Create a TSDB result table

CREATE TABLE stream_test_hitsdb (
    metric      VARCHAR,   
    `timestamp` INTEGER,
    `value`     DOUBLE,
    tagk1       VARCHAR,     
    tagk2       VARCHAR,
    tagk3       VARCHAR
) WITH (
    type='hitsdb',
    host='<yourHostName>',
    virtualDomainSwitch = 'false',
    httpConnectionPool = '20',
    batchPutSize = '1000'
);

• Column 0: metric (VARCHAR).
• Column 1: timestamp (INTEGER). Unit: seconds.
• Column 2: value (DOUBLE).
• Columns 3 to N: tag keys. The field names in the time series database are used as the tag keys.

• Write single-value data points where no tags are included. To use this model, use the specified schema that consists of three fields. The names of these fields cannot be changed. You must use the following schema format for this model:

  metric,timestamp,value

• Write single-value data points where tags are included. To use this model, use the specified schema. In the schema, the names of the following fields cannot be changed: metric, timestamp, and value. You can specify the tag names based on your business requirements. You must use the following schema format for this model:

  metric,timestamp,value,tagKey1,....,tagKeyN

• Write single-value data points where the number of tags is unknown. To use this model, use the specified schema that consists of four fields. The names of the four fields cannot be changed. You must use the following schema format for this model:

  metric,timestamp,value,tags

  In the schema, specify the value of the tags parameter as a JSON string. The JSON string allows you to specify an unknown number of tags. If you do not use the JSON string, you must specify a fixed number of tags for the Blink table schema. In the schema, specify the JSON string in the following format:

  {"tagKey1":"tagValue1","tagKey2":"tagValue2",……,"tagKeyN":"tagValueN"}

• Write multi-value data points where tags are not included. You must use the following schema format for this model:

  metric,timestamp,field_name1,field_name2,……,field_nameN

  The names of the metric and timestamp fields cannot be changed. For multi-value fields, the field_ prefix is used for each field to distinguish fields from tags and to support single-value data writes. For example, when the field_name1 field is written to TSDB, the prefix field_ is automatically removed. In the preceding schema, name1 and name2 are the names of the multi-value fields. For the preceding schema, the following format is used to write data to TSDB:

  metric,timestamp,name1,name2,……,nameN

• Write multi-value data points where tags are included. To use this model, use the specified schema. In the schema, the names of the following fields cannot be changed: metric and timestamp. You can specify the tag names based on your business requirements. You must use the following schema format for this model:

  metric,timestamp,tagKey1,....,tagKeyN,field_name1,field_name2,……,field_nameN

• Write single-value data points where the number of tags is unknown. You must use the following schema format for this model:

  metric,timestamp,tags,field_name1,field_name2,……,field_nameN

  The content of tags is a JSON string shown in the following example. Therefore, the limit that the Blink table schema requires a fixed number of tags does not apply.

  {"tagKey1":"tagValue1","tagKey2":"tagValue2",……,"tagKeyN":"tagValueN"}