Repository navigation
Reuse dictionary hashes and group mappings in aggregation - #26826
Draft
Rich-T-kid wants to merge 4 commits into
Draft
Rich-T-kid wants to merge 4 commits into
Rich-T-kid wants to merge 4 commits into
Conversation
Contributor
Author
|
run benchmark tpch tpcds clickbench_partitioned env:
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "e1aa5f4bed1c2880b1043b0d87502e5c984fa9cc" |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing e1aa5f4 (e1aa5f4) to 4ede583 diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "e1aa5f4bed1c2880b1043b0d87502e5c984fa9cc"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing e1aa5f4 (e1aa5f4) to 4ede583 diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "e1aa5f4bed1c2880b1043b0d87502e5c984fa9cc"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing e1aa5f4 (e1aa5f4) to 4ede583 diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "e1aa5f4bed1c2880b1043b0d87502e5c984fa9cc"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpch — base (merge-base)
tpch — branch
File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing e1aa5f4 (e1aa5f4) to 4ede583 diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "e1aa5f4bed1c2880b1043b0d87502e5c984fa9cc"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing e1aa5f4 (e1aa5f4) to 4ede583 diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "e1aa5f4bed1c2880b1043b0d87502e5c984fa9cc"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpcds — base (merge-base)
tpcds — branch
File an issue against this benchmark runner |
Rich-T-kid
force-pushed
the
rich-T-kid/aggregate-dictionary-hash-cache
branch
from
October 27, 2026 20:22
e1aa5f4 to
d49dd6d
Compare
Contributor
Author
|
run benchmark tpch tpcds clickbench_partitioned env:
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37" |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpch — base (merge-base)
tpch — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpcds — base (merge-base)
tpcds — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing e1aa5f4 (e1aa5f4) to 4ede583 diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "e1aa5f4bed1c2880b1043b0d87502e5c984fa9cc"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usageclickbench_partitioned — base (merge-base)
clickbench_partitioned — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usageclickbench_partitioned — base (merge-base)
clickbench_partitioned — branch
File an issue against this benchmark runner |
Contributor
Author
|
run benchmark dictionary_group_values env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37" |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark dictionary_group_values
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing d49dd6d (d49dd6d) to 4ede583 diff Run configurationrun benchmark dictionary_group_values
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "d49dd6d5614860d21328eb63633fd9d8e08fbd37"CPU Details (lscpu)Details
Resource Usagedictionary_group_values — base (merge-base)
dictionary_group_values — branch
File an issue against this benchmark runner |
Rich-T-kid
commented
Oct 27, 2026
Comment on lines
+368
to
+370
| fn create_hashes(&mut self, cols: &[ArrayRef], hashes: &mut [u64]) -> Result<()> { | ||
| if !self.has_dictionary { | ||
| create_hashes(cols, &self.random_state, hashes)?; |
Contributor
Author
There was a problem hiding this comment.
this gets re-used across all 3 PR's I think we should move it up into the hash util PR
Rich-T-kid
force-pushed
the
rich-T-kid/aggregate-dictionary-hash-cache
branch
from
October 27, 2026 23:09
d49dd6d to
0044519
Compare
Contributor
Author
|
run benchmark dictionary_group_values env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39" |
Contributor
Author
|
run benchmark tpch tpcds clickbench_partitioned env:
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39" |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 0044519 (0044519) to 4ede583 diff Run configurationrun benchmark dictionary_group_values
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 0044519 (0044519) to 4ede583 diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 0044519 (0044519) to 4ede583 diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 0044519 (0044519) to 4ede583 diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 0044519 (0044519) to 4ede583 diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpch — base (merge-base)
tpch — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 0044519 (0044519) to 4ede583 diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "4ede5833f68dad5bc8169492dedc5749fcc4b8d5"
changed:
ref: "0044519d3b609ae6fbb0cfc59500b322ea368d39"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpcds — base (merge-base)
tpcds — branch
File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 0676443 (0676443) to 9c32dba diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "9c32dbaaf42c30282abb4be294ea69aa9f53aa4d"
changed:
ref: "0676443827e3a102a9acd344e8e6a290943d989e"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 0676443 (0676443) to 9c32dba diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "9c32dbaaf42c30282abb4be294ea69aa9f53aa4d"
changed:
ref: "0676443827e3a102a9acd344e8e6a290943d989e"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpch — base (merge-base)
tpch — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 0676443 (0676443) to 9c32dba diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "9c32dbaaf42c30282abb4be294ea69aa9f53aa4d"
changed:
ref: "0676443827e3a102a9acd344e8e6a290943d989e"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpcds — base (merge-base)
tpcds — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 0676443 (0676443) to 9c32dba diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "9c32dbaaf42c30282abb4be294ea69aa9f53aa4d"
changed:
ref: "0676443827e3a102a9acd344e8e6a290943d989e"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usageclickbench_partitioned — base (merge-base)
clickbench_partitioned — branch
File an issue against this benchmark runner |
Reuse dictionary value hashes by column while the values Arc is unchanged, without retaining input dictionaries. Expose cache state for callers that account for memory and document the public hashing contract.
For a single non-streaming dictionary group key, the outer table already guarantees that appended values are distinct groups. Reuse the dictionary column's cached value-to-slot mapping as group indices so later batches can skip hashing and probing.
Rich-T-kid
force-pushed
the
rich-T-kid/aggregate-dictionary-hash-cache
branch
from
October 30, 2026 22:53
369ac6f to
f349d1f
Compare
Contributor
Author
|
run benchmark tpch tpcds clickbench_partitioned |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing f349d1f (f349d1f) to a48c3cf diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "f349d1fa1339952045759b1e1d60fc761ef6402c"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing f349d1f (f349d1f) to a48c3cf diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "f349d1fa1339952045759b1e1d60fc761ef6402c"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing f349d1f (f349d1f) to a48c3cf diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "f349d1fa1339952045759b1e1d60fc761ef6402c"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing f349d1f (f349d1f) to a48c3cf diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "f349d1fa1339952045759b1e1d60fc761ef6402c"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpch — base (merge-base)
tpch — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing f349d1f (f349d1f) to a48c3cf diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "f349d1fa1339952045759b1e1d60fc761ef6402c"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpcds — base (merge-base)
tpcds — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing f349d1f (f349d1f) to a48c3cf diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "f349d1fa1339952045759b1e1d60fc761ef6402c"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usageclickbench_partitioned — base (merge-base)
clickbench_partitioned — branch
File an issue against this benchmark runner |
Contributor
Author
|
run benchmark tpch tpcds clickbench_partitioned |
Use the caching hasher only for non-streaming group schemas that contain dictionary columns, and share cached hashes only with those columns. Streaming and non-dictionary aggregates keep the existing hashing path without per-batch cache plumbing.
Rich-T-kid
force-pushed
the
rich-T-kid/aggregate-dictionary-hash-cache
branch
from
October 30, 2026 23:32
d14a173 to
409c45d
Compare
Contributor
Author
|
run benchmark tpch tpcds clickbench_partitioned |
|
Benchmark for this request failed before finishing (Kubernetes reason: Benchmarks requested: Runner log (last 40 lines)Kubernetes messageFile an issue against this benchmark runner |
|
Benchmark for this request failed before finishing (Kubernetes reason: Benchmarks requested: Runner log (last 40 lines)Kubernetes messageFile an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing d14a173 (d14a173) to a48c3cf diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "d14a173971f4a0f451dc5e253ad3bfc61e6011c6"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 409c45d (409c45d) to a48c3cf diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "409c45d9576334f69cab2c09fa70d68ff64db1e3"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 409c45d (409c45d) to a48c3cf diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "409c45d9576334f69cab2c09fa70d68ff64db1e3"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark running (GKE) | trigger CPU Details (lscpu)Comparing 409c45d (409c45d) to a48c3cf diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "409c45d9576334f69cab2c09fa70d68ff64db1e3"Results will be posted here when complete File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing d14a173 (d14a173) to a48c3cf diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "d14a173971f4a0f451dc5e253ad3bfc61e6011c6"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpch — base (merge-base)
tpch — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 409c45d (409c45d) to a48c3cf diff Run configurationrun benchmark tpch
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "409c45d9576334f69cab2c09fa70d68ff64db1e3"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpch — base (merge-base)
tpch — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 409c45d (409c45d) to a48c3cf diff Run configurationrun benchmark tpcds
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "409c45d9576334f69cab2c09fa70d68ff64db1e3"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usagetpcds — base (merge-base)
tpcds — branch
File an issue against this benchmark runner |
|
🤖 Benchmark completed (GKE) | trigger Instance: Comparing 409c45d (409c45d) to a48c3cf diff Run configurationrun benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true"
DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G"
baseline:
ref: "a48c3cf5aa944823e58f683839253a3aef96193b"
changed:
ref: "409c45d9576334f69cab2c09fa70d68ff64db1e3"CPU Details (lscpu)Details
Memory Pool PeaksPeak Base:
Pool accounting vs. process RSS Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it.
Resource Usageclickbench_partitioned — base (merge-base)
clickbench_partitioned — branch
File an issue against this benchmark runner |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Which issue does this PR close?
Closes #26812.
Closes #26813.
Depends on #26817.
Rationale for this change
Dictionary group keys were hashing the same values twice: once for the aggregate group table and again in
DictionaryGroupValuesColumn. A sole non-streaming dictionary key also repeated its inner value-dedup lookup, and later batches sharing the same values array still probed the outer group table row by row.For a sole group key, the outer table already guarantees that every appended value is a distinct group. That makes the dictionary inner slot identical to the aggregate group index.
What changes are included in this PR?
DictionaryGroupValuesColumn.Arcand every referenced key is known, fill group indices directly without hashing, probing, equality checks, or appends.Results
The focused Criterion benchmark compares #26817's exact head
a48c3cf5with this PR at409c45d95. Each case processes 32 batches of 8,192 rows sharing one dictionary values array.Hosted benchmarks compare #26817's exact head with this PR and set
datafusion.execution.parquet.enable_rle_to_dictionary=trueon both sides:ClickBench improves Q5, Q12, Q13, Q34, Q36, and Q39 by 6–10%. Across the three suites, 12 queries improve with no reported regressions.
What is the testing strategy for this PR?
The focused test covers shared values, a reordered replacement values array, null keys and null values, the direct mapping path, and invalidation after partial emit.
multi_group_bytests with forced hash collisionscargo fmt --all -- --checkcargo clippy --all-targets --all-features -- -D warningsAre there any user-facing changes?
No.