Ballista Rust ClientΒΆ

To connect to a Ballista cluster from Rust, first start by creating a BallistaContext.

let config = BallistaConfig::builder()
    .set("ballista.shuffle.partitions", "4")
    .build()?;

// connect to Ballista scheduler
let ctx = BallistaContext::remote("localhost", 50050, &config);

Here is a full example using the DataFrame API.

#[tokio::main]
async fn main() -> Result<()> {
    let config = BallistaConfig::builder()
        .set("ballista.shuffle.partitions", "4")
        .build()?;

    // connect to Ballista scheduler
    let ctx = BallistaContext::remote("localhost", 50050, &config);

    let testdata = datafusion::test_util::parquet_test_data();

    let filename = &format!("{}/alltypes_plain.parquet", testdata);

    // define the query using the DataFrame trait
    let df = ctx
        .read_parquet(filename)?
        .select_columns(&["id", "bool_col", "timestamp_col"])?
        .filter(col("id").gt(lit(1)))?;

    // print the results
    df.show().await?;

    Ok(())
}

Here is a full example demonstrating SQL usage.

#[tokio::main]
async fn main() -> Result<()> {
    let config = BallistaConfig::builder()
        .set("ballista.shuffle.partitions", "4")
        .build()?;

    // connect to Ballista scheduler
    let ctx = BallistaContext::remote("localhost", 50050, &config);

    let testdata = datafusion::test_util::arrow_test_data();

    // register csv file with the execution context
    ctx.register_csv(
        "aggregate_test_100",
        &format!("{}/csv/aggregate_test_100.csv", testdata),
        CsvReadOptions::new(),
    )?;

    // execute the query
    let df = ctx.sql(
        "SELECT c1, MIN(c12), MAX(c12) \
        FROM aggregate_test_100 \
        WHERE c11 > 0.1 AND c11 < 0.9 \
        GROUP BY c1",
    )?;

    // print the results
    df.show().await?;

    Ok(())
}