Runtime Stats

Runtime stats are used to collect the per-query velox runtime events for offline query analysis purpose. The collected stats can provide insights into the operator level query execution internals, such as how much time a query operator spent in disk spilling. The collected stats are organized in a free-form key-value for easy extension. The key is the event name and the value is defined as RuntimeCounter which is used to store and aggregate a particular event occurrences during the operator execution. RuntimeCounter has three types: kNone used to record event count, kNanos used to record event time in nanoseconds and kBytes used to record memory or storage size in bytes. It records the count of events, and the min/max/sum of the event values. The stats are stored in OperatorStats structure. The query system can aggregate the operator level stats collected from each driver by pipeline and task for analysis.

Memory Arbitration

These stats are reported by all operators.

Stats	Unit	Description
memoryReclaimCount		The number of times that the memory arbitration to reclaim memory from an spillable operator. This stats only applies for spillable operators.
memoryReclaimWallNanos	nanos	The memory reclaim execution time of an operator during the memory arbitration. It collects time spent on disk spilling or file write. This stats only applies for spillable operators.
reclaimedMemoryBytes	bytes	The reclaimed memory bytes of an operator during the memory arbitration. This stats only applies for spillable operators.
globalArbitrationCount		The number of times a request for more memory hit the arbitrator’s capacity limit and initiated a global arbitration attempt where memory is reclaimed from viable candidates chosen among all running queries based on a criterion.
localArbitrationCount		The number of times a request for more memory hit the query memory limit and initiated a local arbitration attempt where memory is reclaimed from the requestor itself.

HashBuild, HashAggregation

These stats are reported only by HashBuild and HashAggregation operators.

Stats	Unit	Description
hashtable.capacity		Number of slots across all buckets in the hash table.
hashtable.numRehashes		Number of rehash() calls.
hashtable.numDistinct		Number of distinct keys in the hash table.
hashtable.numTombstones		Number of tombstone slots in the hash table.
hashtable.buildWallNanos	nanos	Time spent on building the hash table from rows collected by all the hash build operators. This stat is only reported by the HashBuild operator.

TableWriter

These stats are reported only by TableWriter operator

Stats	Unit	Description
earlyFlushedRawBytes	bytes	Number of bytes pre-maturely flushed from file writers because of memory reclaiming.

Spilling

These stats are reported by operators that support spilling.

Stats	Unit	Description
spillFillTime	microseconds	The time spent on filling rows for spilling.
spillSortTime	microseconds	The time spent on sorting rows for spilling.
spillSerializationTime	microseconds	The time spent on serializing rows for spilling.
spillFlushTime	microseconds	The time spent on copy out serialized rows for disk write. If compression is enabled, this includes the compression time.
spillWrites		The number of spill writer flushes, equivalent to number of write calls to underlying filesystem.
spillWriteTime	microseconds	The time spent on writing spilled rows to disk.
spillRuns		The number of times that spilling runs on an operator.
exceededMaxSpillLevel		The number of times that an operator exceeds the max spill limit.
spillReadBytes	bytes	The number of bytes read from spilled files.
spillReads		The number of spill reader reads, equivalent to the number of read calls to the underlying filesystem.
spillReadTimeUs	microseconds	The time spent on read data from spilled files.
spillDeserializationTimeUs	microseconds	The time spent on deserializing rows read from spilled files.