Documentation

Partition templates

Use partition templates to define the patterns used to generate partition keys. A partition key uniquely identifies a partition and is used to name the partition Parquet file in the Object store.

A partition template defines how InfluxDB groups data into partitions by specifying 1-8 template parts. Each template part represents a dimension to partition data by.

Template part types

InfluxDB supports three types of partition template parts:

  • Tag part: Partitions data by the unique values of an InfluxDB tag. For example, using region as a tag part creates separate partitions for each region value (us-west, us-east, eu-central).

  • Tag bucket part: Partitions data by “buckets” of InfluxDB tag values. Instead of creating a partition for every unique tag value, tag values are hashed and grouped into a specified number of buckets. Use this for high-cardinality tags or when the number of distinct values is unknown.

  • * Required Time part: Partitions data by time intervals using a Rust strftime date and time format string. The smallest time unit in your format determines the granularity of time partitioning (yearly with %Y, monthly with %Y-%m, or daily with %Y-%m-%d).

Requirements and guidelines

When creating a partition template:

  1. Include exactly one time part

    • Always specify a time part in your template

    • With influxctl, always include --template-timeformat with a valid format

    • Without a time part, InfluxDB won’t compact partitions, impacting performance

    • If you include more than one time part, InfluxDB uses the smallest unit of time

    • Use one of the following Rust strftime date and time strings:

      • %Y-%m-%d (daily)
      • %Y-%m (monthly)
      • %Y (annually)
  2. Tag and tag bucket limitations

    • Include up to seven tag and tag bucket parts
    • Don’t use the same tag key in both a tag part and a tag bucket part–for example, if your template uses region as a tag part, you cannot use region as a tag bucket part
  3. Maximum template parts: 8 total (1 time part + up to 7 tag and tag bucket parts)

Restrictions

Template part size limit

Each template part is limited to 200 bytes in length. Anything longer will be truncated at 200 bytes and appended with #.

Partition key size limit

With the truncation of template parts, the maximum length of a partition key is 1,607 bytes (1.57 KiB).

Reserved keywords

The following reserved keywords cannot be used in partition templates:

  • time

Reserved Characters

If used in template parts, non-ASCII characters and the following reserved characters must be percent encoded:

  • |: Partition key part delimiter
  • !: Null or missing partition key part
  • ^: Empty string partition key part
  • #: Key part truncation marker
  • %: Required for unambiguous reversal of percent encoding

Tag part templates

Tag part templates consist of a tag key to partition by. Generated partition keys include the unique tag value specific to each partition.

A partition template may include a given tag key only once in template parts that operate on tags (tag value and tag bucket)–for example, if a template partitions on unique values of tag_A, then you can’t use tag_A as a tag bucket part.

Tag bucket part templates

Tag bucket part templates consist of a tag key to partition by and the number of “buckets” to partition tag values into–for example:

customerID,500
  • Copy
  • Fill window

Values of the customerID tag are bucketed into 500 distinct “buckets.” Each bucket is identified by the remainder of the tag value hashed into a 32-bit integer divided by the specified number of buckets:

hash(tagValue) % N
  • Copy
  • Fill window

Generated partition keys include the unique tag bucket identifier specific to each partition.

Supported number of tag buckets: 1-1,000

Tag buckets should be used to partition by high cardinality tags or tags with an unknown number of distinct values.

A partition template may include a given tag key only once in template parts that operate on tags (tag value and tag bucket)–for example, if a template partitions on unique values of tag_A, then you can’t use tag_A as a tag bucket part.

Time part templates

Time part templates use a limited subset of the Rust strftime date and time formatting syntax to specify time format in partition keys. Time part templates can be daily (%Y-%m-%d), monthly (%Y-%m), or yearly (%Y). InfluxDB partitions data by the smallest unit of time included in the time part template.

InfluxDB supports only date specifiers in time part templates.

Date specifiers

Time part templates allow only the following date specifiers:

VariableExampleDescription
%Y2001The full proleptic Gregorian year, zero-padded to 4 digits. chrono supports years from -262144 to 262143. Note: years before 1 BCE or after 9999 CE, require an initial sign (+/-).
%m07Month number (01–12), zero-padded to 2 digits.
%d08Day number (01–31), zero-padded to 2 digits.

Was this page helpful?

Thank you for your feedback!


The future of Flux

Flux is going into maintenance mode. You can continue using it as you currently are without any changes to your code.

Read more

Now Generally Available

InfluxDB 3 Core and Enterprise

Start fast. Scale faster.

Get the Updates

InfluxDB 3 Core is an open source, high-speed, recent-data engine that collects and processes data in real-time and persists it to local disk or object storage. InfluxDB 3 Enterprise builds on Core’s foundation, adding high availability, read replicas, enhanced security, and data compaction for faster queries and optimized storage. A free tier of InfluxDB 3 Enterprise is available for non-commercial at-home or hobbyist use.

For more information, check out: