SQL Reference

In this part, we will introduce you IoTDB’s Query Language. IoTDB offers you a SQL-like query language for interacting with IoTDB, the query language can be devided into 4 major parts:

Schema Statement: statements about schema management are all listed in this section.
Data Management Statement: statements about data management (such as: data insertion, data query, etc.) are all listed in this section.
Database Management Statement: statements about database management and authentication are all listed in this section.
Functions: functions that IoTDB offers are all listed in this section.

All of these statements are write in IoTDB’s own syntax, for details about the syntax composition, please check the Reference section.

Show Version

show version

+---------------+
|        version|
+---------------+
|0.11.1-SNAPSHOT|
+---------------+
Total line number = 1
It costs 0.417s

Schema Statement

Set Storage Group

SET STORAGE GROUP TO <FullPath>
Eg: IoTDB > SET STORAGE GROUP TO root.ln.wf01.wt01
Note: FullPath can not include `*`

Delete Storage Group

DELETE STORAGE GROUP <FullPath> [COMMA <FullPath>]*
Eg: IoTDB > DELETE STORAGE GROUP root.ln.wf01.wt01
Eg: IoTDB > DELETE STORAGE GROUP root.ln.wf01.wt01, root.ln.wf01.wt02
Eg: IoTDB > DELETE STORAGE GROUP root.ln.wf01.*
Eg: IoTDB > DELETE STORAGE GROUP root.*

Create Timeseries Statement

CREATE TIMESERIES <FullPath> WITH <AttributeClauses>
alias
    : LR_BRACKET ID RR_BRACKET
    ;
attributeClauses
    : DATATYPE OPERATOR_EQ dataType COMMA ENCODING OPERATOR_EQ encoding
    (COMMA (COMPRESSOR | COMPRESSION) OPERATOR_EQ compressor=propertyValue)?
    (COMMA property)*
    tagClause
    attributeClause
    ;
attributeClause
    : (ATTRIBUTES LR_BRACKET property (COMMA property)* RR_BRACKET)?
    ;
tagClause
    : (TAGS LR_BRACKET property (COMMA property)* RR_BRACKET)?
    ;
DataTypeValue: BOOLEAN | DOUBLE | FLOAT | INT32 | INT64 | TEXT
EncodingValue: GORILLA | PLAIN | RLE | TS_2DIFF | REGULAR
CompressorValue: UNCOMPRESSED | SNAPPY
Eg: CREATE TIMESERIES root.ln.wf01.wt01.status WITH DATATYPE=BOOLEAN, ENCODING=PLAIN
Eg: CREATE TIMESERIES root.ln.wf01.wt01.temperature WITH DATATYPE=FLOAT, ENCODING=RLE
Eg: CREATE TIMESERIES root.ln.wf01.wt01.temperature WITH DATATYPE=FLOAT, ENCODING=RLE, COMPRESSOR=SNAPPY, MAX_POINT_NUMBER=3
Eg: create timeseries root.turbine.d0.s0(temperature) with datatype=FLOAT, encoding=RLE, compression=SNAPPY tags(unit=f, description='turbine this is a test1') attributes(H_Alarm=100, M_Alarm=50)
Note: Datatype and encoding type must be corresponding. Please check Chapter 3 Encoding Section for details.

Delete Timeseries Statement

DELETE TIMESERIES <PrefixPath> [COMMA <PrefixPath>]*
Eg: IoTDB > DELETE TIMESERIES root.ln.wf01.wt01.status
Eg: IoTDB > DELETE TIMESERIES root.ln.wf01.wt01.status, root.ln.wf01.wt01.temperature
Eg: IoTDB > DELETE TIMESERIES root.ln.wf01.wt01.*

Alter Timeseries Statement

ALTER TIMESERIES fullPath alterClause
alterClause
    : RENAME beforeName=ID TO currentName=ID
    | SET property (COMMA property)*
    | DROP ID (COMMA ID)*
    | ADD TAGS property (COMMA property)*
    | ADD ATTRIBUTES property (COMMA property)*
    | UPSERT tagClause attributeClause
    ;
attributeClause
    : (ATTRIBUTES LR_BRACKET property (COMMA property)* RR_BRACKET)?
    ;
tagClause
    : (TAGS LR_BRACKET property (COMMA property)* RR_BRACKET)?
    ;
Eg: ALTER timeseries root.turbine.d1.s1 RENAME tag1 TO newTag1
Eg: ALTER timeseries root.turbine.d1.s1 SET tag1=newV1, attr1=newV1
Eg: ALTER timeseries root.turbine.d1.s1 DROP tag1, tag2
Eg: ALTER timeseries root.turbine.d1.s1 ADD TAGS tag3=v3, tag4=v4
Eg: ALTER timeseries root.turbine.d1.s1 ADD ATTRIBUTES attr3=v3, attr4=v4
EG: ALTER timeseries root.turbine.d1.s1 UPSERT TAGS(tag2=newV2, tag3=v3) ATTRIBUTES(attr3=v3, attr4=v4)

Show All Timeseries Statement

SHOW TIMESERIES
Eg: IoTDB > SHOW TIMESERIES
Note: This statement can only be used in IoTDB Client. If you need to show all timeseries in JDBC, please use `DataBaseMetadata` interface.

Show Specific Timeseries Statement

SHOW TIMESERIES <Path>
Eg: IoTDB > SHOW TIMESERIES root
Eg: IoTDB > SHOW TIMESERIES root.ln
Eg: IoTDB > SHOW TIMESERIES root.ln.*.*.status
Eg: IoTDB > SHOW TIMESERIES root.ln.wf01.wt01.status
Note: The path can be prefix path, star path or timeseries path
Note: This statement can be used in IoTDB Client and JDBC.

Show Specific Timeseries Statement with where clause

SHOW TIMESERIES prefixPath? showWhereClause?
showWhereClause
    : WHERE (property | containsExpression)
    ;
containsExpression
    : name=ID OPERATOR_CONTAINS value=propertyValue
    ;
Eg: show timeseries root.ln where unit='c'
Eg: show timeseries root.ln where description contains 'test1'

Show Specific Timeseries Statement with where clause start from offset and limit the total number of result

SHOW TIMESERIES prefixPath? showWhereClause? limitClause?
showWhereClause
    : WHERE (property | containsExpression)
    ;
containsExpression
    : name=ID OPERATOR_CONTAINS value=propertyValue
    ;
limitClause
    : LIMIT INT offsetClause?
    | offsetClause? LIMIT INT
    ;
Eg: show timeseries root.ln where unit='c'
Eg: show timeseries root.ln where description contains 'test1'
Eg: show timeseries root.ln where unit='c' limit 10 offset 10

Show Storage Group Statement

SHOW STORAGE GROUP
Eg: IoTDB > SHOW STORAGE GROUP
Note: This statement can be used in IoTDB Client and JDBC.

Show Specific Storage Group Statement

SHOW STORAGE GROUP <PrefixPath>
Eg: IoTDB > SHOW STORAGE GROUP root.*
Eg: IoTDB > SHOW STORAGE GROUP root.ln
Note: The path can be prefix path or star path.
Note: This statement can be used in IoTDB Client and JDBC.

Show Merge Status Statement

SHOW MERGE
Eg: IoTDB > SHOW MERGE
Note: This statement can be used in IoTDB Client and JDBC.

Count Timeseries Statement

COUNT TIMESERIES <Path>
Eg: IoTDB > COUNT TIMESERIES root
Eg: IoTDB > COUNT TIMESERIES root.ln
Eg: IoTDB > COUNT TIMESERIES root.ln.*.*.status
Eg: IoTDB > COUNT TIMESERIES root.ln.wf01.wt01.status
Note: The path can be prefix path, star path or timeseries path.
Note: This statement can be used in IoTDB Client and JDBC.

COUNT TIMESERIES <Path> GROUP BY LEVEL=<INTEGER>
Eg: IoTDB > COUNT TIMESERIES root GROUP BY LEVEL=1
Eg: IoTDB > COUNT TIMESERIES root.ln GROUP BY LEVEL=2
Eg: IoTDB > COUNT TIMESERIES root.ln.wf01 GROUP BY LEVEL=3
Note: The path can be prefix path or timeseries path.
Note: This statement can be used in IoTDB Client and JDBC.

Count Nodes Statement

COUNT NODES <Path> LEVEL=<INTEGER>
Eg: IoTDB > COUNT NODES root LEVEL=2
Eg: IoTDB > COUNT NODES root.ln LEVEL=2
Eg: IoTDB > COUNT NODES root.ln.* LEVEL=3
Eg: IoTDB > COUNT NODES root.ln.wf01 LEVEL=3
Note: The path can be prefix path or timeseries path.
Note: This statement can be used in IoTDB Client and JDBC.

Show All Devices Statement

SHOW Devices
Eg: IoTDB > SHOW Devices
Note: This statement can be used in IoTDB Client and JDBC.

Show Specific Devices Statement

SHOW DEVICES <PrefixPath>
Eg: IoTDB > SHOW DEVICES root
Eg: IoTDB > SHOW DEVICES root.ln
Eg: IoTDB > SHOW DEVICES root.*.wf01
Note: The path can be prefix path or star path.
Note: This statement can be used in IoTDB Client and JDBC.

Show Child Paths of Root Statement

SHOW CHILD PATHS
Eg: IoTDB > SHOW CHILD PATHS
Note: This statement can be used in IoTDB Client and JDBC.

Show Child Paths Statement

SHOW CHILD PATHS <Path>
Eg: IoTDB > SHOW CHILD PATHS root
Eg: IoTDB > SHOW CHILD PATHS root.ln
Eg: IoTDB > SHOW CHILD PATHS root.*.wf01
Eg: IoTDB > SHOW CHILD PATHS root.ln.wf*
Note: The path can be prefix path or star path, the nodes can be in a "prefix + star" format. 
Note: This statement can be used in IoTDB Client and JDBC.

Create snapshot for schema

CREATE SNAPSHOT FOR SCHEMA

Data Management Statement

Insert Record Statement

INSERT INTO <PrefixPath> LPAREN TIMESTAMP COMMA <Sensor> [COMMA <Sensor>]* RPAREN VALUES LPAREN <TimeValue>, <PointValue> [COMMA <PointValue>]* RPAREN
Sensor : Identifier
Eg: IoTDB > INSERT INTO root.ln.wf01.wt01(timestamp,status) values(1509465600000,true)
Eg: IoTDB > INSERT INTO root.ln.wf01.wt01(timestamp,status) VALUES(NOW(), false)
Eg: IoTDB > INSERT INTO root.ln.wf01.wt01(timestamp,temperature) VALUES(2017-11-01T00:17:00.000+08:00,24.22028)
Eg: IoTDB > INSERT INTO root.ln.wf01.wt01(timestamp, status, temperature) VALUES (1509466680000, false, 20.060787);
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>
Note: The order of Sensor and PointValue need one-to-one correspondence

Delete Record Statement

DELETE FROM <PrefixPath> [COMMA <PrefixPath>]* [WHERE <WhereClause>]?
WhereClause : <Condition> [(AND) <Condition>]*
Condition  : <TimeExpr> [(AND) <TimeExpr>]*
TimeExpr : TIME PrecedenceEqualOperator (<TimeValue> | <RelativeTime>)
Eg: DELETE FROM root.ln.wf01.wt01.temperature WHERE time > 2016-01-05T00:15:00+08:00 and time < 2017-11-1T00:05:00+08:00
Eg: DELETE FROM root.ln.wf01.wt01.status, root.ln.wf01.wt01.temperature WHERE time < NOW()
Eg: DELETE FROM root.ln.wf01.wt01.* WHERE time >= 1509466140000

Select Record Statement

SELECT <SelectClause> FROM <FromClause> [WHERE <WhereClause>]?
SelectClause : <SelectPath> (COMMA <SelectPath>)*
SelectPath : <FUNCTION> LPAREN <Path> RPAREN | <Path>
FUNCTION : ‘COUNT’ , ‘MIN_TIME’, ‘MAX_TIME’, ‘MIN_VALUE’, ‘MAX_VALUE’
FromClause : <PrefixPath> (COMMA <PrefixPath>)?
WhereClause : <Condition> [(AND | OR) <Condition>]*
Condition  : <Expression> [(AND | OR) <Expression>]*
Expression : [NOT | !]? <TimeExpr> | [NOT | !]? <SensorExpr>
TimeExpr : TIME PrecedenceEqualOperator (<TimeValue> | <RelativeTime>)
RelativeTimeDurationUnit = Integer ('Y'|'MO'|'W'|'D'|'H'|'M'|'S'|'MS'|'US'|'NS')
RelativeTime : (now() | <TimeValue>) [(+|-) RelativeTimeDurationUnit]+
SensorExpr : (<Timeseries> | <Path>) PrecedenceEqualOperator <PointValue>
Eg: IoTDB > SELECT status, temperature FROM root.ln.wf01.wt01 WHERE temperature < 24 and time > 2017-11-1 0:13:00
Eg. IoTDB > SELECT * FROM root
Eg. IoTDB > SELECT * FROM root where time > now() - 5m
Eg. IoTDB > SELECT * FROM root.ln.*.wf*
Eg. IoTDB > SELECT COUNT(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 25
Eg. IoTDB > SELECT MIN_TIME(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 25
Eg. IoTDB > SELECT MAX_TIME(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature > 24
Eg. IoTDB > SELECT MIN_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature > 23
Eg. IoTDB > SELECT MAX_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 25
Eg. IoTDB > SELECT COUNT(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 25 GROUP BY LEVEL=1
Note: the statement needs to satisfy this constraint: <Path>(SelectClause) + <PrefixPath>(FromClause) = <Timeseries>
Note: If the <SensorExpr>(WhereClause) is started with <Path> and not with ROOT, the statement needs to satisfy this constraint: <PrefixPath>(FromClause) + <Path>(SensorExpr) = <Timeseries>
Note: In Version 0.7.0, if <WhereClause> includes `OR`, time filter can not be used.
Note: There must be a space on both sides of the plus and minus operator appearing in the time expression

Group By Statement

SELECT <SelectClause> FROM <FromClause> WHERE  <WhereClause> GROUP BY <GroupByTimeClause>
SelectClause : <Function> [COMMA < Function >]*
Function : <AggregationFunction> LPAREN <Path> RPAREN
FromClause : <PrefixPath>
WhereClause : <Condition> [(AND | OR) <Condition>]*
Condition  : <Expression> [(AND | OR) <Expression>]*
Expression : [NOT | !]? <TimeExpr> | [NOT | !]? <SensorExpr>
TimeExpr : TIME PrecedenceEqualOperator (<TimeValue> | <RelativeTime>)
RelativeTimeDurationUnit = Integer ('Y'|'MO'|'W'|'D'|'H'|'M'|'S'|'MS'|'US'|'NS')
RelativeTime : (now() | <TimeValue>) [(+|-) RelativeTimeDurationUnit]+
SensorExpr : (<Timeseries> | <Path>) PrecedenceEqualOperator <PointValue>
GroupByTimeClause : LPAREN <TimeInterval> COMMA <TimeUnit> (COMMA <TimeUnit>)? RPAREN
TimeInterval: LSBRACKET <TimeValue> COMMA <TimeValue> RRBRACKET | LRBRACKET <TimeValue> COMMA <TimeValue> RSBRACKET
TimeUnit : Integer <DurationUnit>
DurationUnit : "ms" | "s" | "m" | "h" | "d" | "w"
Eg: SELECT COUNT(status), COUNT(temperature) FROM root.ln.wf01.wt01 where temperature < 24 GROUP BY([1509465720000, 1509466380000), 5m)
Eg: SELECT COUNT(status), COUNT(temperature) FROM root.ln.wf01.wt01 where temperature < 24 GROUP BY((1509465720000, 1509466380000], 5m)
Eg. SELECT COUNT (status), MAX_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE time < 1509466500000 GROUP BY([1509465720000, 1509466380000), 5m, 10m)
Eg. SELECT MIN_TIME(status), MIN_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE temperature < 25 GROUP BY ([1509466140000, 1509466380000), 3m, 5ms)
Eg. SELECT MIN_TIME(status), MIN_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE temperature < 25 GROUP BY ((1509466140000, 1509466380000], 3m, 5ms)
Note: the statement needs to satisfy this constraint: <Path>(SelectClause) + <PrefixPath>(FromClause) = <Timeseries>
Note: If the <SensorExpr>(WhereClause) is started with <Path> and not with ROOT, the statement needs to satisfy this constraint: <PrefixPath>(FromClause) + <Path>(SensorExpr) = <Timeseries>
Note: <TimeValue>(TimeInterval) needs to be greater than 0
Note: First <TimeValue>(TimeInterval) in needs to be smaller than second <TimeValue>(TimeInterval)
Note: <TimeUnit> needs to be greater than 0
Note: Third <TimeUnit> if set shouldn't be smaller than second <TimeUnit>

Fill Statement

SELECT <SelectClause> FROM <FromClause> WHERE <WhereClause> FILL <FillClause>
SelectClause : <Path> [COMMA <Path>]*
FromClause : < PrefixPath > [COMMA < PrefixPath >]*
WhereClause : <WhereExpression>
WhereExpression : TIME EQUAL <TimeValue>
FillClause : LPAREN <TypeClause> [COMMA <TypeClause>]* RPAREN
TypeClause : <Int32Clause> | <Int64Clause> | <FloatClause> | <DoubleClause> | <BoolClause> | <TextClause>
Int32Clause: INT32 LBRACKET (<LinearClause> | <PreviousClause>)  RBRACKET
Int64Clause: INT64 LBRACKET (<LinearClause> | <PreviousClause>)  RBRACKET
FloatClause: FLOAT LBRACKET (<LinearClause> | <PreviousClause>)  RBRACKET
DoubleClause: DOUBLE LBRACKET (<LinearClause> | <PreviousClause>)  RBRACKET
BoolClause: BOOLEAN LBRACKET (<LinearClause> | <PreviousClause>)  RBRACKET
TextClause: TEXT LBRACKET (<LinearClause> | <PreviousClause>)  RBRACKET
PreviousClause : PREVIOUS [COMMA <ValidPreviousTime>]?
LinearClause : LINEAR [COMMA <ValidPreviousTime> COMMA <ValidBehindTime>]?
ValidPreviousTime, ValidBehindTime: <TimeUnit>
TimeUnit : Integer <DurationUnit>
DurationUnit : "ms" | "s" | "m" | "h" | "d" | "w"
Eg: SELECT temperature FROM root.ln.wf01.wt01 WHERE time = 2017-11-01T16:37:50.000 FILL(float[previous, 1m])
Eg: SELECT temperature,status FROM root.ln.wf01.wt01 WHERE time = 2017-11-01T16:37:50.000 FILL (float[linear, 1m, 1m], boolean[previous, 1m])
Eg: SELECT temperature,status,hardware FROM root.ln.wf01.wt01 WHERE time = 2017-11-01T16:37:50.000 FILL (float[linear, 1m, 1m], boolean[previous, 1m], text[previous])
Eg: SELECT temperature,status,hardware FROM root.ln.wf01.wt01 WHERE time = 2017-11-01T16:37:50.000 FILL (float[linear], boolean[previous, 1m], text[previous])
Note: the statement needs to satisfy this constraint: <PrefixPath>(FromClause) + <Path>(SelectClause) = <Timeseries>
Note: Integer in <TimeUnit> needs to be greater than 0

Group By Fill Statement

SELECT <SelectClause> FROM <FromClause> WHERE  <WhereClause> GROUP BY <GroupByClause> (FILL <GROUPBYFillClause>)?
GroupByClause : LPAREN <TimeInterval> COMMA <TimeUnit> RPAREN
GROUPBYFillClause : LPAREN <TypeClause> RPAREN
TypeClause : <AllClause> | <Int32Clause> | <Int64Clause> | <FloatClause> | <DoubleClause> | <BoolClause> | <TextClause> 
AllClause: ALL LBRACKET (<PreviousUntilLastClause> | <PreviousClause>)  RBRACKET
Int32Clause: INT32 LBRACKET (<PreviousUntilLastClause> | <PreviousClause>)  RBRACKET
Int64Clause: INT64 LBRACKET (<PreviousUntilLastClause> | <PreviousClause>)  RBRACKET
FloatClause: FLOAT LBRACKET (<PreviousUntilLastClause> | <PreviousClause>)  RBRACKET
DoubleClause: DOUBLE LBRACKET (<PreviousUntilLastClause> | <PreviousClause>)  RBRACKET
BoolClause: BOOLEAN LBRACKET (<PreviousUntilLastClause> | <PreviousClause>)  RBRACKET
TextClause: TEXT LBRACKET (<PreviousUntilLastClause> | <PreviousClause>)  RBRACKET
PreviousClause : PREVIOUS
PreviousUntilLastClause : PREVIOUSUNTILLAST
Eg: SELECT last_value(temperature) FROM root.ln.wf01.wt01 GROUP BY([20, 100), 5m) FILL (float[PREVIOUS])
Eg: SELECT last_value(temperature) FROM root.ln.wf01.wt01 GROUP BY((15, 100], 5m) FILL (float[PREVIOUS])
Eg: SELECT last_value(power) FROM root.ln.wf01.wt01 GROUP BY([20, 100), 5m) FILL (int32[PREVIOUSUNTILLAST])
Eg: SELECT last_value(power) FROM root.ln.wf01.wt01 GROUP BY([20, 100), 5m) FILL (int32[PREVIOUSUNTILLAST, 5m])
Eg: SELECT last_value(temperature), last_value(power) FROM root.ln.wf01.wt01 GROUP BY([20, 100), 5m) FILL (ALL[PREVIOUS])
Eg: SELECT last_value(temperature), last_value(power) FROM root.ln.wf01.wt01 GROUP BY([20, 100), 5m) FILL (ALL[PREVIOUS, 5m])
Note: In group by fill, sliding step is not supported in group by clause
Note: Now, only last_value aggregation function is supported in group by fill.
Note: Linear fill is not supported in group by fill.

Order by time Statement

SELECT <SelectClause> FROM <FromClause> WHERE  <WhereClause> GROUP BY <GroupByClause> (FILL <GROUPBYFillClause>)? orderByTimeClause?
orderByTimeClause: order by time (asc | desc)?
Eg: SELECT last_value(temperature) FROM root.ln.wf01.wt01 GROUP BY([20, 100), 5m) FILL (float[PREVIOUS]) order by time desc
Eg: SELECT * from root order by time desc
Eg: SELECT * from root order by time desc align by device 
Eg: SELECT * from root order by time desc disable align
Eg: SELECT last * from root order by time desc

Limit Statement

SELECT <SelectClause> FROM <FromClause> [WHERE <WhereClause>] [<LIMITClause>] [<SLIMITClause>]
SelectClause : [<Path> | Function]+
Function : <AggregationFunction> LPAREN <Path> RPAREN
FromClause : <Path>
WhereClause : <Condition> [(AND | OR) <Condition>]*
Condition : <Expression> [(AND | OR) <Expression>]*
Expression: [NOT|!]?<TimeExpr> | [NOT|!]?<SensorExpr>
TimeExpr : TIME PrecedenceEqualOperator (<TimeValue> | <RelativeTime>)
RelativeTimeDurationUnit = Integer ('Y'|'MO'|'W'|'D'|'H'|'M'|'S'|'MS'|'US'|'NS')
RelativeTime : (now() | <TimeValue>) [(+|-) RelativeTimeDurationUnit]+
SensorExpr : (<Timeseries>|<Path>) PrecedenceEqualOperator <PointValue>
LIMITClause : LIMIT <N> [OFFSETClause]?
N : Integer
OFFSETClause : OFFSET <OFFSETValue>
OFFSETValue : Integer
SLIMITClause : SLIMIT <SN> [SOFFSETClause]?
SN : Integer
SOFFSETClause : SOFFSET <SOFFSETValue>
SOFFSETValue : Integer
Eg: IoTDB > SELECT status, temperature FROM root.ln.wf01.wt01 WHERE temperature < 24 and time > 2017-11-1 0:13:00 LIMIT 3 OFFSET 2
Eg. IoTDB > SELECT COUNT (status), MAX_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE time < 1509466500000 GROUP BY([1509465720000, 1509466380000], 5m) LIMIT 3
Note: N, OFFSETValue, SN and SOFFSETValue must be greater than 0.
Note: The order of <LIMITClause> and <SLIMITClause> does not affect the grammatical correctness.
Note: <FillClause> can not use <LIMITClause> but not <SLIMITClause>.

Align By Device Statement

AlignbyDeviceClause : ALIGN BY DEVICE
Rules:  
1. Both uppercase and lowercase are ok.  
Correct example: select * from root.sg1 align by device  
Correct example: select * from root.sg1 ALIGN BY DEVICE  
2. AlignbyDeviceClause can only be used at the end of a query statement.  
Correct example: select * from root.sg1 where time > 10 align by device  
Wrong example: select * from root.sg1 align by device where time > 10  
3. The paths of the SELECT clause can only be single level. In other words, the paths of the SELECT clause can only be measurements or STAR, without DOT.
Correct example: select s0,s1 from root.sg1.* align by device  
Correct example: select s0,s1 from root.sg1.d0, root.sg1.d1 align by device  
Correct example: select * from root.sg1.* align by device  
Correct example: select * from root align by device  
Correct example: select s0,s1,* from root.*.* align by device  
Wrong example: select d0.s1, d0.s2, d1.s0 from root.sg1 align by device  
Wrong example: select *.s0, *.s1 from root.* align by device  
Wrong example: select *.*.* from root align by device
4. The data types of the same measurement column should be the same across devices. 
Note that when it comes to aggregated paths, the data type of the measurement column will reflect 
the aggregation function rather than the original timeseries.
Correct example: select s0 from root.sg1.d0,root.sg1.d1 align by device   
root.sg1.d0.s0 and root.sg1.d1.s0 are both INT32.  
Correct example: select count(s0) from root.sg1.d0,root.sg1.d1 align by device   
count(root.sg1.d0.s0) and count(root.sg1.d1.s0) are both INT64.  
Wrong example: select s0 from root.sg1.d0, root.sg2.d3 align by device  
root.sg1.d0.s0 is INT32 while root.sg2.d3.s0 is FLOAT. 
5. The display principle of the result table is that all the columns (no matther whther a column has has existing data) will be shown, with nonexistent cells being null. Besides, the select clause support const column (e.g., 'a', '123' etc..).  
For example, "select s0,s1,s2,'abc',s1,s2 from root.sg.d0, root.sg.d1, root.sg.d2 align by device". Suppose that the actual existing timeseries are as follows:  
- root.sg.d0.s0
- root.sg.d0.s1
- root.sg.d1.s0
Then you could expect a table like:  
| Time | Device   | s0 | s1 |  s2  | 'abc' | s1 |  s2  |
| ---  | ---      | ---| ---| null | 'abc' | ---| null |
|  1   |root.sg.d0| 20 | 2.5| null | 'abc' | 2.5| null |
|  2   |root.sg.d0| 23 | 3.1| null | 'abc' | 3.1| null |
| ...  | ...      | ...| ...| null | 'abc' | ...| null |
|  1   |root.sg.d1| 12 |null| null | 'abc' |null| null |
|  2   |root.sg.d1| 19 |null| null | 'abc' |null| null |
| ...  | ...      | ...| ...| null | 'abc' | ...| null |
Note that the cells of measurement 's0' and device 'root.sg.d1' are all null.    
6. The duplicated devices in the prefix paths are neglected.  
For example, "select s0,s1 from root.sg.d0,root.sg.d0,root.sg.d1 align by device" is equal to "select s0,s1 from root.sg.d0,root.sg.d1 align by device".  
For example. "select s0,s1 from root.sg.*,root.sg.d0 align by device" is equal to "select s0,s1 from root.sg.* align by device".  
7. The duplicated measurements in the suffix paths are not neglected.  
For example, "select s0,s0,s1 from root.sg.* align by device" is not equal to "select s0,s1 from root.sg.* align by device".
8. Both time predicates and value predicates are allowed in Where Clause. The paths of the value predicates can be the leaf node or full path started with ROOT. And wildcard is not allowed here. For example:
- select * from root.sg.* where time = 1 align by device
- select * from root.sg.* where s0 < 100 align by device
- select * from root.sg.* where time < 20 AND s0 > 50 align by device
- select * from root.sg.d0 where root.sg.d0.s0 = 15 align by device
9. More correct examples:
   - select * from root.vehicle align by device
   - select s0,s0,s1 from root.vehicle.* align by device
   - select s0,s1 from root.vehicle.* limit 10 offset 1 align by device
   - select * from root.vehicle slimit 10 soffset 2 align by device
   - select * from root.vehicle where time > 10 align by device
   - select * from root.vehicle.* where time < 10 AND s0 > 25 align by device
   - select * from root.vehicle where root.vehicle.d0.s0>0 align by device
   - select count(*) from root.vehicle align by device
   - select sum(*) from root.vehicle GROUP BY (20ms,0,[2,50]) align by device
   - select * from root.vehicle where time = 3 Fill(int32[previous, 5ms]) align by device

Disable Align Statement

Disable Align Clause: DISABLE ALIGN
Rules:  
1. Both uppercase and lowercase are ok.  
Correct example: select * from root.sg1 disable align  
Correct example: select * from root.sg1 DISABLE ALIGN  
2. Disable Align Clause can only be used at the end of a query statement.  
Correct example: select * from root.sg1 where time > 10 disable align 
Wrong example: select * from root.sg1 disable align where time > 10 
3. Disable Align Clause cannot be used with Aggregation, Fill Statements, Group By or Group By Device Statements, but can with Limit Statements.
Correct example: select * from root.sg1 limit 3 offset 2 disable align
Correct example: select * from root.sg1 slimit 3 soffset 2 disable align
Wrong example: select count(s0),count(s1) from root.sg1.d1 disable align
Wrong example: select * from root.vehicle where root.vehicle.d0.s0>0 disable align
Wrong example: select * from root.vehicle align by device disable align
4. The display principle of the result table is that only when the column (or row) has existing data will the column (or row) be shown, with nonexistent cells being empty.
You could expect a table like:
| Time | root.sg.d0.s1 | Time | root.sg.d0.s2 | Time | root.sg.d1.s1 |
| ---  | ---           | ---  | ---           | ---  | ---           |
|  1   | 100           | 20   | 300           | 400  | 600           |
|  2   | 300           | 40   | 800           | 700  | 900           |
|  4   | 500           |      |               | 800  | 1000          |
|      |               |      |               | 900  | 8000          |
5. More correct examples: 
   - select * from root.vehicle disable align
   - select s0,s0,s1 from root.vehicle.* disable align
   - select s0,s1 from root.vehicle.* limit 10 offset 1 disable align
   - select * from root.vehicle slimit 10 soffset 2 disable align
   - select * from root.vehicle where time > 10 disable align

Select Last Record Statement

The LAST function returns the last time-value pair of the given timeseries. Currently filters are not supported in LAST queries.

SELECT LAST <SelectClause> FROM <FromClause>
Select Clause : <Path> [COMMA <Path>]*
FromClause : < PrefixPath > [COMMA < PrefixPath >]*
WhereClause : <TimeExpr> [(AND | OR) <TimeExpr>]*
TimeExpr : TIME PrecedenceEqualOperator (<TimeValue> | <RelativeTime>)
Eg. SELECT LAST s1 FROM root.sg.d1
Eg. SELECT LAST s1, s2 FROM root.sg.d1
Eg. SELECT LAST s1 FROM root.sg.d1, root.sg.d2
Eg. SELECT LAST s1 FROM root.sg.d1 where time > 100
Eg. SELECT LAST s1, s2 FROM root.sg.d1 where time >= 500
Rules:
1. the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>
2. SELECT LAST only supports time filter that contains '>' or '>=' currently.
3. The result set of last query will always be displayed in a fixed three column table format.
For example, "select last s1, s2 from root.sg.d1, root.sg.d2", the query result would be:
| Time | Path         | Value |
| ---  | ------------ | ----- |
|  5   | root.sg.d1.s1| 100   |
|  2   | root.sg.d1.s2| 400   |
|  4   | root.sg.d2.s1| 250   |
|  9   | root.sg.d2.s2| 600   |
4. It is not supported to use "diable align" in LAST query.

As Statement

As statement assigns an alias to time seires queried in SELECT statement

You can use as statement in all queries, but some rules are restricted about wildcard.
1. Raw data query
select s1 as speed, s2 as temperature from root.sg.d1
The result set will be like：
| Time | speed | temperature |
|  ... |  ...  |     ....    |
2. Aggregation query
select count(s1) as s1_num, max_value(s2) as s2_max from root.sg.d1
3. Down-frequence query
select count(s1) as s1_num from root.sg.d1 group by ([100,500), 80ms)
4. Align by device query
select s1 as speed, s2 as temperature from root.sg.d1 align by device
select count(s1) as s1_num, count(s2), count(s3) as s3_num from root.sg.d2 align by device
5. Last Record query
select last s1 as speed, s2 from root.sg.d1
Rules：
1. In addition to Align by device query，each AS statement has to corresponding to one time series exactly.
E.g. select s1 as temperature from root.sg.*
At this time if `root.sg.*` includes more than one device，then an exception will be thrown。
2. In align by device query，the prefix path that each AS statement corresponding to can includes multiple device, but the suffix path can only be single sensor.
E.g. select s1 as temperature from root.sg.*
In this situation, it will be show correctly even if multiple devices are selected.
E.g. select * as temperature from root.sg.d1
In this situation, it will throws an exception if * corresponds to multiple sensors.

Database Management Statement

Create User

CREATE USER <userName> <password>;  
userName:=identifier  
password:=string
Eg: IoTDB > CREATE USER thulab 'pwd';

Delete User

DROP USER <userName>;  
userName:=identifier
Eg: IoTDB > DROP USER xiaoming;

Create Role

CREATE ROLE <roleName>;  
roleName:=identifie
Eg: IoTDB > CREATE ROLE admin;

Delete Role

DROP ROLE <roleName>;  
roleName:=identifier
Eg: IoTDB > DROP ROLE admin;

Grant User Privileges

GRANT USER <userName> PRIVILEGES <privileges> ON <nodeName>;  
userName:=identifier  
nodeName:=identifier (DOT identifier)*  
privileges:= string (COMMA string)*
Eg: IoTDB > GRANT USER tempuser PRIVILEGES 'DELETE_TIMESERIES' on root.ln;

Grant Role Privileges

GRANT ROLE <roleName> PRIVILEGES <privileges> ON <nodeName>;  
privileges:= string (COMMA string)*  
roleName:=identifier  
nodeName:=identifier (DOT identifier)*
Eg: IoTDB > GRANT ROLE temprole PRIVILEGES 'DELETE_TIMESERIES' ON root.ln;

Grant User Role

GRANT <roleName> TO <userName>;  
roleName:=identifier  
userName:=identifier
Eg: IoTDB > GRANT temprole TO tempuser;

Revoke User Privileges

REVOKE USER <userName> PRIVILEGES <privileges> ON <nodeName>;   
privileges:= string (COMMA string)*  
userName:=identifier  
nodeName:=identifier (DOT identifier)*
Eg: IoTDB > REVOKE USER tempuser PRIVILEGES 'DELETE_TIMESERIES' on root.ln;

Revoke Role Privileges

REVOKE ROLE <roleName> PRIVILEGES <privileges> ON <nodeName>;  
privileges:= string (COMMA string)*  
roleName:= identifier  
nodeName:=identifier (DOT identifier)*
Eg: IoTDB > REVOKE ROLE temprole PRIVILEGES 'DELETE_TIMESERIES' ON root.ln;

Revoke Role From User

REVOKE <roleName> FROM <userName>;
roleName:=identifier
userName:=identifier
Eg: IoTDB > REVOKE temprole FROM tempuser;

List Users

LIST USER
Eg: IoTDB > LIST USER

List Roles

LIST ROLE
Eg: IoTDB > LIST ROLE

List Privileges

LIST PRIVILEGES USER  <username> ON <path>;    
username:=identifier    
path=‘root’ (DOT identifier)*
Eg: IoTDB > LIST PRIVILEGES USER sgcc_wirte_user ON root.sgcc;

List Privileges of Roles(On Specific Path)

LIST PRIVILEGES ROLE <roleName> ON <path>;    
roleName:=identifier  
path=‘root’ (DOT identifier)*
Eg: IoTDB > LIST PRIVILEGES ROLE wirte_role ON root.sgcc;

List Privileges of Users

LIST USER PRIVILEGES <username> ;   
username:=identifier  
Eg: IoTDB > LIST USER PRIVILEGES tempuser;

List Privileges of Roles

LIST ROLE PRIVILEGES <roleName>
roleName:=identifier
Eg: IoTDB > LIST ROLE PRIVILEGES actor;

List Roles of Users

LIST ALL ROLE OF USER <username> ;  
username:=identifier
Eg: IoTDB > LIST ALL ROLE OF USER tempuser;

List Users of Role

LIST ALL USER OF ROLE <roleName>;
roleName:=identifier
Eg: IoTDB > LIST ALL USER OF ROLE roleuser;

Alter Password

ALTER USER <username> SET PASSWORD <password>;
roleName:=identifier
password:=identifier
Eg: IoTDB > ALTER USER tempuser SET PASSWORD 'newpwd';

Functions

COUNT

The COUNT function returns the value number of timeseries(one or more) non-null values selected by the SELECT statement. The result is a signed 64-bit integer. If there are no matching rows, COUNT () returns 0.

SELECT COUNT(Path) (COMMA COUNT(Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT COUNT(status), COUNT(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

FIRST_VALUE(Rename from FIRST at V0.10.0)

The FIRST_VALUE function returns the first point value of the choosen timeseries(one or more).

SELECT FIRST_VALUE (Path) (COMMA FIRST_VALUE (Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT FIRST_VALUE (status), FIRST_VALUE (temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

LAST_VALUE

The LAST_VALUE function returns the last point value of the choosen timeseries(one or more).

SELECT LAST_VALUE (Path) (COMMA LAST_VALUE (Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT LAST_VALUE (status), LAST_VALUE (temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

MAX_TIME

The MAX_TIME function returns the maximum timestamp of the choosen timeseries(one or more). The result is a signed 64-bit integer, greater than 0.

SELECT MAX_TIME (Path) (COMMA MAX_TIME (Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT MAX_TIME(status), MAX_TIME(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

MAX_VALUE

The MAX_VALUE function returns the maximum value(lexicographically ordered) of the choosen timeseries (one or more).

SELECT MAX_VALUE (Path) (COMMA MAX_VALUE (Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT MAX_VALUE(status), MAX_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

AVG(Rename from MEAN at V0.9.0)

The AVG function returns the arithmetic mean value of the choosen timeseries over a specified period of time. The timeseries must be int32, int64, float, double type, and the other types are not to be calculated. The result is a double type number.

SELECT AVG (Path) (COMMA AVG (Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT AVG (temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

MIN_TIME

The MIN_TIME function returns the minimum timestamp of the choosen timeseries(one or more). The result is a signed 64-bit integer, greater than 0.

SELECT MIN_TIME (Path) (COMMA MIN_TIME (Path))*FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT MIN_TIME(status), MIN_TIME(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

MIN_VALUE

The MIN_VALUE function returns the minimum value(lexicographically ordered) of the choosen timeseries (one or more).

SELECT MIN_VALUE (Path) (COMMA MIN_VALUE (Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT MIN_VALUE(status),MIN_VALUE(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

The NOW function returns the current timestamp. This function can be used in the data operation statement to represent time. The result is a signed 64-bit integer, greater than 0.

NOW()
Eg. INSERT INTO root.ln.wf01.wt01(timestamp,status) VALUES(NOW(), false)
Eg. DELETE FROM root.ln.wf01.wt01.status, root.ln.wf01.wt01.temperature WHERE time < NOW()
Eg. SELECT * FROM root WHERE time < NOW()
Eg. SELECT COUNT(temperature) FROM root.ln.wf01.wt01 WHERE time < NOW()

The SUM function returns the sum of the choosen timeseries (one or more) over a specified period of time. The timeseries must be int32, int64, float, double type, and the other types are not to be calculated. The result is a double type number.

SELECT SUM(Path) (COMMA SUM(Path))* FROM <FromClause> [WHERE <WhereClause>]?
Eg. SELECT SUM(temperature) FROM root.ln.wf01.wt01 WHERE root.ln.wf01.wt01.temperature < 24
Note: the statement needs to satisfy this constraint: <PrefixPath> + <Path> = <Timeseries>

TTL

IoTDB supports storage-level TTL settings, which means it is able to delete old data automatically and periodically. The benefit of using TTL is that hopefully you can control the total disk space usage and prevent the machine from running out of disks. Moreover, the query performance may downgrade as the total number of files goes up and the memory usage also increase as there are more files. Timely removing such files helps to keep at a high query performance level and reduce memory usage. The TTL operations in IoTDB are supported by the following three statements:

Set TTL

SET TTL TO StorageGroupName TTLTime
Eg. SET TTL TO root.group1 3600000
This example means that for data in root.group1, only that of the latest 1 hour will remain, the
older one is removed or made invisible. 
Note: TTLTime should be millisecond timestamp. When TTL is set, insertions that fall
out of TTL will be rejected.

Unset TTL

UNSET TTL TO StorageGroupName
Eg. UNSET TTL TO root.group1
This example means that data of all time will be accepted in this group.

Show TTL

SHOW ALL TTL
SHOW TTL ON StorageGroupNames
Eg.1 SHOW ALL TTL
This example will show TTLs of all storage groups.
Eg.2 SHOW TTL ON root.group1,root.group2,root.group3
This example will show TTLs of the specified 3 groups.
Notice: storage groups without TTL will show a "null"

Notice: When you set TTL to some storage groups, data out of the TTL will be made invisible immediately, but because the data files may contain both out-dated and living data or the data files may be being used by queries, the physical removal of data is stale. If you increase or unset TTL just after setting it previously, some previously invisible data may be seen again, but the physically removed one is lost forever. In other words, different from delete statement, the atomicity of data deletion is not guaranteed for efficiency concerns. So we recommend that you do not change the TTL once it is set or at least do not reset it frequently, unless you are determined to suffer the unpredictability.

Delete Partition (experimental)

DELETE PARTITION StorageGroupName INT(COMMA INT)*
Eg DELETE PARTITION root.sg1 0,1,2
This example will delete the first 3 time partitions of storage group root.sg1.

The partitionId can be found in data folders or converted using timestamp / partitionInterval.

Performance Tracing

IoTDB supports tracking the execution of query statements by using TRACING statements. The number of tsfile files and chunks accessed by the query etc are output through the log file. The default output location is in ./data/tracing. The performance tracing function is turned off by default. Users can use the TRACING ON/OFF command to turn this function on/off.

TRACING ON    // Open performance tracing
TRACING OFF   // Close performance tracing

Reference

Identifiers

QUOTE := '\'';
DOT := '.';
COLON : ':' ;
COMMA := ',' ;
SEMICOLON := ';' ;
LPAREN := '(' ;
RPAREN := ')' ;
LBRACKET := '[';
RBRACKET := ']';
EQUAL := '=' | '==';
NOTEQUAL := '<>' | '!=';
LESSTHANOREQUALTO := '<=';
LESSTHAN := '<';
GREATERTHANOREQUALTO := '>=';
GREATERTHAN := '>';
DIVIDE := '/';
PLUS := '+';
MINUS := '-';
STAR := '*';
Letter := 'a'..'z' | 'A'..'Z';
HexDigit := 'a'..'f' | 'A'..'F';
Digit := '0'..'9';
Boolean := TRUE | FALSE | 0 | 1 (case insensitive)

StringLiteral := ( '\'' ( ~('\'') )* '\'' | '\"' ( ~('\"') )* '\"');
eg. 'abc'
eg. 'abc'

Integer := ('-' | '+')? Digit+;
eg. 123
eg. -222

Float := ('-' | '+')? Digit+ DOT Digit+ (('e' | 'E') ('-' | '+')? Digit+)?;
eg. 3.1415
eg. 1.2E10
eg. -1.33

Identifier := (Letter | '_') (Letter | Digit | '_' | MINUS)*;
eg. a123
eg. _abc123

Literals

PointValue : Integer | Float | StringLiteral | Boolean

TimeValue : Integer | DateTime | ISO8601 | NOW()
Note: Integer means timestamp type.
DateTime : 
eg. 2016-11-16T16:22:33+08:00
eg. 2016-11-16 16:22:33+08:00
eg. 2016-11-16T16:22:33.000+08:00
eg. 2016-11-16 16:22:33.000+08:00
Note: DateTime Type can support several types, see Chapter 3 Datetime section for details.

PrecedenceEqualOperator : EQUAL | NOTEQUAL | LESSTHANOREQUALTO | LESSTHAN | GREATERTHANOREQUALTO | GREATERTHAN

Timeseries : ROOT [DOT <LayerName>]* DOT <SensorName>
LayerName : Identifier
SensorName : Identifier
eg. root.ln.wf01.wt01.status
eg. root.sgcc.wf03.wt01.temperature
Note: Timeseries must be start with `root`(case insensitive) and end with sensor name.

PrefixPath : ROOT (DOT <LayerName>)*
LayerName : Identifier | STAR
eg. root.sgcc
eg. root.*

Path: (ROOT | <LayerName>) (DOT <LayerName>)* 
LayerName: Identifier | STAR
eg. root.ln.wf01.wt01.status
eg. root.*.wf01.wt01.status
eg. root.ln.wf01.wt01.*
eg. *.wt01.*
eg. *