xai-org/x-algorithm · error · anyhow::Error

no user_actions

Error message

no user_actions

What it means

extract_single_user_arrow expects an Arrow RecordBatch containing a column named 'user_actions'. When column_by_name returns None (the column is absent from the batch schema), this error is raised. It typically means upstream data production and this consumer disagree on the expected schema.

Source

Thrown at bdsm/rust/abuse-v3-features/src/lib.rs:260

    let decompressed = zstd::decode_all(Cursor::new(raw))?;
    let cursor = Cursor::new(decompressed);
    let reader = FileReader::try_new(cursor, None)?;
    let mut batches = Vec::new();
    for b in reader {
        batches.push(b?);
    }
    if batches.is_empty() {
        anyhow::bail!("empty");
    }
    let batch = if batches.len() == 1 {
        batches.into_iter().next().unwrap()
    } else {
        arrow::compute::concat_batches(&batches[0].schema(), &batches)?
    };

    let actions_col = batch
        .column_by_name("user_actions")
        .ok_or_else(|| anyhow::anyhow!("no user_actions"))?;
    let list = actions_col
        .as_any()
        .downcast_ref::<ListArray>()
        .ok_or_else(|| anyhow::anyhow!("not ListArray"))?;
    if list.is_empty() || list.is_null(0) {
        anyhow::bail!("empty list");
    }

    let vals = list.value(0);
    let sa = vals
        .as_any()
        .downcast_ref::<StructArray>()
        .ok_or_else(|| anyhow::anyhow!("not StructArray"))?;
    let total = sa.len();
    if total == 0 {
        anyhow::bail!("zero actions");
    }

View on GitHub (pinned to 24c60942c5)

Solutions

  1. Inspect batch.schema() field names to confirm what the producer actually sent
  2. Align the consumer schema with the producer: fix the column name or update this code to the new field name
  3. Add a schema assertion/validation step before calling extract to fail fast with a clearer message
  4. Pin producer and consumer to compatible schema versions in deployment

Example fix

// before
let actions_col = batch
    .column_by_name("user_actions")
    .ok_or_else(|| anyhow::anyhow!("no user_actions"))?;

// after
let actions_col = batch
    .column_by_name("user_actions")
    .ok_or_else(|| anyhow::anyhow!(
        "no user_actions: batch schema has fields {:?}",
        batch.schema().fields().iter().map(|f| f.name()).collect::<Vec<_>>()
    ))?;
Defensive patterns

Strategy: validation

Validate before calling

// Validate the batch schema before extraction
let expected = Schema::new(vec![Field::new("user_actions", DataType::List(Arc::new(Field::new("item", ActionType::data_type(), true))), false)]);
anyhow::ensure!(batch.schema().contains(&expected) || batch.column_by_name("user_actions").is_some(), "unexpected batch schema");

Type guard

fn has_user_actions(batch: &RecordBatch) -> bool {
    batch.column_by_name("user_actions").is_some()
}

Try / catch

// catch and report the actual schema for fast diagnosis
match extract_batch(batches) {
    Err(e) if e.to_string().contains("no user_actions") => {
        tracing::error!("schema drift: got fields {:?}", batch.schema());
        return Err(e);
    }
    r => r,
}

Prevention

When it happens

Trigger: Calling extract_batch / extract_single_user_arrow on a batch whose schema lacks a 'user_actions' field — e.g. an empty concat input, a renamed column, or a producer writing a different feature schema version.

Common situations: Upstream feature pipeline renamed or dropped the user_actions column, schema drift between producer and consumer versions, or an accidentally empty/mis-routed dataset being fed into the extractor.

Understand the failure class

Background: Schema validation failed / invalid input schema: payload rejected because its shape doesn't match the expected schema — this error's family across 28 libraries.

Related errors

AI-assisted analysis of xai-org/x-algorithm@24c60942c5 (2026-08-28). Data as JSON: /api/errors/4ea5850f80db0d3f. Report an issue: GitHub.