marks, excluding reserved words.
LEGEND:

* Uppercase means literal * Lowercase means not literal * Italics mean optional * The pipe (|) symbol means OR or AND/OR * Ellipsis (...) means repeatable * orange ( ) indicate scope, not literal

Apache Cassandra Query Language (CQL)

SPECIFICATION 3.1.0

REFERENCE GUIDE - P.2

DELETE

DELETE column_name, ... | ( column_name term )

FROM keyspace_name.table_nameTM
USING TIMESTAMP integer
WHERE row_specification
term is:
[ list_position ] | key_value

value is one of:

a literal
a set
{ literal, literal, ... }
a list
[ literal, literal, ... ]
a map, described in ALTER KEYSPACE

row_specication is one of:

primary_key_name = key_value
primary_key_name IN ( key_value, key_value, ...)

option is one of:

TIMESTAMP microseconds
TTL seconds

INSERT (continued)

LIST PERMISSIONS
DROP INDEX

DROP INDEX IF EXISTS index_name

DROP KEYSPACE

DROP ( KEYSPACE | SCHEMA ) IF EXISTS keyspace_name

LIST permission_name PERMISSION

| ( LIST ALL PERMISSIONS )
ON resource OF user_name
NORECURSIVE
permission_name and resource are shown in GRANT.

DROP TABLE

DROP TABLE IF EXISTS keyspace_name.table_name

DROP TRIGGER

DROP TRIGGER trigger_name ON table_name

DROP USER

DROP USER user_name

LIST USERS

LIST USERS

REVOKE

REVOKE ( permission_name PERMISSION )

| ( ALL PERMISSIONS )
ON resource FROM user_name
permission_name and resource are shown in GRANT.

GRANT

GRANT permission_name PERMISSION

| ( GRANT ALL PERMISSIONS ) ON resource TO user_name

SELECT

permission_name is one of:

ALL
ALTER
AUTHORIZE
CREATE
DROP
MODIFY
SELECT

select_expression is:
selection_list | ( COUNT ( * | 1 ) )
selection_list is:
selector AS alias, selector AS alias, ... | *

resource is one of:

ALL KEYSPACES
KEYSPACE keyspace_name
TABLEkeyspace_name.table_name

alias is an alias for a column.

selector is:
column_name
| ( WRITETIME (column_name) )
| TTL
| ( TTL (column_name) )
| ( function (selector, selector, ...) )

INSERT INTO keyspace_name.table_name

( column_name, column_name ... )
VALUES ( value, value, ... ) IF NOT EXISTS
USING option AND option

INSERT

(continued)

(continued)

LEGEND:

SELECT select_expression
FROM keyspace_name.table_name
WHERE relation AND relation ...
ORDER BY (clustering_key ( ASC | DESC ), ... )
LIMIT n
ALLOW FILTERING

* Uppercase means literal * Lowercase means not literal * Italics mean optional * The pipe (|) symbol means OR or AND/OR * Ellipsis (...) means repeatable * orange ( ) indicate scope, not literal

Apache Cassandra Query Language (CQL)

SPECIFICATION 3.1.0

REFERENCE GUIDE - P.3

SELECT (continued)

function is a timeuuid function, a

TMtoken function, or a blob
conversion function.

CQL Table Storage Properties

relation is:
column_name ( = | < | > | <= | >= ) key_value
| column_name IN ( ( key_value, ... ) )
| TOKEN ( column_name, ... ) ( = | < | > | <= | >= )
( term | TOKEN ( term, ... ) )

bloom_filter_fp_chance
Desired false-positive probability for SSTable Bloom lters.
caching
Cache memory settings. Values: all, keys_only, rows_only, none.
comment
A human readable comment describing the table.

term is a constant, such as a true or false, a bind marker (?),

or a set, list, or map.
TRUNCATE

TRUNCATE keyspace_name.table_name

UPDATE

UPDATE keyspace_name.table_name
USING option AND option
SET assignment, assignment ...
WHERE row_specification
IF column_name = literal
AND column_name = literal ...

compaction
Options for SSTable compaction:
bucket_high
bucket_low
max_threshold
min_threshold
min_sstable_size
sstable_size_in_mb
tombstone_compaction_interval
tombstone_threshold

option is one of:

TIMESTAMP microseconds
TTL seconds

compression
The compression algorithm and subproperties for the table:
sstable_compression
chunk_length_kb
crc_check_chance

assignment is one of:

column_name = value
set_or_list_item = set_or_list_item (+ | -) ...
map_name = map_name (+ | -) ...
column_name [ term ] = value
counter_column_name = counter_column_name
( + | - ) integer

default_time_to_live
The default expiration time in seconds for a table. Used in MapReduce/Hive
scenarios in which you have no control of TTL.
dclocal_read_repair_chance
The probability of read repairs being invoked over all replicas in the current data
center.

set, list, map are dened in INSERT.

gc_grace_seconds
The time to wait before garbage collecting tombstones (deletion markers).

term is:
[ list_position ] | key_value

index_interval
The number of keys that are sampled to create the partition summary, a subset
of the partition index.

row_specification is one of:

primary_key_name = key_value
primary_key_name IN (key_value ,...)
USE

memtable_flush_period_in_ms
Forces flushing of the memtable after the number of specied milliseconds
elapses.

USE keyspace_name

populate_io_cache_on_flush
Populates the page cache on memtable flush and compaction. Enable only
when all data on the node ts in memory. Defaults is false. Used to t only one
of several tables inside single keyspace.

LEGEND:

* Uppercase means literal * Lowercase means not literal * Italics mean optional * The pipe (|) symbol means OR or AND/OR * Ellipsis (...) means repeatable * orange ( ) indicate scope, not literal

Apache Cassandra Query Language (CQL)

SPECIFICATION 3.1.0

REFERENCE GUIDE - P.4

CQL Table Storage Properties (continued)

Syntax elements

TM

read_repair_chance
Species the probability for invoking read repairs on non-quorum reads.

Generally, the elements used in the command syntax have the following
denitions. A few elements have a slightly different meaning when used with a
particular command and are redened in the synopsis of the command.

replicate_on_write
Writes counter table data to all affected replicas regardless of the consistency
level specied by the client for a write request.

clustering_key
Columns in addition to the partitioning key component of a primary key that
determine clustering.

speculative_retry
Overrides normal read timeout when read_repair is not 1.0, sending another
request to read. Options:
ALL Retry reads of all replicas.
Xpercentile Retry reads based on the effect on throughput and latency.
Yms Retry reads after specified milliseconds
NONE Do not retry reads.

column_name
Alphanumeric column name, case-insensitive unless enclosed in double
quotation marks. No reserved keywords. Unreserved keywords enclosed in
quotation marks are ok. Enclose names having unparseable characters in
double quotation marks.
constant
A string, integer, float, boolean, UUID, or blob.
counter_column_name
A column_name of a column of type counter.

Functions

keyspace_name
A keyspace name, starting with an alpha character, consisting of 32 or fewer
alpha-numeric characters and underscores. Case-insensitive unless enclosed
in double quotation marks.

Blob conversion
Converts native types into binary data (blob).
typeAsBlob() takes a native type and returns it as a blob
bigintAsBlob(3) returns 0x0000000000000003
blobAsType takes a 64-bit blob argument and converts it to a bigint value
blobAsBigint(0x0000000000000003) returns 3

key_value
The value of a primary key.
literal
Data that is of a supported data type
Float constant in E notation
Numeric constant
A letter, followed by any sequence of letters, digits, or the underscore
A string, characters enclosed in single quotation marks
Whitespace that separates of terms, otherwise ignored

dateOf()
Used in a SELECT clause to extract the timestamp of a timeuuid column in a
resultset. Returns the extracted timestamp as a date.
minTimeuuid() and maxTimeuuid()
Returns a UUID-like result given a conditional time component as an argument.
Example:
SELECT * FROM myTable
WHERE t > maxTimeuuid('2013-01-01 00:05+0000')
AND t < minTimeuuid('2013-02-02 10:00+0000')

partitioning_key
The primary key column that determines on which node data is stored.
property
A CQL storage property, such as caching = "all".

now()
Generates a new unique timeuuid, useful for inserting values. Returns a unique
value.

table_name
Valid table names are strings of alphanumeric characters and underscores,
which begin with a letter.

unixTimestampOf()
Used in a SELECT clause to extract the timestamp of a timeuuid column in a
resultset. Returns a raw, 64-bit integer timestamp.

timestamp
Microseconds representing the standard base time since epoch: January 1
1970 at 00:00:00 GMT.
variable
A bind variable, such as ?, used with a prepared statement.

LEGEND:

* Uppercase means literal * Lowercase means not literal * Italics mean optional * The pipe (|) symbol means OR or AND/OR * Ellipsis (...) means repeatable * orange ( ) indicate scope, not literal

Apache Cassandra Query Language (CQL)

SPECIFICATION 3.1.0

REFERENCE GUIDE - SUPPORTED DATA TYPES

CQL Type

Description

ASCII

US-ASCII character string

BIGINT

64-bit signed long

BLOB

Arbitrary bytes (no validation), expressed as hexadecimal

BOOLEAN

true or false

COUNTER

Distributed counter value (64-bit long)

DECIMAL

Variable-precision decimal

DOUBLE

64-bit IEEE-754 floating point

FLOAT

32-bit IEEE-754 floating point

INET

IP address string in IPv4 or IPv6 form

INT

32-bit signed integer

LIST

A collection of one or more ordered elements

MAP

A collection of one or more timestamp, value pairs

SET

A collection of one or more elements

TEXT

UTF-8 encoded string

TIMESTAMP

Date plus time, encoded as 8 bytes since epoch

UUID

A UUID in standard UUID format

TIMEUUID

Type 1 UUID only

VARCHAR

UTF-8 encoded string

VARINT

Arbitrary-precision integer

Office Locations
DATASTAX HQ - SF BAY AREA
777 Mariners Island Blvd #510
San Mateo, CA 94404
650-389-6000

LEGEND:

DATASTAX TX
902 East 5th St. #202
Austin, TX 78702
512-537-7809

* Uppercase means literal * Lowercase means not literal * Italics mean optional * The pipe (|) symbol means OR or AND/OR * Ellipsis (...) means repeatable * orange ( ) indicate scope, not literal