Skip to main content

Billing overview

Imply Polaris operates on a consumption-based pricing model. Each project is billed based on the project size, batch and streaming ingestion volumes, the amount of file storage and deep storage used, and asynchronous (async) queries.

Users with the AdministerBilling permission can view total accrued charges and per-project usage charges in the Billing page of the Polaris UI.

For current pricing, see Imply Polaris pricing.

Billing dimensions

Polaris tracks usage across the following billing dimensions:

Project size

Polaris measures the number of hours a project has been operating, starting at the beginning of each hour. Monthly project size resource consumption is calculated by multiplying the project size price by the total number of hours used in the month.

You can choose between starter, series A, and series D project types:

  • The starter project is best suited for functional evaluation of Polaris during trial or proof of concept. It is limited to 25 GB storage capacity, 4 vCPU, and 16 GB memory.
  • Series A are best suited for compute-intensive analytical workloads that power critical business applications.
  • Series D are best suited for storage-intensive production applications with lower cost and moderate performance needs relative to the amount of queryable data.

Series A

The following table shows series A project sizes:

Project sizeCapacityvCPU allocationMemory
A.01100 GB10 vCPU64 GB RAM
A.02200 GB20 vCPU128 GB RAM
A.04400 GB24 vCPU160 GB RAM
A.08800 GB40 vCPU288 GB RAM
A.121,200 GB60 vCPU432 GB RAM
A.161,600 GB76 vCPU560 GB RAM
A.242,400 GB108 vCPU816 GB RAM
A.323,200 GB144 vCPU1088 GB RAM
A.484,800 GB216 vCPU1632 GB RAM
A.646,400 GB320 vCPU2432 GB RAM
A.969,600 GB432 vCPU3264 GB RAM

Series D

The following table shows series D project sizes:

Project sizeCapacityvCPU allocationMemory
D.04400 GB8 vCPU48 GB RAM
D.08800 GB12 vCPU80 GB RAM
D.121,200 GB20 vCPU128 GB RAM
D.161,600 GB24 vCPU160 GB RAM
D.242,400 GB32 vCPU224 GB RAM
D.323,200 GB40 vCPU288 GB RAM
D.484,800 GB60 vCPU432 GB RAM
D.646,400 GB76 vCPU560 GB RAM
D.969,600 GB112 vCPU832 GB RAM

If you are unsure of your data usage, sign up for a free trial to ingest and size your data. After you move out of the trial and become a paid customer, you can select from the following default project sizes: starter, A.01, A.02, A.04, A.08, D.04, D.08, D.12. To change to a different project size, contact Imply.

Data ingestion

Polaris measures your consumption of batch and streaming ingestion resources based on the processed bytes, which refer to the uncompressed size of your input data in a successful ingestion job. The processed size may differ from the original input size due to format and compression.

Polaris bills you for every ingestion job even if you're ingesting the same data into different tables. For example, if you ingest a stream as both detailed and aggregate tables, this increments the billable processed bytes twice. Reingesting data that already exists in Polaris also increases your billable processed bytes.

Batch and streaming ingestion resources are measured and billed independently using the following tiered model:

  • First terabyte (1000 GB) in a month
  • Next 9 terabytes (9000 GB) in a month
  • Anything over 10 terabytes in a month

The price for streaming ingestion remains the same regardless of the streaming ingestion source. For supported streaming ingestion sources, see Ingestion sources.

File storage

Polaris charges you for file storage when you upload files to the staging area for batch ingestion. File storage consumption is calculated by multiplying the file storage price by the total volume of data stored per month (in GB-month).

Deep storage

Deep storage encompasses all persisted data, including precached and noncached data. When you query deep storage data, the results also include real-time data. Unlike data loaded into precache, data stored in deep storage doesn't count towards your project size. It is billed at a much lower rate, making it a cost effective solution for long term data retention.

Polaris calculates deep storage consumption by multiplying the deep storage price by the total data volume stored per month (in GB-month).

To reduce costs and conserve resources, especially for infrequently accessed data, you can implement a precache policy to automatically stop caching older data. For information on precache policies, see Data lifecycle management.

Async query

Polaris measures the compute resources consumed by async queries in Data Processing Unit (DPU) minutes. A DPU is a normalized unit of processing power in Polaris. One DPU-minute corresponds to one task consuming one DPU for one minute. Each async query task running for one minute is billed as one DPU-minute.

Because async queries run on dedicated hardware that Polaris provisions on demand, you only pay for the query while it’s running, without any ongoing computing or maintenance costs.

You can manage how many compute resources an async query consumes by controlling task concurrency. Higher concurrency typically results in faster execution times. For instance, a query with a concurrency of two that takes 12 minutes to execute can take around six minutes with a concurrency of four.

Cloud provider marketplace

The method you use to sign up for Polaris determines how your organization is billed. If you sign up directly on Polaris, Polaris bills you for the resources you consume. If you sign up through a cloud provider, the provider manages all billing and invoicing for your Polaris subscription. For current provider pricing, refer to the Polaris marketplace listing on Amazon Web Services (AWS) and Microsoft Azure. To change your billing method, contact Polaris support.

Learn more

See the following topics for more information: