influxdata/influxdb · error

time column

Error message

time column

What it means

filter_time unwraps batch.time_column() with the message 'time column'. IOx batches are invariantly required to carry a non-nullable time column, so its absence is a programming error and the library panics rather than returning an error.

Solutions

1. Ensure the batch always contains the time column before filtering
2. Check the code path that produced the batch for a dropped/renamed time column
3. In tests, use the batch builders that include the time column

Example fix

// before
let out = filter_ranges(&batch, &[0..10], |t| t > 0);
// after
assert!(batch.time_column().is_some(), "batch missing time column");
let out = filter_ranges(&batch, &[0..10], |t| t > 0);

if batch.time_column().is_none() { return Err("batch missing time column"); }

fn has_time_column(batch: &RecordBatch) -> bool { batch.time_column().is_some() }

When it happens

Trigger: Calling filter_time (via filter_ranges/filter_batch) on a RecordBatch that has no time column, e.g. a batch built manually without it or with the time column dropped.

Common situations: Test helpers constructing batches from only tag/field columns; upstream pipeline bug dropping the time column before filtering.

Understand the failure class

Background: "This is a bug, please report it": internal invariant violations, unreachable panics, and SNH errors explained — this error's family across 47 libraries.

Related errors

• no time column in metadata statistics
• By the point that we're doing partitioning, we should've…
• Existing transaction for table should not exist
• mapped to out-of-bounds shard
• ordered catalog batch should contain changes

AI-assisted analysis of influxdata/influxdb@06200ef96b (2026-09-19). Data as JSON: /api/errors/89b3e45f9aa5e41b. Report an issue: GitHub.

Appendix: source

Thrown at core/partition/src/filter.rs:23

use crate::Batch;
use mutable_batch::MutableBatch;
use std::ops::Range;

/// Given a [`MutableBatch`] a time predicate and a set of row ranges, returns the row
/// indexes that pass the predicate
///
/// # Panic
///
/// Panics if `batch` does not contain a time column of the correct type
pub(crate) fn filter_time<'a, F>(
    batch: &'a MutableBatch,
    ranges: &'a [Range<usize>],
    mut predicate: F,
) -> Vec<Range<usize>>
where
    F: FnMut(i64) -> bool,
{
    let col_data = batch.time_column().expect("time column");

    // Time column is not nullable so can skip checking mask
    let mut ret = vec![];
    for range in ranges {
        let offset = range.start;
        ret.extend(
            filter_slice(&col_data[range.clone()], &mut predicate)
                .map(|r| (r.start + offset)..(r.end + offset)),
        )
    }
    ret
}

fn filter_slice<'a, T, F>(
    col_data: &'a [T],
    predicate: &'a mut F,
) -> impl Iterator<Item = Range<usize>> + 'a
where

View on GitHub (pinned to 06200ef96b)