Struct QueryTuning 

pub struct QueryTuning {

    pub sort_run_size: usize,
    pub stream_chunk_size: usize,
    pub aggregate_scan_cap: usize,
    pub arrow_batch_max_rows: usize,
    pub arrow_batch_max_bytes: usize,
    pub bitmap_over_fetch_factor: usize,
    pub bfs_memory_budget_bytes: usize,
    pub bfs_bytes_per_node: usize,
    pub doc_cache_entries: usize,
    pub columnar_flush_threshold: usize,
    pub compaction_target_bytes: usize,
    pub groupby_max_groups_in_mem: usize,
    pub aggregate_chunk_size: usize,
    pub max_scan_result_bytes: usize,
}

Query execution tuning for the Data Plane executor.

Fields

sort_run_size: usize

stream_chunk_size: usize

aggregate_scan_cap: usize

arrow_batch_max_rows: usize

arrow_batch_max_bytes: usize

bitmap_over_fetch_factor: usize

bfs_memory_budget_bytes: usize

bfs_bytes_per_node: usize

doc_cache_entries: usize

Per-core LRU document cache size (number of entries). See DocCache::new in core_loop.rs.

columnar_flush_threshold: usize

Columnar memtable flush threshold in rows. See nodedb_columnar::memtable::DEFAULT_FLUSH_THRESHOLD.

compaction_target_bytes: usize

Target segment size in bytes after compaction. See nodedb::storage::compaction::CompactionConfig.

groupby_max_groups_in_mem: usize

Maximum number of distinct group keys held in memory during GROUP BY before the spiller triggers a spill run to disk. When exceeded, the current in-memory accumulator map is serialized to a temp file and the map is cleared. All spill runs are k-way merged at finalize time. Default: 1_000_000.

aggregate_chunk_size: usize

Number of documents the streaming aggregate processes per chunk before checking whether a spill run should be triggered. A smaller value reduces peak memory at the cost of more frequent spill checks; a larger value amortises check overhead at the cost of holding more documents in flight between checks. The LIMIT clause is applied after all chunks are accumulated, so this knob does not affect query correctness. Default: 10_000.

max_scan_result_bytes: usize

Maximum total bytes a single unbounded (no-LIMIT) document scan may materialize on a Data Plane core before the handler aborts with a deterministic ResourcesExhausted error. A SELECT * FROM <coll> without a LIMIT is no longer silently truncated; instead it returns every row that fits this budget and surfaces a typed error (suggesting the client add a LIMIT) if the result would exceed it. This is a per-core, per-query bound that sits below the Control-Plane aggregate network.max_query_result_bytes ceiling. Default: 512 MiB.

Trait Implementations

impl Clone for QueryTuning

fn clone(&self) -> QueryTuning

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for QueryTuning

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for QueryTuning

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for QueryTuning

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for QueryTuning

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.