ReferenceApplies to: Unified engine deployments
[result_read_limits]
Settings
max_batches_per_page
Limits how many stored result batches can be returned in one result-read page.
- Default:
16
- Type: integer; positive
Example
16 allows up to 16 batches. If the byte limit is reached first, a page may contain fewer batches.
max_bytes_per_page
Limits the combined schema and result payload bytes returned in one result-read page.
- Default:
75497472
- Type: integer; positive
Example
75497472 gives a page a 72 MiB budget. A client may need to read more pages to retrieve a larger result.
max_continuation_token_bytes
Limits the size of the token used to request the next result page.
- Default:
4096
- Type: integer (bytes); positive
Example
4096 bounds the token to 4 KiB. This token records where reading should continue; it does not contain all result rows.
Limits the size of result metadata that the reader will accept.
- Default:
8388608
- Type: integer (bytes); positive
Example
8388608 allows up to 8 MiB describing the result and its batches. The batch payloads have separate read limits.
max_result_batches
Limits how many batch references one saved result may list.
- Default:
100000
- Type: integer; positive
Example
100000 bounds the metadata for a result made of many small batches. It does not limit every batch to one row.
Last modified on November 4, 2026
