Skip to main content

Configure the stream generator

Please check Load sample streaming data for the overview of this data source.

Sample configuration

Here is a sample of stream generator configuraion json:

"batch_size": 1,
"interval": 200,
"fields": [
"name": "number",
"type": "int",
"limit": [
"name": "_time",
"type": "timestamp",
"timestamp_format": "2006-01-02 15:04:05.000"

The usage of each attribute:

Field NameDescriptionSample Value
batch_sizehow many batched to be generated for each iteration2
intervalthe interval between each iteration in millisecond1000
interval_deltaa random delta for the interval of each iteration300
counthow many events to be generated, if 0 is set, there is no limits for the events1000
fieldsa list of fields definition

For fields, the following attributes are supported:

Field NameDescription
namename of the field
typewhat types of data to be generated, support timestamp, string, int, float, bool, map, array, generate, regex
rangeoptional for string, int and float, which is list of value that can be generated
limitoptional for int and float, a list with two values that specify the min/max of the generated data
timestamp_formatoptional for timestamp type, following golang time string format rules
timestamp_delay_minminimal delay for timestamp in ms
timestamp_delay_maxmaximal delay for timestamp in ms
rulea generation rule in case the type is generate or regex


By default, the timestamp will be generated as a int which is the unit timestamp that is the time in ms since 1970.1.1. if timestamp_format is specified, it will be generated as a string using this format. Refer to this document for details.

Regex Rule

In case the field type is set as regex, the generated data will be based on the regular expression specified in rule fields.

Generator Rule

In case the field type is set as generator, the generated data will be based on the generation rule of gofakeit, refer to rule