Polars On-Prem

Polars On-Prem lets you host a Polars cluster on your own hardware or cloud provider. Once connected, you can run queries remotely, either executing multiple queries in parallel, or distributing a single query across multiple nodes for large-scale workloads.

Interested in running Polars On-Prem? Sign up here to apply.

After that you can install by getting the latest version here.

After installing Polars On-Prem either on bare-metal or on Kubernetes, you can connect to your cluster using the Polars Cloud Python client.

import polars as pl
import polars_cloud as pc

# Connect to your Polars On-Prem cluster
ctx = pc.ClusterContext(compute_address="your-cluster-compute-address")
result = (
    pl.LazyFrame()
    .with_columns(a=pl.arange(0, 100000000).sum())
    .remote(ctx)
    .distributed()
    .execute()
)
print(result)