Default timestamp

This guide explains how to choose the default timestamp for your Data Pool in Propel’s Serverless ClickHouse.

​

What is a Default Timestamp?

A default timestamp is an optional property of a Data Pool that:

• Provides a time dimension for Metrics
• Simplifies time-based queries for developers

​

How to use the default timestamp

​

1. Metric creation

When defining a Metric, Propel automatically uses the Data Pool’s default timestamp as the time dimension.

​

2. Streamlined time series queries

With a default timestamp, you can omit the timestamp specification in time range queries.

Example query using the default timestamp. Note that the timestamp field is not specified:

query TimeSeriesQuery {
  timeSeries(input: {
    metric: {
      count: {
        dataPool: {
          name: "TacoSoft Demo Data"
        }
      }
    },
    timeZone: "UTC",
    granularity: "DAY",
    timeRange: {
      relative: "LAST_N_DAYS",
      n: 30,
    },
    filters: []
  }) {
    labels
    values
  }
}

​

How to select a good default timestamp for your Data Pool

To select a good default timestamp, consider two factors:

1. The default timestamp should be the primary time dimension that you filter on.
2. It should be part of the sorting key to ensure good query performance. For more information on the sorting key, see How to select a table engine and sorting key.

​

Querying with a different timestamp

Even when you define a default timestamp for a Data Pool, you can still query it using a different timestamp by specifying it in the query.

The query below shows how to pass a different timestamp at query time:

query TimeSeriesQuery {
  timeSeries(input: {
    metric: {
      count: {
        dataPool: {
          name: "TacoSoft Demo Data"
        }
      }
    },
    timeZone: "UTC",
    granularity: "DAY",
    timeRange: {
      relative: "LAST_N_DAYS",
      n: 30,
      timestamp: "order_item_generated_at"
    },
    filters: []
  }) {
    labels
    values
  }
}