You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@arrow.apache.org by "sdf-jkl (via GitHub)" <gi...@apache.org> on 2026/10/08 18:00:11 UTC
[PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
sdf-jkl opened a new pull request, #11460:
URL: https://github.com/apache/arrow-rs/pull/11460
## Which issue does this PR close?
None. This is a sibling experiment to @HippoBaro’s [#11449](https://github.com/apache/arrow-rs/pull/11449).
## Rationale for this change
#11449 reports substantial execution-order-dependent variance in Arrow writer benchmarks and addresses it using jemalloc with page purging disabled.
This PR explores mimalloc as an alternative, consistent with the allocator used by DataFusion’s CLI and benchmark harness. We’ll compare default mimalloc against mimalloc with purging disabled to investigate performance and benchmark stability.
## What changes are included in this PR?
Use mimalloc as the global allocator for the Linux `arrow_writer` benchmark executable. The dependency is development-only; library consumers are unaffected.
Purging remains configurable through `MIMALLOC_PURGE_DELAY`. Two bot runs with four shards each are planned:
- Default mimalloc settings.
- `MIMALLOC_PURGE_DELAY=-1` on the PR side, retaining freed pages for reuse.
Both runs will compare against the same baseline. These initial comparisons will help assess performance; repeated runs are needed to establish whether variance improves.
## Are these changes tested?
- Formatting and compilation with the bot’s benchmark features passed.
- The `primitive/default` smoke test passed with `MIMALLOC_PURGE_DELAY=-1`; allocator diagnostics confirmed purging was disabled.
- Bot results pending.
## Are there any significant user-facing changes?
None.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "HippoBaro (via GitHub)" <gi...@apache.org>.
HippoBaro commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6066993550
@sdf-jkl Thanks for picking this up! I think evaluating `mimalloc` makes sense if it aligns us better with what DataFusion does.
The benchmark bot may not be the best environment to evaluate whether your changes eliminate the order-dependent measurement artifacts, though. I’ll try to find some time to benchmark this on my end and report whether I can still reproduce the large performance drift I otherwise observe on `main`. Might not happen this week however!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6068969422
run benchmark arrow_writer
shards: 4
baseline:
ref: https://github.com/apache/arrow-rs/pull/11460/commits/b1c118bd52655725911d374aff4d1a0d0b85a4c9
changed:
ref: https://github.com/apache/arrow-rs/pull/11460/commits/b1c118bd52655725911d374aff4d1a0d0b85a4c9
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "HippoBaro (via GitHub)" <gi...@apache.org>.
HippoBaro commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6206877385
Thank you @sdf-jkl!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065966059
run benchmark arrow_writer
shards: 4
baseline:
ref: "7fc78ca9ecf907de396119a25bc689527f0fd0ea"
changed:
ref: "b1c118bd52655725911d374aff4d1a0d0b85a4c9"
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6201338306
lgtm
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6215738765
Thanks everyone :pray:
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6066767540
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065966059)
**Target:** arrow_writer
Comparing b1c118bd52655725911d374aff4d1a0d0b85a4c9 (b1c118bd52655725911d374aff4d1a0d0b85a4c9) to 7fc78ca9ecf907de396119a25bc689527f0fd0ea [diff](https://github.com/apache/arrow-rs/compare/7fc78ca9ecf907de396119a25bc689527f0fd0ea..b1c118bd52655725911d374aff4d1a0d0b85a4c9)
<details><summary>Run configuration</summary>
```yaml
run benchmark arrow_writer
baseline:
ref: 7fc78ca9ecf907de396119a25bc689527f0fd0ea
changed:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
shards: 4
```
</details>
<details><summary>Details</summary>
```
group base changed
----- ---- -------
bool/bloom_filter 1.00 11.1±0.05ms 22.6 MB/sec 1.00 11.1±0.05ms 22.5 MB/sec
bool/cdc 1.00 13.4±0.13ms 18.7 MB/sec 1.01 13.6±0.11ms 18.4 MB/sec
bool/default 1.02 9.0±0.10ms 27.7 MB/sec 1.00 8.8±0.02ms 28.3 MB/sec
bool/number_distinct_values 1.01 19.1±0.21ms 13.1 MB/sec 1.00 18.8±0.03ms 13.3 MB/sec
bool/parquet_2 1.00 10.8±0.04ms 23.2 MB/sec 1.02 11.0±0.07ms 22.6 MB/sec
bool/zstd 1.01 9.6±0.11ms 26.2 MB/sec 1.00 9.4±0.03ms 26.5 MB/sec
bool/zstd_parquet_2 1.00 11.2±0.05ms 22.3 MB/sec 1.02 11.5±0.08ms 21.8 MB/sec
bool_non_null/bloom_filter 1.00 5.2±0.02ms 24.3 MB/sec 1.00 5.2±0.04ms 24.2 MB/sec
bool_non_null/cdc 1.00 4.1±0.04ms 30.7 MB/sec 1.00 4.1±0.04ms 30.8 MB/sec
bool_non_null/default 1.00 2.3±0.02ms 54.4 MB/sec 1.01 2.3±0.02ms 54.1 MB/sec
bool_non_null/number_distinct_values 1.00 15.7±0.12ms 8.0 MB/sec 1.00 15.6±0.04ms 8.0 MB/sec
bool_non_null/parquet_2 1.00 4.6±0.01ms 27.1 MB/sec 1.00 4.6±0.03ms 27.1 MB/sec
bool_non_null/zstd 1.00 2.6±0.01ms 47.7 MB/sec 1.01 2.7±0.01ms 47.1 MB/sec
bool_non_null/zstd_parquet_2 1.00 5.0±0.01ms 24.8 MB/sec 1.00 5.0±0.02ms 24.9 MB/sec
bool_ree/bloom_filter 1.00 27.8±0.09ms 7.7 MB/sec 1.01 28.1±0.27ms 7.6 MB/sec
bool_ree/cdc 1.00 28.5±0.10ms 7.5 MB/sec 1.02 29.1±0.42ms 7.3 MB/sec
bool_ree/default 1.01 25.3±0.13ms 8.4 MB/sec 1.00 25.1±0.15ms 8.5 MB/sec
bool_ree/number_distinct_values 1.00 37.3±0.07ms 5.7 MB/sec 1.01 37.6±0.22ms 5.7 MB/sec
bool_ree/parquet_2 1.00 25.9±0.06ms 8.2 MB/sec 1.01 26.2±0.35ms 8.1 MB/sec
bool_ree/zstd 1.00 25.7±0.05ms 8.3 MB/sec 1.01 25.9±0.30ms 8.2 MB/sec
bool_ree/zstd_parquet_2 1.01 26.5±0.14ms 8.0 MB/sec 1.00 26.3±0.29ms 8.1 MB/sec
decimal/bloom_filter 1.20 171.0±2.35ms 351.0 MB/sec 1.00 142.0±7.84ms 422.5 MB/sec
decimal/cdc 1.00 104.4±0.59ms 574.8 MB/sec 1.02 106.0±2.68ms 565.9 MB/sec
decimal/default 1.00 84.8±0.98ms 707.8 MB/sec 1.02 86.6±4.25ms 692.7 MB/sec
decimal/number_distinct_values 1.00 318.5±2.08ms 188.4 MB/sec 1.07 339.7±43.44ms 176.7 MB/sec
decimal/parquet_2 1.20 207.9±3.13ms 288.6 MB/sec 1.00 173.7±5.87ms 345.4 MB/sec
decimal/zstd 1.39 135.5±1.18ms 442.8 MB/sec 1.00 97.6±4.37ms 614.8 MB/sec
decimal/zstd_parquet_2 1.17 230.8±2.97ms 260.0 MB/sec 1.00 197.6±8.96ms 303.7 MB/sec
fixed_size_binary_ree/bloom_filter 1.00 63.9±0.40ms 15.8 MB/sec 1.00 63.9±1.03ms 15.8 MB/sec
fixed_size_binary_ree/cdc 1.00 72.3±0.41ms 13.9 MB/sec 1.00 72.7±1.57ms 13.9 MB/sec
fixed_size_binary_ree/default 1.00 65.2±1.77ms 15.5 MB/sec 1.02 66.5±3.94ms 15.1 MB/sec
fixed_size_binary_ree/number_distinct_values 1.00 80.5±0.38ms 12.5 MB/sec 1.01 81.6±3.36ms 12.3 MB/sec
fixed_size_binary_ree/parquet_2 1.00 62.6±0.93ms 16.1 MB/sec 1.01 63.2±3.16ms 15.9 MB/sec
fixed_size_binary_ree/zstd 1.00 62.8±0.36ms 16.0 MB/sec 1.01 63.3±0.60ms 15.9 MB/sec
fixed_size_binary_ree/zstd_parquet_2 1.00 63.3±0.92ms 15.9 MB/sec 1.00 63.6±2.79ms 15.8 MB/sec
float_with_nans/bloom_filter 1.09 89.9±1.82ms 155.7 MB/sec 1.00 82.6±2.55ms 169.5 MB/sec
float_with_nans/cdc 1.00 78.8±0.92ms 177.8 MB/sec 1.02 80.4±1.78ms 174.2 MB/sec
float_with_nans/default 1.00 70.0±0.95ms 200.0 MB/sec 1.03 71.9±2.22ms 194.8 MB/sec
float_with_nans/number_distinct_values 1.00 169.6±0.38ms 82.5 MB/sec 1.06 179.6±10.67ms 77.9 MB/sec
float_with_nans/parquet_2 1.00 73.0±0.78ms 191.7 MB/sec 1.01 73.6±1.28ms 190.3 MB/sec
float_with_nans/zstd 1.00 107.4±1.02ms 130.4 MB/sec 1.02 109.6±2.17ms 127.7 MB/sec
float_with_nans/zstd_parquet_2 1.07 118.0±2.02ms 118.7 MB/sec 1.00 110.0±2.80ms 127.2 MB/sec
fsb/bloom_filter 1.00 13.9±0.35ms 1161.5 MB/sec 1.05 14.5±0.64ms 1110.3 MB/sec
fsb/cdc 1.00 16.6±0.22ms 969.6 MB/sec 1.01 16.9±0.31ms 957.0 MB/sec
fsb/default 1.00 11.1±0.08ms 1446.7 MB/sec 1.00 11.1±0.12ms 1450.5 MB/sec
fsb/number_distinct_values 1.00 15.5±0.12ms 1041.0 MB/sec 1.01 15.6±0.63ms 1031.2 MB/sec
fsb/parquet_2 1.00 13.4±0.16ms 1201.2 MB/sec 1.07 14.3±0.29ms 1125.7 MB/sec
fsb/zstd 1.00 14.1±0.37ms 1142.2 MB/sec 1.03 14.5±0.47ms 1112.0 MB/sec
fsb/zstd_parquet_2 1.01 13.8±0.11ms 1165.2 MB/sec 1.00 13.7±0.15ms 1178.1 MB/sec
int32_ree/bloom_filter 1.01 35.0±0.20ms 11.6 MB/sec 1.00 34.7±0.26ms 11.7 MB/sec
int32_ree/cdc 1.00 37.3±0.04ms 10.9 MB/sec 1.01 37.8±0.40ms 10.8 MB/sec
int32_ree/default 1.01 34.6±0.21ms 11.8 MB/sec 1.00 34.1±0.33ms 11.9 MB/sec
int32_ree/number_distinct_values 1.00 50.2±0.11ms 8.1 MB/sec 1.01 50.9±0.49ms 8.0 MB/sec
int32_ree/parquet_2 1.00 33.6±0.07ms 12.1 MB/sec 1.01 34.1±0.34ms 11.9 MB/sec
int32_ree/zstd 1.00 34.2±0.07ms 11.9 MB/sec 1.01 34.5±0.47ms 11.8 MB/sec
int32_ree/zstd_parquet_2 1.02 35.2±0.21ms 11.6 MB/sec 1.00 34.5±0.25ms 11.8 MB/sec
int32_ree_95pct_null/bloom_filter 1.00 21.4±0.03ms 19.0 MB/sec 1.00 21.4±0.10ms 19.0 MB/sec
int32_ree_95pct_null/cdc 1.00 23.4±0.10ms 17.3 MB/sec 1.03 24.1±0.10ms 16.9 MB/sec
int32_ree_95pct_null/default 1.00 21.2±0.04ms 19.1 MB/sec 1.00 21.2±0.13ms 19.2 MB/sec
int32_ree_95pct_null/number_distinct_values 1.00 22.0±0.02ms 18.5 MB/sec 1.00 22.0±0.05ms 18.5 MB/sec
int32_ree_95pct_null/parquet_2 1.00 21.3±0.06ms 19.1 MB/sec 1.00 21.3±0.10ms 19.1 MB/sec
int32_ree_95pct_null/zstd 1.00 21.3±0.03ms 19.1 MB/sec 1.01 21.4±0.06ms 19.0 MB/sec
int32_ree_95pct_null/zstd_parquet_2 1.00 21.3±0.04ms 19.1 MB/sec 1.00 21.4±0.06ms 19.1 MB/sec
large_string_distinct/delta_byte_array 1.00 72.8±4.31ms 3.4 GB/sec 1.07 77.7±5.02ms 3.2 GB/sec
large_string_distinct/plain 1.04 61.5±0.83ms 4.1 GB/sec 1.00 59.0±5.41ms 4.2 GB/sec
large_string_distinct_nullable/delta_byte_array 1.01 71.2±1.96ms 3.5 GB/sec 1.00 70.7±5.78ms 3.5 GB/sec
large_string_non_null/bloom_filter 1.07 54.8±1.04ms 4.6 GB/sec 1.00 51.4±1.26ms 4.9 GB/sec
large_string_non_null/cdc 1.02 243.0±1.41ms 1053.5 MB/sec 1.00 238.6±1.08ms 1073.0 MB/sec
large_string_non_null/default 1.00 51.8±0.07ms 4.8 GB/sec 1.00 52.0±0.67ms 4.8 GB/sec
large_string_non_null/number_distinct_values 1.04 70.5±1.03ms 3.5 GB/sec 1.00 67.7±0.60ms 3.7 GB/sec
large_string_non_null/parquet_2 1.00 51.8±0.06ms 4.8 GB/sec 1.00 52.0±0.76ms 4.8 GB/sec
large_string_non_null/zstd 1.03 53.2±0.87ms 4.7 GB/sec 1.00 51.5±0.78ms 4.9 GB/sec
large_string_non_null/zstd_parquet_2 1.07 54.3±0.98ms 4.6 GB/sec 1.00 50.9±0.52ms 4.9 GB/sec
large_string_shared_prefix/delta_byte_array 1.03 43.6±0.15ms 5.7 GB/sec 1.00 42.3±0.27ms 5.9 GB/sec
large_string_shared_prefix/plain 1.05 95.2±0.68ms 2.6 GB/sec 1.00 90.8±5.22ms 2.8 GB/sec
large_string_shared_prefix_list/delta_byte_array 1.00 55.5±0.36ms 4.5 GB/sec 1.00 55.4±3.07ms 4.5 GB/sec
large_string_shared_prefix_nullable/delta_byte_array 1.00 42.8±1.16ms 5.8 GB/sec 1.00 42.9±0.73ms 5.8 GB/sec
large_string_shared_prefix_nullable/plain 1.00 66.2±0.52ms 3.8 GB/sec 1.03 67.9±5.82ms 3.7 GB/sec
large_string_shared_prefix_nullable_dense/delta_byte_array 1.00 21.8±0.04ms 5.7 GB/sec 1.01 22.0±0.21ms 5.7 GB/sec
large_string_shared_prefix_nullable_trailing/delta_byte_array 1.01 45.2±1.67ms 5.5 GB/sec 1.00 44.8±1.72ms 5.6 GB/sec
list_nested/bloom_filter 1.00 132.2±0.19ms 218.3 MB/sec 1.07 141.3±5.03ms 204.3 MB/sec
list_nested/cdc 1.00 153.0±0.91ms 188.6 MB/sec 1.01 155.0±2.92ms 186.2 MB/sec
list_nested/default 1.15 142.4±0.46ms 202.6 MB/sec 1.00 123.4±3.83ms 233.9 MB/sec
list_nested/number_distinct_values 1.04 323.2±7.45ms 89.3 MB/sec 1.00 310.5±12.65ms 92.9 MB/sec
list_nested/parquet_2 1.00 129.2±0.16ms 223.3 MB/sec 1.02 132.1±3.01ms 218.5 MB/sec
list_nested/zstd 1.00 129.9±0.68ms 222.1 MB/sec 1.03 133.7±4.41ms 215.9 MB/sec
list_nested/zstd_parquet_2 1.13 166.0±0.68ms 173.9 MB/sec 1.00 147.4±3.50ms 195.8 MB/sec
list_primitive/bloom_filter 1.00 300.8±18.44ms 1813.2 MB/sec 1.02 305.6±21.87ms 1784.6 MB/sec
list_primitive/cdc 1.00 298.3±1.27ms 1828.2 MB/sec 1.06 317.4±20.88ms 1718.0 MB/sec
list_primitive/default 1.00 191.7±1.48ms 2.8 GB/sec 1.07 205.2±16.88ms 2.6 GB/sec
list_primitive/number_distinct_values 1.00 403.2±8.55ms 1352.5 MB/sec 1.07 433.0±27.57ms 1259.6 MB/sec
list_primitive/parquet_2 1.00 205.0±1.21ms 2.6 GB/sec 1.08 220.9±19.80ms 2.4 GB/sec
list_primitive/zstd 1.00 460.2±5.32ms 1185.0 MB/sec 1.10 507.7±36.18ms 1074.3 MB/sec
list_primitive/zstd_parquet_2 1.00 432.3±1.86ms 1261.5 MB/sec 1.04 450.3±17.40ms 1211.1 MB/sec
list_primitive_non_null/bloom_filter 1.00 388.7±17.91ms 1400.1 MB/sec 1.06 410.4±42.61ms 1326.1 MB/sec
list_primitive_non_null/cdc 1.00 376.9±7.33ms 1444.2 MB/sec 1.11 418.3±31.48ms 1301.2 MB/sec
list_primitive_non_null/default 1.10 229.9±2.81ms 2.3 GB/sec 1.00 208.6±25.96ms 2.5 GB/sec
list_primitive_non_null/number_distinct_values 1.08 656.1±40.43ms 829.5 MB/sec 1.00 609.0±24.91ms 893.7 MB/sec
list_primitive_non_null/parquet_2 1.00 230.2±6.27ms 2.3 GB/sec 1.06 245.2±41.00ms 2.2 GB/sec
list_primitive_non_null/zstd 1.00 681.9±3.73ms 798.1 MB/sec 1.03 701.1±37.84ms 776.3 MB/sec
list_primitive_non_null/zstd_parquet_2 1.04 644.7±18.69ms 844.2 MB/sec 1.00 621.5±8.37ms 875.7 MB/sec
list_primitive_sparse_99pct_null/bloom_filter 1.00 13.5±0.11ms 2.7 GB/sec 1.01 13.6±0.19ms 2.7 GB/sec
list_primitive_sparse_99pct_null/cdc 1.00 24.1±0.07ms 1551.4 MB/sec 1.00 24.0±0.15ms 1555.8 MB/sec
list_primitive_sparse_99pct_null/default 1.00 12.9±0.11ms 2.8 GB/sec 1.00 13.0±0.16ms 2.8 GB/sec
list_primitive_sparse_99pct_null/number_distinct_values 1.00 13.2±0.08ms 2.8 GB/sec 1.00 13.2±0.16ms 2.8 GB/sec
list_primitive_sparse_99pct_null/parquet_2 1.00 12.7±0.10ms 2.9 GB/sec 1.00 12.7±0.14ms 2.9 GB/sec
list_primitive_sparse_99pct_null/zstd 1.00 14.9±0.09ms 2.4 GB/sec 1.00 14.9±0.17ms 2.4 GB/sec
list_primitive_sparse_99pct_null/zstd_parquet_2 1.00 12.5±0.03ms 2.9 GB/sec 1.00 12.5±0.10ms 2.9 GB/sec
list_struct_with_list/bloom_filter 1.00 255.6±0.61ms 178.4 MB/sec 1.07 272.3±6.95ms 167.5 MB/sec
list_struct_with_list/cdc 1.00 300.7±1.18ms 151.6 MB/sec 1.04 313.3±10.66ms 145.5 MB/sec
list_struct_with_list/default 1.00 234.4±1.25ms 194.5 MB/sec 1.04 243.4±7.65ms 187.4 MB/sec
list_struct_with_list/number_distinct_values 1.00 549.2±11.93ms 83.0 MB/sec 1.04 569.8±15.63ms 80.0 MB/sec
list_struct_with_list/parquet_2 1.00 244.3±0.41ms 186.6 MB/sec 1.04 254.3±11.58ms 179.3 MB/sec
list_struct_with_list/zstd 1.00 252.4±0.37ms 180.7 MB/sec 1.05 263.8±6.30ms 172.8 MB/sec
list_struct_with_list/zstd_parquet_2 1.00 266.8±3.13ms 170.9 MB/sec 1.04 277.3±9.05ms 164.4 MB/sec
medium_string_shared_prefix_nullable/delta_byte_array 1.00 38.1±0.63ms 6.6 GB/sec 1.12 42.8±0.76ms 5.8 GB/sec
ndv/bool/ndv_on 1.00 9.8±0.02ms 8.1 MB/sec 1.01 9.8±0.10ms 8.1 MB/sec
ndv/fsb16_high_cardinality/ndv_on 1.00 42.5±1.45ms 235.1 MB/sec 1.02 43.4±1.05ms 230.7 MB/sec
ndv/fsb16_low_cardinality/ndv_on 1.00 35.1±0.20ms 285.1 MB/sec 1.00 35.2±0.28ms 284.2 MB/sec
ndv/int32_high_cardinality/ndv_on 1.00 28.8±0.60ms 86.9 MB/sec 1.05 30.2±0.80ms 82.8 MB/sec
ndv/int32_low_cardinality/ndv_on 1.00 16.2±0.06ms 154.4 MB/sec 1.00 16.2±0.05ms 154.6 MB/sec
ndv/large_binary_high_cardinality/ndv_on 1.00 37.7±0.08ms 198.7 MB/sec 1.00 37.7±0.11ms 198.9 MB/sec
ndv/large_binary_low_cardinality/ndv_on 1.00 24.9±0.09ms 601.3 MB/sec 1.02 25.4±0.11ms 591.2 MB/sec
ndv/utf8_high_cardinality/ndv_on 1.00 79.2±3.92ms 2.0 GB/sec 1.05 83.1±5.61ms 1955.1 MB/sec
ndv/utf8_low_cardinality/ndv_on 1.00 23.8±0.07ms 315.3 MB/sec 1.02 24.2±0.08ms 309.7 MB/sec
ndv/utf8_view_high_cardinality/ndv_on 1.00 41.2±2.17ms 242.5 MB/sec 1.06 43.6±2.42ms 229.4 MB/sec
ndv/utf8_view_low_cardinality/ndv_on 1.00 23.6±0.08ms 423.3 MB/sec 1.00 23.7±0.08ms 422.3 MB/sec
primitive/bloom_filter 1.00 127.6±0.61ms 351.6 MB/sec 1.06 135.7±2.20ms 330.7 MB/sec
primitive/cdc 1.00 148.8±1.22ms 301.6 MB/sec 1.02 152.3±2.45ms 294.7 MB/sec
primitive/default 1.00 107.0±0.91ms 419.4 MB/sec 1.02 109.4±2.45ms 410.2 MB/sec
primitive/number_distinct_values 1.01 343.1±4.28ms 130.8 MB/sec 1.00 339.7±5.97ms 132.1 MB/sec
primitive/parquet_2 1.00 116.0±0.43ms 386.7 MB/sec 1.02 118.8±2.09ms 377.7 MB/sec
primitive/zstd 1.00 138.1±1.32ms 325.0 MB/sec 1.04 143.9±3.08ms 311.9 MB/sec
primitive/zstd_parquet_2 1.00 151.0±0.96ms 297.3 MB/sec 1.03 155.4±2.43ms 288.8 MB/sec
primitive_all_null/bloom_filter 1.10 905.7±5.94µs 48.4 GB/sec 1.00 826.7±5.03µs 53.0 GB/sec
primitive_all_null/cdc 1.02 19.0±0.30ms 2.3 GB/sec 1.00 18.7±0.29ms 2.3 GB/sec
primitive_all_null/default 1.29 270.7±1.80µs 161.9 GB/sec 1.00 210.3±1.63µs 208.3 GB/sec
primitive_all_null/number_distinct_values 1.28 268.8±1.84µs 163.0 GB/sec 1.00 210.2±1.45µs 208.5 GB/sec
primitive_all_null/parquet_2 1.28 275.4±2.31µs 159.1 GB/sec 1.00 214.8±2.48µs 204.0 GB/sec
primitive_all_null/zstd 1.31 419.7±1.94µs 104.4 GB/sec 1.00 319.3±2.16µs 137.2 GB/sec
primitive_all_null/zstd_parquet_2 1.31 374.2±2.26µs 117.1 GB/sec 1.00 286.1±2.39µs 153.2 GB/sec
primitive_non_null/bloom_filter 1.00 89.7±0.36ms 490.4 MB/sec 1.12 100.2±2.12ms 439.2 MB/sec
primitive_non_null/cdc 1.00 78.3±0.42ms 561.8 MB/sec 1.03 80.9±1.53ms 543.6 MB/sec
primitive_non_null/default 1.00 57.7±0.69ms 762.2 MB/sec 1.05 60.4±1.99ms 728.9 MB/sec
primitive_non_null/number_distinct_values 1.11 503.2±10.71ms 87.4 MB/sec 1.00 455.0±8.63ms 96.7 MB/sec
primitive_non_null/parquet_2 1.00 71.7±0.20ms 614.0 MB/sec 1.03 73.9±1.64ms 595.1 MB/sec
primitive_non_null/zstd 1.00 90.4±0.74ms 486.7 MB/sec 1.06 95.6±2.54ms 460.2 MB/sec
primitive_non_null/zstd_parquet_2 1.00 108.8±1.34ms 404.6 MB/sec 1.04 113.2±2.96ms 388.6 MB/sec
primitive_sparse_99pct_null/bloom_filter 1.00 12.6±0.39ms 3.5 GB/sec 1.00 12.6±0.20ms 3.5 GB/sec
primitive_sparse_99pct_null/cdc 1.00 30.9±0.37ms 1454.0 MB/sec 1.06 32.9±0.49ms 1365.9 MB/sec
primitive_sparse_99pct_null/default 1.01 11.3±0.10ms 3.9 GB/sec 1.00 11.2±0.07ms 3.9 GB/sec
primitive_sparse_99pct_null/number_distinct_values 1.00 13.8±0.04ms 3.2 GB/sec 1.00 13.8±0.06ms 3.2 GB/sec
primitive_sparse_99pct_null/parquet_2 1.01 11.2±0.16ms 3.9 GB/sec 1.00 11.0±0.10ms 4.0 GB/sec
primitive_sparse_99pct_null/zstd 1.00 14.5±0.26ms 3.0 GB/sec 1.02 14.8±0.11ms 3.0 GB/sec
primitive_sparse_99pct_null/zstd_parquet_2 1.00 13.4±0.09ms 3.3 GB/sec 1.01 13.5±0.11ms 3.3 GB/sec
short_string_non_null/bloom_filter 1.00 26.3±0.08ms 456.7 MB/sec 1.02 26.9±0.34ms 446.3 MB/sec
short_string_non_null/cdc 1.00 19.5±0.04ms 616.4 MB/sec 1.03 20.1±0.20ms 596.6 MB/sec
short_string_non_null/default 1.00 15.5±0.04ms 773.1 MB/sec 1.03 16.0±0.24ms 749.4 MB/sec
short_string_non_null/number_distinct_values 1.00 69.9±0.47ms 171.7 MB/sec 1.10 76.9±9.68ms 156.0 MB/sec
short_string_non_null/parquet_2 1.00 24.7±0.03ms 485.9 MB/sec 1.03 25.4±0.15ms 471.7 MB/sec
short_string_non_null/zstd 1.00 35.9±0.15ms 333.9 MB/sec 1.03 37.0±0.38ms 324.4 MB/sec
short_string_non_null/zstd_parquet_2 1.00 27.6±0.05ms 434.2 MB/sec 1.03 28.3±0.15ms 423.3 MB/sec
small_string_distinct/delta_byte_array 1.02 1129.0±51.98µs 6.9 GB/sec 1.00 1111.3±63.12µs 7.1 GB/sec
small_string_distinct/plain 1.00 876.3±26.17µs 9.0 GB/sec 1.00 876.8±33.89µs 8.9 GB/sec
small_string_partial_prefix/delta_byte_array 1.03 1298.2±92.04µs 6.0 GB/sec 1.00 1261.0±74.58µs 6.2 GB/sec
small_string_partial_prefix/plain 1.01 1286.4±85.44µs 6.1 GB/sec 1.00 1271.0±72.68µs 6.2 GB/sec
small_string_shared_prefix/delta_byte_array 1.00 1012.6±14.94µs 7.7 GB/sec 1.02 1028.4±12.86µs 7.6 GB/sec
small_string_shared_prefix/plain 1.00 1035.9±5.26µs 7.6 GB/sec 1.12 1157.0±10.19µs 6.8 GB/sec
string/bloom_filter 1.22 220.5±24.29ms 2.3 GB/sec 1.00 180.4±15.23ms 2.8 GB/sec
string/cdc 1.00 208.1±7.64ms 2.5 GB/sec 1.01 210.7±15.76ms 2.4 GB/sec
string/default 1.57 136.1±24.41ms 3.8 GB/sec 1.00 86.8±7.38ms 5.9 GB/sec
string/number_distinct_values 1.05 223.6±4.49ms 2.3 GB/sec 1.00 213.0±25.16ms 2.4 GB/sec
string/parquet_2 1.72 168.3±2.34ms 3.0 GB/sec 1.00 97.7±3.69ms 5.2 GB/sec
string/zstd 1.05 463.7±10.44ms 1130.6 MB/sec 1.00 443.7±22.51ms 1181.6 MB/sec
string/zstd_parquet_2 1.00 383.9±7.63ms 1365.4 MB/sec 1.05 402.2±29.17ms 1303.5 MB/sec
string_and_binary_view/bloom_filter 1.00 66.3±1.31ms 486.2 MB/sec 1.02 68.0±1.68ms 474.3 MB/sec
string_and_binary_view/cdc 1.00 62.0±1.34ms 520.3 MB/sec 1.01 62.8±1.69ms 513.7 MB/sec
string_and_binary_view/default 1.00 51.0±1.36ms 632.0 MB/sec 1.01 51.7±1.51ms 623.5 MB/sec
string_and_binary_view/number_distinct_values 1.00 114.6±1.96ms 281.5 MB/sec 1.01 116.0±2.30ms 277.9 MB/sec
string_and_binary_view/parquet_2 1.01 58.7±0.39ms 549.1 MB/sec 1.00 58.1±0.71ms 554.8 MB/sec
string_and_binary_view/zstd 1.00 86.7±0.73ms 372.1 MB/sec 1.02 88.7±1.38ms 363.5 MB/sec
string_and_binary_view/zstd_parquet_2 1.00 72.1±0.23ms 447.0 MB/sec 1.01 72.8±1.11ms 443.0 MB/sec
string_dictionary/bloom_filter 1.31 135.0±6.33ms 1956.5 MB/sec 1.00 103.4±6.99ms 2.5 GB/sec
string_dictionary/cdc 1.65 87.8±5.30ms 2.9 GB/sec 1.00 53.3±5.73ms 4.8 GB/sec
string_dictionary/default 2.05 94.3±2.18ms 2.7 GB/sec 1.00 46.1±2.35ms 5.6 GB/sec
string_dictionary/number_distinct_values 1.00 98.2±1.57ms 2.6 GB/sec 1.08 106.3±11.50ms 2.4 GB/sec
string_dictionary/parquet_2 1.00 46.8±0.12ms 5.5 GB/sec 1.04 48.6±3.55ms 5.3 GB/sec
string_dictionary/zstd 1.06 244.2±4.54ms 1081.7 MB/sec 1.00 229.7±17.85ms 1149.8 MB/sec
string_dictionary/zstd_parquet_2 1.16 230.7±1.67ms 1145.1 MB/sec 1.00 198.3±9.46ms 1332.1 MB/sec
string_dictionary_low_cardinality_100/bloom_filter 1.00 21.5±0.08ms 186.8 MB/sec 1.00 21.5±0.05ms 186.3 MB/sec
string_dictionary_low_cardinality_100/cdc 1.00 23.6±0.09ms 169.9 MB/sec 1.01 23.9±0.12ms 167.7 MB/sec
string_dictionary_low_cardinality_100/default 1.00 21.3±0.07ms 187.7 MB/sec 1.00 21.4±0.06ms 187.2 MB/sec
string_dictionary_low_cardinality_100/number_distinct_values 1.00 44.7±0.15ms 89.6 MB/sec 1.00 44.9±1.00ms 89.1 MB/sec
string_dictionary_low_cardinality_100/parquet_2 1.00 21.3±0.09ms 188.3 MB/sec 1.01 21.6±0.22ms 185.6 MB/sec
string_dictionary_low_cardinality_100/zstd 1.00 22.2±0.05ms 180.8 MB/sec 1.00 22.1±0.05ms 181.2 MB/sec
string_dictionary_low_cardinality_100/zstd_parquet_2 1.00 22.1±0.05ms 181.0 MB/sec 1.00 22.1±0.05ms 181.6 MB/sec
string_dictionary_low_cardinality_20/bloom_filter 1.00 21.4±0.07ms 187.2 MB/sec 1.02 21.9±0.21ms 182.8 MB/sec
string_dictionary_low_cardinality_20/cdc 1.00 23.6±0.08ms 169.6 MB/sec 1.02 24.1±0.09ms 166.2 MB/sec
string_dictionary_low_cardinality_20/default 1.00 21.3±0.08ms 188.2 MB/sec 1.02 21.8±0.21ms 184.0 MB/sec
string_dictionary_low_cardinality_20/number_distinct_values 1.03 45.3±0.08ms 88.5 MB/sec 1.00 44.1±1.04ms 90.8 MB/sec
string_dictionary_low_cardinality_20/parquet_2 1.02 21.5±0.06ms 185.9 MB/sec 1.00 21.2±0.12ms 188.9 MB/sec
string_dictionary_low_cardinality_20/zstd 1.01 21.7±0.04ms 184.7 MB/sec 1.00 21.4±0.06ms 187.5 MB/sec
string_dictionary_low_cardinality_20/zstd_parquet_2 1.01 21.7±0.06ms 184.8 MB/sec 1.00 21.4±0.15ms 187.5 MB/sec
string_dictionary_low_cardinality_400/bloom_filter 1.00 22.1±0.05ms 181.2 MB/sec 1.00 22.2±0.10ms 181.1 MB/sec
string_dictionary_low_cardinality_400/cdc 1.00 24.2±0.05ms 165.5 MB/sec 1.00 24.3±0.11ms 165.0 MB/sec
string_dictionary_low_cardinality_400/default 1.00 21.9±0.10ms 183.6 MB/sec 1.00 21.9±0.05ms 183.5 MB/sec
string_dictionary_low_cardinality_400/number_distinct_values 1.00 45.8±0.12ms 87.6 MB/sec 1.00 45.6±0.89ms 87.9 MB/sec
string_dictionary_low_cardinality_400/parquet_2 1.00 22.2±0.11ms 180.8 MB/sec 1.01 22.4±0.06ms 179.0 MB/sec
string_dictionary_low_cardinality_400/zstd 1.00 22.5±0.05ms 178.0 MB/sec 1.01 22.7±0.17ms 176.9 MB/sec
string_dictionary_low_cardinality_400/zstd_parquet_2 1.00 22.5±0.05ms 178.5 MB/sec 1.00 22.5±0.06ms 177.9 MB/sec
string_non_null/bloom_filter 1.03 274.0±20.50ms 1912.1 MB/sec 1.00 265.1±28.71ms 1976.8 MB/sec
string_non_null/cdc 1.01 256.6±9.81ms 2042.3 MB/sec 1.00 254.5±24.39ms 2.0 GB/sec
string_non_null/default 1.16 109.9±10.96ms 4.7 GB/sec 1.00 94.4±15.75ms 5.4 GB/sec
string_non_null/number_distinct_values 1.00 285.8±10.99ms 1833.6 MB/sec 1.01 288.5±40.46ms 1816.5 MB/sec
string_non_null/parquet_2 1.05 115.2±6.98ms 4.4 GB/sec 1.00 109.8±17.12ms 4.7 GB/sec
string_non_null/zstd 1.00 550.5±1.97ms 951.9 MB/sec 1.05 577.7±34.83ms 907.0 MB/sec
string_non_null/zstd_parquet_2 1.00 501.3±10.21ms 1045.4 MB/sec 1.04 519.0±24.58ms 1009.7 MB/sec
string_ree/bloom_filter 1.70 186.8±9.72ms 87.8 MB/sec 1.00 109.8±8.79ms 149.5 MB/sec
string_ree/cdc 1.20 227.2±0.60ms 72.2 MB/sec 1.00 188.6±18.28ms 87.0 MB/sec
string_ree/default 1.54 129.2±1.38ms 127.0 MB/sec 1.00 83.8±13.60ms 195.9 MB/sec
string_ree/number_distinct_values 1.12 138.5±2.14ms 118.4 MB/sec 1.00 124.0±12.47ms 132.4 MB/sec
string_ree/parquet_2 1.64 128.8±0.40ms 127.4 MB/sec 1.00 78.4±1.83ms 209.3 MB/sec
string_ree/zstd 1.30 192.0±3.68ms 85.4 MB/sec 1.00 147.3±5.36ms 111.4 MB/sec
string_ree/zstd_parquet_2 1.49 139.0±0.72ms 118.0 MB/sec 1.00 93.2±2.13ms 176.0 MB/sec
string_ree_95pct_null/bloom_filter 1.00 24.4±0.04ms 57.7 MB/sec 1.00 24.5±0.15ms 57.5 MB/sec
string_ree_95pct_null/cdc 1.00 32.1±0.12ms 43.9 MB/sec 1.01 32.3±0.18ms 43.5 MB/sec
string_ree_95pct_null/default 1.00 25.0±0.09ms 56.2 MB/sec 1.00 25.1±0.44ms 56.0 MB/sec
string_ree_95pct_null/number_distinct_values 1.00 26.6±0.12ms 52.8 MB/sec 1.00 26.6±0.43ms 52.8 MB/sec
string_ree_95pct_null/parquet_2 1.00 25.0±0.09ms 56.3 MB/sec 1.00 25.1±0.38ms 56.0 MB/sec
string_ree_95pct_null/zstd 1.00 25.0±0.05ms 56.4 MB/sec 1.00 25.1±0.25ms 56.1 MB/sec
string_ree_95pct_null/zstd_parquet_2 1.00 25.6±0.14ms 54.9 MB/sec 1.00 25.6±0.47ms 54.9 MB/sec
struct_all_null/bloom_filter 1.08 377.0±1.33µs 41.8 GB/sec 1.00 348.1±1.22µs 45.2 GB/sec
struct_all_null/cdc 1.03 7.7±0.08ms 2.0 GB/sec 1.00 7.5±0.07ms 2.1 GB/sec
struct_all_null/default 1.26 115.5±0.73µs 136.4 GB/sec 1.00 91.5±0.58µs 172.1 GB/sec
struct_all_null/number_distinct_values 1.27 116.3±0.66µs 135.4 GB/sec 1.00 91.9±0.68µs 171.4 GB/sec
struct_all_null/parquet_2 1.28 119.4±0.95µs 131.9 GB/sec 1.00 93.1±1.22µs 169.1 GB/sec
struct_all_null/zstd 1.31 179.7±0.78µs 87.7 GB/sec 1.00 137.0±0.83µs 114.9 GB/sec
struct_all_null/zstd_parquet_2 1.29 158.8±0.78µs 99.2 GB/sec 1.00 123.2±1.11µs 127.9 GB/sec
struct_non_null/bloom_filter 1.00 39.9±0.35ms 401.1 MB/sec 1.07 42.8±1.09ms 374.1 MB/sec
struct_non_null/cdc 1.00 39.7±0.19ms 402.6 MB/sec 1.03 40.9±0.97ms 391.2 MB/sec
struct_non_null/default 1.00 27.3±0.46ms 585.5 MB/sec 1.01 27.7±0.60ms 577.0 MB/sec
struct_non_null/number_distinct_values 1.00 188.3±2.83ms 85.0 MB/sec 1.10 208.0±29.70ms 76.9 MB/sec
struct_non_null/parquet_2 1.00 33.8±0.40ms 472.7 MB/sec 1.02 34.7±0.88ms 461.2 MB/sec
struct_non_null/zstd 1.00 35.3±0.30ms 452.7 MB/sec 1.08 38.1±1.00ms 420.4 MB/sec
struct_non_null/zstd_parquet_2 1.00 46.3±0.08ms 345.4 MB/sec 1.04 48.2±0.71ms 331.9 MB/sec
struct_sparse_99pct_null/bloom_filter 1.00 6.5±0.04ms 2.4 GB/sec 1.02 6.6±0.06ms 2.4 GB/sec
struct_sparse_99pct_null/cdc 1.00 13.9±0.22ms 1160.4 MB/sec 1.12 15.5±0.22ms 1038.4 MB/sec
struct_sparse_99pct_null/default 1.00 6.0±0.03ms 2.6 GB/sec 1.03 6.1±0.06ms 2.6 GB/sec
struct_sparse_99pct_null/number_distinct_values 1.00 7.8±0.24ms 2.0 GB/sec 1.11 8.7±0.17ms 1857.2 MB/sec
struct_sparse_99pct_null/parquet_2 1.00 5.9±0.02ms 2.7 GB/sec 1.05 6.2±0.06ms 2.5 GB/sec
struct_sparse_99pct_null/zstd 1.00 7.3±0.01ms 2.2 GB/sec 1.04 7.6±0.08ms 2.1 GB/sec
struct_sparse_99pct_null/zstd_parquet_2 1.00 6.8±0.02ms 2.3 GB/sec 1.02 6.9±0.07ms 2.3 GB/sec
```
</details>
<details><summary>Per-runner information</summary>
<details><summary>arrow_writer — shard 1/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-9ad7f920-gxhw
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065966059-3178-brnvq 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 805.2s |
| Peak memory | 2.8 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 771.7s |
| CPU sys | 30.3s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 795.2s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.4 GiB |
| CPU user | 776.3s |
| CPU sys | 16.6s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 2/4</summary>
**Node:** gk3-benchmark-cluster-nap-1p8mzl3a-8368b426-czzt
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065966059-3179-w97sb 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 1130.2s |
| Peak memory | 2.9 GiB |
| Avg memory | 2.6 GiB |
| CPU user | 1080.2s |
| CPU sys | 46.1s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 1130.3s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 1087.7s |
| CPU sys | 38.7s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 3/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-59aaa3df-hg9x
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065966059-3180-d7j2x 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 1240.3s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.6 GiB |
| CPU user | 1193.7s |
| CPU sys | 40.8s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 1230.3s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 1194.9s |
| CPU sys | 30.6s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 4/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-963d8389-zdwp
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065966059-3181-bg7mv 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 1020.2s |
| Peak memory | 2.9 GiB |
| Avg memory | 2.3 GiB |
| CPU user | 938.8s |
| CPU sys | 75.4s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 980.2s |
| Peak memory | 3.1 GiB |
| Avg memory | 2.3 GiB |
| CPU user | 961.4s |
| CPU sys | 16.8s |
| Peak spill | 0 B |
</details>
</details>
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
<!-- benchmark-request:6065966059:finish:ca9159f5c5f8c4141b00aa40c1a88788 -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6067291362
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065969074)
**Target:** arrow_writer
Comparing b1c118bd52655725911d374aff4d1a0d0b85a4c9 (b1c118bd52655725911d374aff4d1a0d0b85a4c9) to 7fc78ca9ecf907de396119a25bc689527f0fd0ea [diff](https://github.com/apache/arrow-rs/compare/7fc78ca9ecf907de396119a25bc689527f0fd0ea..b1c118bd52655725911d374aff4d1a0d0b85a4c9)
<details><summary>Run configuration</summary>
```yaml
run benchmark arrow_writer
baseline:
ref: 7fc78ca9ecf907de396119a25bc689527f0fd0ea
changed:
env:
MIMALLOC_PURGE_DELAY: '-1'
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
shards: 4
```
</details>
<details><summary>Details</summary>
```
group base changed
----- ---- -------
bool/bloom_filter 1.06 11.6±0.09ms 21.6 MB/sec 1.00 10.9±0.02ms 22.9 MB/sec
bool/cdc 1.00 13.0±0.03ms 19.3 MB/sec 1.02 13.3±0.03ms 18.8 MB/sec
bool/default 1.00 8.9±0.05ms 28.1 MB/sec 1.00 8.9±0.05ms 28.1 MB/sec
bool/number_distinct_values 1.00 18.9±0.12ms 13.2 MB/sec 1.00 18.9±0.06ms 13.2 MB/sec
bool/parquet_2 1.06 11.3±0.10ms 22.1 MB/sec 1.00 10.6±0.03ms 23.5 MB/sec
bool/zstd 1.00 9.4±0.03ms 26.6 MB/sec 1.01 9.5±0.07ms 26.3 MB/sec
bool/zstd_parquet_2 1.07 11.7±0.10ms 21.3 MB/sec 1.00 11.0±0.03ms 22.7 MB/sec
bool_non_null/bloom_filter 1.02 5.1±0.01ms 24.4 MB/sec 1.00 5.0±0.01ms 24.9 MB/sec
bool_non_null/cdc 1.02 3.9±0.03ms 31.8 MB/sec 1.00 3.9±0.01ms 32.4 MB/sec
bool_non_null/default 1.00 2.3±0.01ms 55.3 MB/sec 1.01 2.3±0.01ms 54.8 MB/sec
bool_non_null/number_distinct_values 1.00 15.7±0.04ms 8.0 MB/sec 1.00 15.7±0.03ms 8.0 MB/sec
bool_non_null/parquet_2 1.03 4.7±0.02ms 26.7 MB/sec 1.00 4.6±0.01ms 27.3 MB/sec
bool_non_null/zstd 1.01 2.7±0.01ms 46.9 MB/sec 1.00 2.6±0.01ms 47.4 MB/sec
bool_non_null/zstd_parquet_2 1.00 5.0±0.01ms 24.9 MB/sec 1.01 5.1±0.02ms 24.7 MB/sec
bool_ree/bloom_filter 1.00 27.9±0.13ms 7.6 MB/sec 1.00 27.8±0.08ms 7.6 MB/sec
bool_ree/cdc 1.00 28.2±0.07ms 7.6 MB/sec 1.00 28.3±0.05ms 7.5 MB/sec
bool_ree/default 1.00 25.1±0.10ms 8.5 MB/sec 1.01 25.4±0.17ms 8.4 MB/sec
bool_ree/number_distinct_values 1.00 37.1±0.09ms 5.7 MB/sec 1.00 37.0±0.04ms 5.7 MB/sec
bool_ree/parquet_2 1.01 26.0±0.13ms 8.2 MB/sec 1.00 25.9±0.03ms 8.2 MB/sec
bool_ree/zstd 1.00 25.8±0.05ms 8.3 MB/sec 1.00 25.8±0.09ms 8.3 MB/sec
bool_ree/zstd_parquet_2 1.00 26.4±0.18ms 8.1 MB/sec 1.00 26.5±0.17ms 8.0 MB/sec
decimal/bloom_filter 1.30 169.5±1.22ms 353.9 MB/sec 1.00 130.5±1.23ms 459.8 MB/sec
decimal/cdc 1.02 105.5±0.50ms 568.8 MB/sec 1.00 103.3±0.69ms 580.9 MB/sec
decimal/default 1.01 81.3±0.67ms 738.0 MB/sec 1.00 80.6±0.64ms 744.2 MB/sec
decimal/number_distinct_values 1.13 320.8±1.89ms 187.1 MB/sec 1.00 284.8±0.98ms 210.7 MB/sec
decimal/parquet_2 1.22 206.2±2.61ms 290.9 MB/sec 1.00 169.4±2.96ms 354.2 MB/sec
decimal/zstd 1.42 133.8±1.36ms 448.3 MB/sec 1.00 94.1±1.52ms 637.7 MB/sec
decimal/zstd_parquet_2 1.21 228.7±2.95ms 262.3 MB/sec 1.00 189.2±2.33ms 317.1 MB/sec
fixed_size_binary_ree/bloom_filter 1.00 64.7±0.49ms 15.6 MB/sec 1.02 65.7±0.42ms 15.3 MB/sec
fixed_size_binary_ree/cdc 1.04 74.7±0.41ms 13.5 MB/sec 1.00 71.6±0.41ms 14.1 MB/sec
fixed_size_binary_ree/default 1.04 63.9±1.13ms 15.8 MB/sec 1.00 61.7±0.95ms 16.3 MB/sec
fixed_size_binary_ree/number_distinct_values 1.04 82.8±1.03ms 12.2 MB/sec 1.00 79.9±1.00ms 12.6 MB/sec
fixed_size_binary_ree/parquet_2 1.00 63.3±0.41ms 15.9 MB/sec 1.01 63.9±0.42ms 15.8 MB/sec
fixed_size_binary_ree/zstd 1.04 65.3±0.45ms 15.4 MB/sec 1.00 62.7±0.98ms 16.1 MB/sec
fixed_size_binary_ree/zstd_parquet_2 1.00 63.9±0.42ms 15.8 MB/sec 1.01 64.6±0.99ms 15.6 MB/sec
float_with_nans/bloom_filter 1.12 88.1±0.44ms 158.9 MB/sec 1.00 78.4±1.09ms 178.5 MB/sec
float_with_nans/cdc 1.01 78.0±0.73ms 179.4 MB/sec 1.00 77.3±0.54ms 181.1 MB/sec
float_with_nans/default 1.01 70.5±0.35ms 198.7 MB/sec 1.00 70.0±0.38ms 199.9 MB/sec
float_with_nans/number_distinct_values 1.04 172.4±0.82ms 81.2 MB/sec 1.00 166.5±0.47ms 84.1 MB/sec
float_with_nans/parquet_2 1.01 69.8±0.34ms 200.5 MB/sec 1.00 68.8±1.13ms 203.3 MB/sec
float_with_nans/zstd 1.00 107.8±0.36ms 129.9 MB/sec 1.00 107.8±0.40ms 129.9 MB/sec
float_with_nans/zstd_parquet_2 1.09 116.5±0.40ms 120.1 MB/sec 1.00 106.8±0.37ms 131.1 MB/sec
fsb/bloom_filter 1.00 12.6±0.07ms 1276.5 MB/sec 1.01 12.8±0.15ms 1259.8 MB/sec
fsb/cdc 1.00 14.8±0.05ms 1087.3 MB/sec 1.02 15.1±0.17ms 1067.3 MB/sec
fsb/default 1.02 11.3±0.07ms 1432.7 MB/sec 1.00 11.0±0.10ms 1463.7 MB/sec
fsb/number_distinct_values 1.01 15.4±0.06ms 1048.3 MB/sec 1.00 15.3±0.11ms 1056.6 MB/sec
fsb/parquet_2 1.00 13.0±0.10ms 1240.0 MB/sec 1.04 13.5±0.50ms 1194.7 MB/sec
fsb/zstd 1.00 13.0±0.05ms 1237.4 MB/sec 1.00 13.1±0.11ms 1233.3 MB/sec
fsb/zstd_parquet_2 1.02 13.9±0.09ms 1160.1 MB/sec 1.00 13.6±0.13ms 1183.3 MB/sec
int32_ree/bloom_filter 1.01 34.9±0.12ms 11.7 MB/sec 1.00 34.4±0.07ms 11.8 MB/sec
int32_ree/cdc 1.00 37.6±0.12ms 10.8 MB/sec 1.02 38.4±0.15ms 10.6 MB/sec
int32_ree/default 1.02 34.4±0.11ms 11.8 MB/sec 1.00 33.8±0.05ms 12.0 MB/sec
int32_ree/number_distinct_values 1.02 50.8±0.13ms 8.0 MB/sec 1.00 50.1±0.06ms 8.1 MB/sec
int32_ree/parquet_2 1.00 33.9±0.10ms 12.0 MB/sec 1.01 34.2±0.15ms 11.9 MB/sec
int32_ree/zstd 1.00 34.6±0.16ms 11.7 MB/sec 1.00 34.7±0.17ms 11.7 MB/sec
int32_ree/zstd_parquet_2 1.01 34.7±0.11ms 11.7 MB/sec 1.00 34.3±0.07ms 11.9 MB/sec
int32_ree_95pct_null/bloom_filter 1.00 21.3±0.04ms 19.1 MB/sec 1.00 21.3±0.03ms 19.1 MB/sec
int32_ree_95pct_null/cdc 1.00 23.7±0.08ms 17.2 MB/sec 1.01 23.9±0.06ms 17.0 MB/sec
int32_ree_95pct_null/default 1.00 21.1±0.05ms 19.3 MB/sec 1.00 21.1±0.03ms 19.2 MB/sec
int32_ree_95pct_null/number_distinct_values 1.00 22.1±0.05ms 18.4 MB/sec 1.01 22.2±0.03ms 18.3 MB/sec
int32_ree_95pct_null/parquet_2 1.00 21.1±0.06ms 19.3 MB/sec 1.00 21.0±0.02ms 19.4 MB/sec
int32_ree_95pct_null/zstd 1.00 21.4±0.05ms 19.0 MB/sec 1.00 21.5±0.02ms 18.9 MB/sec
int32_ree_95pct_null/zstd_parquet_2 1.01 21.5±0.05ms 18.9 MB/sec 1.00 21.2±0.02ms 19.2 MB/sec
large_string_distinct/delta_byte_array 1.00 59.9±0.19ms 4.2 GB/sec 1.04 62.1±0.53ms 4.0 GB/sec
large_string_distinct/plain 1.00 61.3±0.78ms 4.1 GB/sec 1.00 61.0±0.71ms 4.1 GB/sec
large_string_distinct_nullable/delta_byte_array 1.00 56.4±0.21ms 4.4 GB/sec 1.01 57.2±0.33ms 4.4 GB/sec
large_string_non_null/bloom_filter 1.06 51.3±0.14ms 4.9 GB/sec 1.00 48.3±0.13ms 5.2 GB/sec
large_string_non_null/cdc 1.02 240.7±0.77ms 1063.8 MB/sec 1.00 236.7±0.98ms 1081.4 MB/sec
large_string_non_null/default 1.00 51.0±0.17ms 4.9 GB/sec 1.01 51.7±0.53ms 4.8 GB/sec
large_string_non_null/number_distinct_values 1.02 68.1±0.26ms 3.7 GB/sec 1.00 66.8±0.56ms 3.7 GB/sec
large_string_non_null/parquet_2 1.00 51.1±0.16ms 4.9 GB/sec 1.00 51.2±0.57ms 4.9 GB/sec
large_string_non_null/zstd 1.02 51.3±0.12ms 4.9 GB/sec 1.00 50.2±0.37ms 5.0 GB/sec
large_string_non_null/zstd_parquet_2 1.04 51.4±0.17ms 4.9 GB/sec 1.00 49.3±0.15ms 5.1 GB/sec
large_string_shared_prefix/delta_byte_array 1.02 43.5±0.16ms 5.7 GB/sec 1.00 42.6±0.18ms 5.9 GB/sec
large_string_shared_prefix/plain 1.07 94.8±0.92ms 2.6 GB/sec 1.00 88.9±0.79ms 2.8 GB/sec
large_string_shared_prefix_list/delta_byte_array 1.01 55.2±0.30ms 4.5 GB/sec 1.00 54.8±0.49ms 4.6 GB/sec
large_string_shared_prefix_nullable/delta_byte_array 1.01 41.0±0.29ms 6.1 GB/sec 1.00 40.6±0.26ms 6.2 GB/sec
large_string_shared_prefix_nullable/plain 1.00 63.9±0.45ms 3.9 GB/sec 1.05 67.1±1.10ms 3.7 GB/sec
large_string_shared_prefix_nullable_dense/delta_byte_array 1.03 22.0±0.17ms 5.7 GB/sec 1.00 21.5±0.10ms 5.8 GB/sec
large_string_shared_prefix_nullable_trailing/delta_byte_array 1.02 40.4±0.05ms 6.2 GB/sec 1.00 39.7±0.06ms 6.3 GB/sec
list_nested/bloom_filter 1.00 132.7±0.47ms 217.5 MB/sec 1.01 133.8±0.74ms 215.6 MB/sec
list_nested/cdc 1.00 150.2±0.41ms 192.1 MB/sec 1.00 150.9±0.28ms 191.2 MB/sec
list_nested/default 1.00 119.1±0.21ms 242.2 MB/sec 1.01 119.9±0.73ms 240.7 MB/sec
list_nested/number_distinct_values 1.00 276.1±2.18ms 104.5 MB/sec 1.04 286.5±17.60ms 100.7 MB/sec
list_nested/parquet_2 1.00 129.4±0.22ms 223.1 MB/sec 1.00 129.3±0.51ms 223.1 MB/sec
list_nested/zstd 1.00 127.6±0.25ms 226.1 MB/sec 1.00 128.3±0.49ms 225.0 MB/sec
list_nested/zstd_parquet_2 1.00 142.4±0.23ms 202.6 MB/sec 1.01 144.5±0.96ms 199.7 MB/sec
list_primitive/bloom_filter 1.16 307.2±23.65ms 1775.5 MB/sec 1.00 263.9±1.18ms 2.0 GB/sec
list_primitive/cdc 1.04 306.7±7.94ms 1778.1 MB/sec 1.00 294.0±0.58ms 1855.3 MB/sec
list_primitive/default 1.11 206.1±2.38ms 2.6 GB/sec 1.00 185.4±0.30ms 2.9 GB/sec
list_primitive/number_distinct_values 1.00 368.1±0.68ms 1481.5 MB/sec 1.02 374.6±1.74ms 1455.8 MB/sec
list_primitive/parquet_2 1.27 257.6±0.73ms 2.1 GB/sec 1.00 203.4±0.33ms 2.6 GB/sec
list_primitive/zstd 1.00 435.7±0.99ms 1251.6 MB/sec 1.02 443.3±0.32ms 1230.3 MB/sec
list_primitive/zstd_parquet_2 1.00 425.8±1.31ms 1280.7 MB/sec 1.00 427.8±1.93ms 1274.9 MB/sec
list_primitive_non_null/bloom_filter 1.06 355.1±9.78ms 1532.4 MB/sec 1.00 333.8±3.26ms 1630.3 MB/sec
list_primitive_non_null/cdc 1.01 378.0±7.02ms 1439.8 MB/sec 1.00 372.7±0.78ms 1460.2 MB/sec
list_primitive_non_null/default 1.20 225.8±3.55ms 2.4 GB/sec 1.00 188.7±4.26ms 2.8 GB/sec
list_primitive_non_null/number_distinct_values 1.11 555.8±6.01ms 979.1 MB/sec 1.00 501.9±4.13ms 1084.4 MB/sec
list_primitive_non_null/parquet_2 1.05 224.9±11.03ms 2.4 GB/sec 1.00 214.4±0.49ms 2.5 GB/sec
list_primitive_non_null/zstd 1.07 681.2±5.88ms 798.9 MB/sec 1.00 637.4±1.93ms 853.9 MB/sec
list_primitive_non_null/zstd_parquet_2 1.07 647.8±32.01ms 840.2 MB/sec 1.00 603.7±1.14ms 901.5 MB/sec
list_primitive_sparse_99pct_null/bloom_filter 1.00 12.6±0.04ms 2.9 GB/sec 1.05 13.3±0.03ms 2.7 GB/sec
list_primitive_sparse_99pct_null/cdc 1.00 24.2±0.18ms 1543.2 MB/sec 1.02 24.6±0.09ms 1517.2 MB/sec
list_primitive_sparse_99pct_null/default 1.00 12.7±0.09ms 2.9 GB/sec 1.00 12.6±0.07ms 2.9 GB/sec
list_primitive_sparse_99pct_null/number_distinct_values 1.00 12.4±0.03ms 2.9 GB/sec 1.04 12.9±0.04ms 2.8 GB/sec
list_primitive_sparse_99pct_null/parquet_2 1.04 12.8±0.06ms 2.8 GB/sec 1.00 12.3±0.08ms 3.0 GB/sec
list_primitive_sparse_99pct_null/zstd 1.01 14.6±0.08ms 2.5 GB/sec 1.00 14.5±0.08ms 2.5 GB/sec
list_primitive_sparse_99pct_null/zstd_parquet_2 1.00 12.6±0.16ms 2.9 GB/sec 1.02 12.8±0.10ms 2.9 GB/sec
list_struct_with_list/bloom_filter 1.00 255.9±0.72ms 178.2 MB/sec 1.00 255.5±0.86ms 178.4 MB/sec
list_struct_with_list/cdc 1.00 295.0±0.76ms 154.6 MB/sec 1.00 296.2±1.08ms 153.9 MB/sec
list_struct_with_list/default 1.00 229.5±0.31ms 198.6 MB/sec 1.01 230.8±0.64ms 197.6 MB/sec
list_struct_with_list/number_distinct_values 1.03 493.4±0.96ms 92.4 MB/sec 1.00 477.1±2.75ms 95.6 MB/sec
list_struct_with_list/parquet_2 1.00 244.1±0.50ms 186.8 MB/sec 1.00 244.1±0.62ms 186.8 MB/sec
list_struct_with_list/zstd 1.00 252.0±0.48ms 181.0 MB/sec 1.00 252.6±0.65ms 180.5 MB/sec
list_struct_with_list/zstd_parquet_2 1.00 267.7±0.56ms 170.3 MB/sec 1.00 267.6±0.98ms 170.4 MB/sec
medium_string_shared_prefix_nullable/delta_byte_array 1.00 35.5±0.12ms 7.0 GB/sec 1.12 39.7±0.26ms 6.3 GB/sec
ndv/bool/ndv_on 1.00 9.8±0.02ms 8.1 MB/sec 1.00 9.8±0.14ms 8.1 MB/sec
ndv/fsb16_high_cardinality/ndv_on 1.00 40.7±0.37ms 246.0 MB/sec 1.02 41.6±2.34ms 240.4 MB/sec
ndv/fsb16_low_cardinality/ndv_on 1.00 35.1±0.25ms 285.2 MB/sec 1.00 35.1±0.23ms 285.0 MB/sec
ndv/int32_high_cardinality/ndv_on 1.07 28.7±0.11ms 87.0 MB/sec 1.00 26.8±0.08ms 93.2 MB/sec
ndv/int32_low_cardinality/ndv_on 1.00 16.1±0.05ms 155.2 MB/sec 1.00 16.1±0.07ms 155.2 MB/sec
ndv/large_binary_high_cardinality/ndv_on 1.02 37.4±0.07ms 200.5 MB/sec 1.00 36.8±0.24ms 203.9 MB/sec
ndv/large_binary_low_cardinality/ndv_on 1.00 24.8±0.08ms 605.0 MB/sec 1.02 25.3±0.09ms 592.4 MB/sec
ndv/utf8_high_cardinality/ndv_on 1.04 72.5±2.27ms 2.2 GB/sec 1.00 69.7±1.62ms 2.3 GB/sec
ndv/utf8_low_cardinality/ndv_on 1.00 23.8±0.08ms 315.5 MB/sec 1.02 24.2±0.07ms 310.1 MB/sec
ndv/utf8_view_high_cardinality/ndv_on 1.01 38.7±0.29ms 258.5 MB/sec 1.00 38.4±2.67ms 260.6 MB/sec
ndv/utf8_view_low_cardinality/ndv_on 1.00 23.7±0.08ms 421.5 MB/sec 1.00 23.8±0.27ms 419.7 MB/sec
primitive/bloom_filter 1.07 139.0±1.17ms 322.8 MB/sec 1.00 129.5±0.72ms 346.6 MB/sec
primitive/cdc 1.00 144.7±0.55ms 310.1 MB/sec 1.01 145.8±0.52ms 307.8 MB/sec
primitive/default 1.00 104.7±0.42ms 428.5 MB/sec 1.00 105.0±0.51ms 427.4 MB/sec
primitive/number_distinct_values 1.08 337.5±1.86ms 133.0 MB/sec 1.00 312.9±0.32ms 143.4 MB/sec
primitive/parquet_2 1.01 116.3±0.52ms 385.9 MB/sec 1.00 114.9±0.60ms 390.4 MB/sec
primitive/zstd 1.00 134.9±0.54ms 332.6 MB/sec 1.02 137.1±0.44ms 327.3 MB/sec
primitive/zstd_parquet_2 1.00 148.8±0.43ms 301.5 MB/sec 1.00 149.4±0.30ms 300.5 MB/sec
primitive_all_null/bloom_filter 1.12 911.1±6.68µs 48.1 GB/sec 1.00 812.3±4.23µs 53.9 GB/sec
primitive_all_null/cdc 1.01 18.8±0.26ms 2.3 GB/sec 1.00 18.5±0.24ms 2.4 GB/sec
primitive_all_null/default 1.31 271.2±1.40µs 161.6 GB/sec 1.00 207.8±1.51µs 210.9 GB/sec
primitive_all_null/number_distinct_values 1.30 270.3±1.56µs 162.2 GB/sec 1.00 208.2±1.40µs 210.5 GB/sec
primitive_all_null/parquet_2 1.28 273.6±1.88µs 160.2 GB/sec 1.00 213.5±2.45µs 205.3 GB/sec
primitive_all_null/zstd 1.32 417.9±2.17µs 104.9 GB/sec 1.00 316.1±2.11µs 138.7 GB/sec
primitive_all_null/zstd_parquet_2 1.31 371.1±2.16µs 118.1 GB/sec 1.00 283.5±2.41µs 154.6 GB/sec
primitive_non_null/bloom_filter 1.00 89.8±0.29ms 489.8 MB/sec 1.02 91.4±0.40ms 481.2 MB/sec
primitive_non_null/cdc 1.01 78.9±1.34ms 557.7 MB/sec 1.00 78.3±0.35ms 562.0 MB/sec
primitive_non_null/default 1.00 55.9±0.26ms 787.2 MB/sec 1.01 56.5±0.28ms 779.4 MB/sec
primitive_non_null/number_distinct_values 1.20 480.6±2.09ms 91.6 MB/sec 1.00 399.4±0.79ms 110.2 MB/sec
primitive_non_null/parquet_2 1.00 72.0±0.25ms 611.4 MB/sec 1.00 72.2±0.19ms 609.6 MB/sec
primitive_non_null/zstd 1.02 88.4±0.17ms 497.9 MB/sec 1.00 87.0±0.14ms 505.6 MB/sec
primitive_non_null/zstd_parquet_2 1.01 106.3±0.25ms 414.0 MB/sec 1.00 104.9±0.12ms 419.3 MB/sec
primitive_sparse_99pct_null/bloom_filter 1.10 13.1±0.13ms 3.3 GB/sec 1.00 11.9±0.22ms 3.7 GB/sec
primitive_sparse_99pct_null/cdc 1.00 29.2±0.24ms 1534.4 MB/sec 1.10 32.0±0.27ms 1400.5 MB/sec
primitive_sparse_99pct_null/default 1.00 10.5±0.04ms 4.2 GB/sec 1.04 10.9±0.04ms 4.0 GB/sec
primitive_sparse_99pct_null/number_distinct_values 1.00 14.2±0.53ms 3.1 GB/sec 1.05 14.9±0.22ms 2.9 GB/sec
primitive_sparse_99pct_null/parquet_2 1.00 10.5±0.03ms 4.2 GB/sec 1.04 11.0±0.04ms 4.0 GB/sec
primitive_sparse_99pct_null/zstd 1.00 14.3±0.08ms 3.1 GB/sec 1.00 14.3±0.12ms 3.1 GB/sec
primitive_sparse_99pct_null/zstd_parquet_2 1.00 12.4±0.04ms 3.5 GB/sec 1.06 13.2±0.03ms 3.3 GB/sec
short_string_non_null/bloom_filter 1.07 28.1±0.09ms 426.8 MB/sec 1.00 26.4±0.10ms 455.3 MB/sec
short_string_non_null/cdc 1.00 19.5±0.08ms 615.2 MB/sec 1.03 20.0±0.04ms 599.0 MB/sec
short_string_non_null/default 1.00 15.6±0.04ms 766.9 MB/sec 1.01 15.8±0.03ms 758.3 MB/sec
short_string_non_null/number_distinct_values 1.11 72.7±0.62ms 165.1 MB/sec 1.00 65.4±0.28ms 183.4 MB/sec
short_string_non_null/parquet_2 1.00 24.8±0.04ms 484.7 MB/sec 1.01 24.9±0.08ms 481.5 MB/sec
short_string_non_null/zstd 1.00 35.6±0.11ms 336.9 MB/sec 1.01 35.9±0.10ms 334.1 MB/sec
short_string_non_null/zstd_parquet_2 1.00 27.7±0.06ms 433.4 MB/sec 1.00 27.8±0.07ms 432.1 MB/sec
small_string_distinct/delta_byte_array 1.06 957.0±4.49µs 8.2 GB/sec 1.00 898.7±5.46µs 8.7 GB/sec
small_string_distinct/plain 1.04 778.6±1.98µs 10.1 GB/sec 1.00 746.6±2.92µs 10.5 GB/sec
small_string_partial_prefix/delta_byte_array 1.04 1005.2±3.09µs 7.8 GB/sec 1.00 969.7±3.29µs 8.1 GB/sec
small_string_partial_prefix/plain 1.05 1041.0±3.81µs 7.5 GB/sec 1.00 990.8±4.09µs 7.9 GB/sec
small_string_shared_prefix/delta_byte_array 1.01 994.1±6.11µs 7.9 GB/sec 1.00 980.2±4.66µs 8.0 GB/sec
small_string_shared_prefix/plain 1.00 1070.4±7.10µs 7.3 GB/sec 1.05 1122.1±14.96µs 7.0 GB/sec
string/bloom_filter 1.32 222.8±25.14ms 2.3 GB/sec 1.00 168.9±4.58ms 3.0 GB/sec
string/cdc 1.09 212.4±6.98ms 2.4 GB/sec 1.00 195.7±0.31ms 2.6 GB/sec
string/default 1.72 138.5±25.50ms 3.7 GB/sec 1.00 80.5±0.64ms 6.4 GB/sec
string/number_distinct_values 1.23 227.5±4.97ms 2.3 GB/sec 1.00 185.0±1.78ms 2.8 GB/sec
string/parquet_2 1.81 166.8±1.29ms 3.1 GB/sec 1.00 92.0±0.15ms 5.6 GB/sec
string/zstd 1.17 482.2±16.77ms 1087.2 MB/sec 1.00 410.6±0.45ms 1276.7 MB/sec
string/zstd_parquet_2 1.03 388.9±8.73ms 1348.2 MB/sec 1.00 378.3±1.12ms 1385.6 MB/sec
string_and_binary_view/bloom_filter 1.02 62.7±0.22ms 514.3 MB/sec 1.00 61.3±0.15ms 526.1 MB/sec
string_and_binary_view/cdc 1.02 59.2±0.29ms 544.6 MB/sec 1.00 58.1±0.09ms 555.1 MB/sec
string_and_binary_view/default 1.02 48.4±0.18ms 665.7 MB/sec 1.00 47.4±0.07ms 680.6 MB/sec
string_and_binary_view/number_distinct_values 1.02 111.4±0.38ms 289.5 MB/sec 1.00 109.0±0.50ms 295.9 MB/sec
string_and_binary_view/parquet_2 1.06 60.0±0.57ms 537.7 MB/sec 1.00 56.8±0.09ms 568.0 MB/sec
string_and_binary_view/zstd 1.00 85.4±0.16ms 377.7 MB/sec 1.00 85.7±0.25ms 376.5 MB/sec
string_and_binary_view/zstd_parquet_2 1.04 73.6±0.62ms 438.0 MB/sec 1.00 71.0±0.11ms 454.4 MB/sec
string_dictionary/bloom_filter 1.47 118.8±0.79ms 2.2 GB/sec 1.00 80.9±0.13ms 3.2 GB/sec
string_dictionary/cdc 1.60 73.6±5.49ms 3.5 GB/sec 1.00 46.1±0.08ms 5.6 GB/sec
string_dictionary/default 2.17 90.2±0.63ms 2.9 GB/sec 1.00 41.6±0.17ms 6.2 GB/sec
string_dictionary/number_distinct_values 1.12 105.6±1.56ms 2.4 GB/sec 1.00 94.2±0.25ms 2.7 GB/sec
string_dictionary/parquet_2 1.03 48.1±0.21ms 5.4 GB/sec 1.00 46.7±0.13ms 5.5 GB/sec
string_dictionary/zstd 1.13 233.6±2.75ms 1130.8 MB/sec 1.00 206.8±0.27ms 1277.0 MB/sec
string_dictionary/zstd_parquet_2 1.19 227.4±0.64ms 1161.8 MB/sec 1.00 190.7±0.18ms 1385.2 MB/sec
string_dictionary_low_cardinality_100/bloom_filter 1.05 22.7±0.89ms 176.1 MB/sec 1.00 21.8±0.06ms 184.1 MB/sec
string_dictionary_low_cardinality_100/cdc 1.01 24.0±0.06ms 167.0 MB/sec 1.00 23.9±0.07ms 167.9 MB/sec
string_dictionary_low_cardinality_100/default 1.05 22.6±1.02ms 177.0 MB/sec 1.00 21.6±0.06ms 185.1 MB/sec
string_dictionary_low_cardinality_100/number_distinct_values 1.02 44.9±0.09ms 89.3 MB/sec 1.00 44.2±0.09ms 90.7 MB/sec
string_dictionary_low_cardinality_100/parquet_2 1.02 21.9±0.22ms 183.1 MB/sec 1.00 21.5±0.08ms 186.1 MB/sec
string_dictionary_low_cardinality_100/zstd 1.01 22.1±0.07ms 181.2 MB/sec 1.00 21.8±0.08ms 183.4 MB/sec
string_dictionary_low_cardinality_100/zstd_parquet_2 1.01 22.1±0.07ms 181.3 MB/sec 1.00 21.8±0.08ms 183.6 MB/sec
string_dictionary_low_cardinality_20/bloom_filter 1.01 21.6±0.22ms 185.1 MB/sec 1.00 21.3±0.04ms 187.7 MB/sec
string_dictionary_low_cardinality_20/cdc 1.01 23.7±0.10ms 168.9 MB/sec 1.00 23.5±0.04ms 170.5 MB/sec
string_dictionary_low_cardinality_20/default 1.02 21.5±0.22ms 186.1 MB/sec 1.00 21.1±0.04ms 189.7 MB/sec
string_dictionary_low_cardinality_20/number_distinct_values 1.00 43.7±0.08ms 91.6 MB/sec 1.00 43.5±0.06ms 92.0 MB/sec
string_dictionary_low_cardinality_20/parquet_2 1.00 21.1±0.06ms 189.9 MB/sec 1.00 21.1±0.03ms 189.6 MB/sec
string_dictionary_low_cardinality_20/zstd 1.00 21.3±0.10ms 188.2 MB/sec 1.00 21.3±0.03ms 188.2 MB/sec
string_dictionary_low_cardinality_20/zstd_parquet_2 1.00 21.2±0.06ms 189.0 MB/sec 1.00 21.2±0.04ms 188.5 MB/sec
string_dictionary_low_cardinality_400/bloom_filter 1.01 22.3±0.11ms 179.6 MB/sec 1.00 22.2±0.06ms 180.8 MB/sec
string_dictionary_low_cardinality_400/cdc 1.00 24.1±0.07ms 166.2 MB/sec 1.01 24.4±0.07ms 164.6 MB/sec
string_dictionary_low_cardinality_400/default 1.00 22.3±0.44ms 180.0 MB/sec 1.01 22.5±0.05ms 178.3 MB/sec
string_dictionary_low_cardinality_400/number_distinct_values 1.01 45.8±0.17ms 87.7 MB/sec 1.00 45.1±0.08ms 88.9 MB/sec
string_dictionary_low_cardinality_400/parquet_2 1.00 21.6±0.05ms 185.3 MB/sec 1.02 22.2±0.07ms 181.1 MB/sec
string_dictionary_low_cardinality_400/zstd 1.00 22.7±0.08ms 177.1 MB/sec 1.00 22.7±0.07ms 176.5 MB/sec
string_dictionary_low_cardinality_400/zstd_parquet_2 1.00 22.6±0.08ms 177.6 MB/sec 1.00 22.6±0.07ms 177.5 MB/sec
string_non_null/bloom_filter 1.20 235.1±13.94ms 2.2 GB/sec 1.00 196.2±0.27ms 2.6 GB/sec
string_non_null/cdc 1.07 251.8±11.97ms 2.0 GB/sec 1.00 234.4±0.64ms 2.2 GB/sec
string_non_null/default 1.49 126.3±12.67ms 4.1 GB/sec 1.00 84.9±0.36ms 6.0 GB/sec
string_non_null/number_distinct_values 1.25 305.9±10.88ms 1713.2 MB/sec 1.00 245.1±1.44ms 2.1 GB/sec
string_non_null/parquet_2 1.06 107.8±1.88ms 4.7 GB/sec 1.00 101.9±0.18ms 5.0 GB/sec
string_non_null/zstd 1.09 573.1±5.70ms 914.4 MB/sec 1.00 525.9±0.49ms 996.3 MB/sec
string_non_null/zstd_parquet_2 1.03 496.5±5.52ms 1055.4 MB/sec 1.00 483.8±0.20ms 1083.2 MB/sec
string_ree/bloom_filter 1.75 178.8±7.57ms 91.8 MB/sec 1.00 102.2±0.55ms 160.6 MB/sec
string_ree/cdc 1.31 227.5±2.19ms 72.1 MB/sec 1.00 173.4±0.24ms 94.6 MB/sec
string_ree/default 1.75 134.4±1.57ms 122.1 MB/sec 1.00 76.7±0.37ms 213.9 MB/sec
string_ree/number_distinct_values 1.39 163.4±2.46ms 100.4 MB/sec 1.00 117.6±0.80ms 139.5 MB/sec
string_ree/parquet_2 1.67 127.7±0.60ms 128.5 MB/sec 1.00 76.3±0.30ms 215.1 MB/sec
string_ree/zstd 1.37 181.5±1.72ms 90.4 MB/sec 1.00 132.8±0.23ms 123.5 MB/sec
string_ree/zstd_parquet_2 1.54 141.1±0.56ms 116.3 MB/sec 1.00 91.8±0.36ms 178.6 MB/sec
string_ree_95pct_null/bloom_filter 1.00 25.0±0.06ms 56.3 MB/sec 1.01 25.1±0.07ms 56.0 MB/sec
string_ree_95pct_null/cdc 1.01 32.8±0.07ms 42.8 MB/sec 1.00 32.4±0.07ms 43.5 MB/sec
string_ree_95pct_null/default 1.03 24.8±0.05ms 56.6 MB/sec 1.00 24.1±0.05ms 58.3 MB/sec
string_ree_95pct_null/number_distinct_values 1.03 27.3±0.11ms 51.6 MB/sec 1.00 26.6±0.08ms 52.9 MB/sec
string_ree_95pct_null/parquet_2 1.02 24.6±0.05ms 57.1 MB/sec 1.00 24.1±0.03ms 58.3 MB/sec
string_ree_95pct_null/zstd 1.00 25.4±0.09ms 55.3 MB/sec 1.01 25.6±0.07ms 54.9 MB/sec
string_ree_95pct_null/zstd_parquet_2 1.00 25.3±0.09ms 55.6 MB/sec 1.00 25.3±0.08ms 55.6 MB/sec
struct_all_null/bloom_filter 1.10 380.0±2.62µs 41.4 GB/sec 1.00 346.1±2.16µs 45.5 GB/sec
struct_all_null/cdc 1.03 7.8±0.10ms 2.0 GB/sec 1.00 7.6±0.14ms 2.1 GB/sec
struct_all_null/default 1.28 115.8±0.83µs 136.0 GB/sec 1.00 90.5±0.52µs 173.9 GB/sec
struct_all_null/number_distinct_values 1.28 117.0±0.59µs 134.6 GB/sec 1.00 91.2±0.58µs 172.8 GB/sec
struct_all_null/parquet_2 1.27 117.5±0.86µs 134.0 GB/sec 1.00 92.6±1.21µs 170.1 GB/sec
struct_all_null/zstd 1.32 178.9±0.88µs 88.0 GB/sec 1.00 136.0±0.87µs 115.8 GB/sec
struct_all_null/zstd_parquet_2 1.30 159.2±0.78µs 98.9 GB/sec 1.00 122.1±1.03µs 128.9 GB/sec
struct_non_null/bloom_filter 1.05 38.4±0.12ms 416.4 MB/sec 1.00 36.7±0.25ms 436.0 MB/sec
struct_non_null/cdc 1.01 39.7±0.14ms 403.2 MB/sec 1.00 39.2±0.11ms 408.2 MB/sec
struct_non_null/default 1.01 26.2±0.08ms 611.8 MB/sec 1.00 25.8±0.14ms 619.4 MB/sec
struct_non_null/number_distinct_values 1.10 190.9±4.20ms 83.8 MB/sec 1.00 174.3±0.43ms 91.8 MB/sec
struct_non_null/parquet_2 1.01 32.7±0.08ms 489.3 MB/sec 1.00 32.3±0.10ms 496.0 MB/sec
struct_non_null/zstd 1.00 34.7±0.08ms 460.8 MB/sec 1.03 35.8±0.79ms 447.0 MB/sec
struct_non_null/zstd_parquet_2 1.00 46.4±0.09ms 345.0 MB/sec 1.00 46.4±0.10ms 345.0 MB/sec
struct_sparse_99pct_null/bloom_filter 1.00 6.5±0.08ms 2.4 GB/sec 1.05 6.8±0.07ms 2.3 GB/sec
struct_sparse_99pct_null/cdc 1.00 13.5±0.11ms 1192.2 MB/sec 1.14 15.5±0.12ms 1043.4 MB/sec
struct_sparse_99pct_null/default 1.00 6.0±0.05ms 2.6 GB/sec 1.02 6.1±0.01ms 2.6 GB/sec
struct_sparse_99pct_null/number_distinct_values 1.00 7.4±0.08ms 2.1 GB/sec 1.12 8.3±0.37ms 1948.5 MB/sec
struct_sparse_99pct_null/parquet_2 1.00 5.9±0.06ms 2.7 GB/sec 1.06 6.3±0.06ms 2.5 GB/sec
struct_sparse_99pct_null/zstd 1.00 7.4±0.07ms 2.1 GB/sec 1.06 7.8±0.05ms 2.0 GB/sec
struct_sparse_99pct_null/zstd_parquet_2 1.00 6.7±0.06ms 2.4 GB/sec 1.04 6.9±0.05ms 2.3 GB/sec
```
</details>
<details><summary>Per-runner information</summary>
<details><summary>arrow_writer — shard 1/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-d1b4f5f6-klwp
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065969074-3182-gch9v 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 805.2s |
| Peak memory | 2.9 GiB |
| Avg memory | 2.6 GiB |
| CPU user | 768.4s |
| CPU sys | 33.5s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 780.2s |
| Peak memory | 3.7 GiB |
| Avg memory | 3.2 GiB |
| CPU user | 777.4s |
| CPU sys | 0.6s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 2/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-d1b4f5f6-klwp
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065969074-3183-vv4pq 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 1140.2s |
| Peak memory | 2.9 GiB |
| Avg memory | 2.6 GiB |
| CPU user | 1089.5s |
| CPU sys | 45.9s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 1090.2s |
| Peak memory | 3.7 GiB |
| Avg memory | 3.2 GiB |
| CPU user | 1085.9s |
| CPU sys | 0.4s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 3/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-9ad7f920-gxhw
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065969074-3184-6m6b4 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 1195.3s |
| Peak memory | 2.9 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 1146.2s |
| CPU sys | 43.2s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 1175.3s |
| Peak memory | 3.5 GiB |
| Avg memory | 3.1 GiB |
| CPU user | 1171.8s |
| CPU sys | 0.6s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 4/4</summary>
**Node:** gk3-benchmark-cluster-nap-1p8mzl3a-44333f37-pt5v
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6065969074-3185-scs8b 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 990.2s |
| Peak memory | 2.9 GiB |
| Avg memory | 2.4 GiB |
| CPU user | 925.2s |
| CPU sys | 63.0s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 945.2s |
| Peak memory | 3.5 GiB |
| Avg memory | 3.1 GiB |
| CPU user | 939.1s |
| CPU sys | 0.9s |
| Peak spill | 0 B |
</details>
</details>
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
<!-- benchmark-request:6065969074:finish:db1d5188f75d22c6492c410cb642f39e -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065969074
run benchmark arrow_writer
shards: 4
baseline:
ref: "7fc78ca9ecf907de396119a25bc689527f0fd0ea"
changed:
ref: "b1c118bd52655725911d374aff4d1a0d0b85a4c9"
env:
MIMALLOC_PURGE_DELAY: "-1"
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6083801014
@adriangb @HippoBaro [mimalloc](https://github.com/apache/arrow-rs/pull/11460#issuecomment-6069786064) overall seems to shows less variance than both [glibc's malloc](https://github.com/apache/arrow-rs/pull/11449#issuecomment-6067746393) and [jemalloc](https://github.com/apache/arrow-rs/pull/11449#issuecomment-6068328880) runs:
Metric | Main vs itself | Jemalloc vs itself | Mimalloc vs itself
-- | -- | -- | --
Median drift | 1.52% | 0.78% | 0.85%
95th percentile | 11.46% | 7.14% | 3.87%
Largest drift | 55.8% | 52.3% | 6.1%
Cases above 5% | 48 | 21 | 4
Cases above 10% | 18 | 10 | 0
This is not really about @HippoBaro's point about order-dependent measurement artifacts, but if it shows less drift between runs it could be an overall improvement
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6069094612
🤖 Benchmark starting (GKE) | [trigger](https://github.com/apache/arrow-rs/pull/11460#issuecomment-6069093342)
**Target:** arrow_writer
**Sharding factor: 4** (4 workers).
Comparing b1c118bd52655725911d374aff4d1a0d0b85a4c9 (b1c118bd52655725911d374aff4d1a0d0b85a4c9) to b1c118bd52655725911d374aff4d1a0d0b85a4c9 [diff](https://github.com/apache/arrow-rs/compare/b1c118bd52655725911d374aff4d1a0d0b85a4c9..b1c118bd52655725911d374aff4d1a0d0b85a4c9)
<details><summary>Run configuration</summary>
```yaml
run benchmark arrow_writer
baseline:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
changed:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
shards: 4
```
</details>
Results will be posted when all workers finish.
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
<!-- benchmark-request:6069093342:start:53016f8b0bd0042a3357b8308a08a6c4 -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065967565
🤖 Benchmark starting (GKE) | [trigger](https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065966059)
**Target:** arrow_writer
**Sharding factor: 4** (4 workers).
Comparing b1c118bd52655725911d374aff4d1a0d0b85a4c9 (b1c118bd52655725911d374aff4d1a0d0b85a4c9) to 7fc78ca9ecf907de396119a25bc689527f0fd0ea [diff](https://github.com/apache/arrow-rs/compare/7fc78ca9ecf907de396119a25bc689527f0fd0ea..b1c118bd52655725911d374aff4d1a0d0b85a4c9)
<details><summary>Run configuration</summary>
```yaml
run benchmark arrow_writer
baseline:
ref: 7fc78ca9ecf907de396119a25bc689527f0fd0ea
changed:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
shards: 4
```
</details>
Results will be posted when all workers finish.
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
<!-- benchmark-request:6065966059:start:627c07964d066e3e512f5979860bc66a -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065970747
🤖 Benchmark starting (GKE) | [trigger](https://github.com/apache/arrow-rs/pull/11460#issuecomment-6065969074)
**Target:** arrow_writer
**Sharding factor: 4** (4 workers).
Comparing b1c118bd52655725911d374aff4d1a0d0b85a4c9 (b1c118bd52655725911d374aff4d1a0d0b85a4c9) to 7fc78ca9ecf907de396119a25bc689527f0fd0ea [diff](https://github.com/apache/arrow-rs/compare/7fc78ca9ecf907de396119a25bc689527f0fd0ea..b1c118bd52655725911d374aff4d1a0d0b85a4c9)
<details><summary>Run configuration</summary>
```yaml
run benchmark arrow_writer
baseline:
ref: 7fc78ca9ecf907de396119a25bc689527f0fd0ea
changed:
env:
MIMALLOC_PURGE_DELAY: '-1'
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
shards: 4
```
</details>
Results will be posted when all workers finish.
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
<!-- benchmark-request:6065969074:start:2ee252fed8b81ee5a1b633e6f5f0937d -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "HippoBaro (via GitHub)" <gi...@apache.org>.
HippoBaro commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6159747811
@sdf-jkl I think this is ready to review?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6175672537
> is there any reason to not use this for our other benchmarks as well?
Looking into it rn!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6068967009
> The benchmark bot may not be the best environment to evaluate whether your changes eliminate the order-dependent measurement artifacts, though. I’ll try to find some time to benchmark this on my end and report whether I can still reproduce the large performance drift I otherwise observe on main. Might not happen this week however!
Thanks @HippoBaro that makes sense. I'd run another bench comparing same branch to itself just to see if there's a big difference between the same runs using it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6069786064
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/arrow-rs/pull/11460#issuecomment-6069093342)
**Target:** arrow_writer
Comparing b1c118bd52655725911d374aff4d1a0d0b85a4c9 (b1c118bd52655725911d374aff4d1a0d0b85a4c9) to b1c118bd52655725911d374aff4d1a0d0b85a4c9 [diff](https://github.com/apache/arrow-rs/compare/b1c118bd52655725911d374aff4d1a0d0b85a4c9..b1c118bd52655725911d374aff4d1a0d0b85a4c9)
<details><summary>Run configuration</summary>
```yaml
run benchmark arrow_writer
baseline:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
changed:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
shards: 4
```
</details>
<details><summary>Details</summary>
```
group base changed
----- ---- -------
bool/bloom_filter 1.04 11.4±0.06ms 22.0 MB/sec 1.00 11.0±0.04ms 22.8 MB/sec
bool/cdc 1.04 13.7±0.15ms 18.2 MB/sec 1.00 13.2±0.08ms 19.0 MB/sec
bool/default 1.01 8.9±0.07ms 28.0 MB/sec 1.00 8.9±0.03ms 28.2 MB/sec
bool/number_distinct_values 1.00 18.9±0.12ms 13.3 MB/sec 1.00 18.9±0.03ms 13.2 MB/sec
bool/parquet_2 1.04 11.1±0.06ms 22.6 MB/sec 1.00 10.7±0.03ms 23.5 MB/sec
bool/zstd 1.01 9.5±0.14ms 26.2 MB/sec 1.00 9.5±0.04ms 26.4 MB/sec
bool/zstd_parquet_2 1.04 11.4±0.06ms 21.9 MB/sec 1.00 11.0±0.04ms 22.7 MB/sec
bool_non_null/bloom_filter 1.00 5.0±0.02ms 24.8 MB/sec 1.00 5.0±0.04ms 24.8 MB/sec
bool_non_null/cdc 1.00 3.9±0.05ms 32.3 MB/sec 1.00 3.9±0.04ms 32.3 MB/sec
bool_non_null/default 1.01 2.3±0.02ms 54.7 MB/sec 1.00 2.3±0.01ms 55.0 MB/sec
bool_non_null/number_distinct_values 1.00 15.6±0.04ms 8.0 MB/sec 1.00 15.7±0.12ms 8.0 MB/sec
bool_non_null/parquet_2 1.00 4.7±0.04ms 26.7 MB/sec 1.00 4.7±0.03ms 26.7 MB/sec
bool_non_null/zstd 1.02 2.8±0.02ms 45.2 MB/sec 1.00 2.7±0.02ms 46.1 MB/sec
bool_non_null/zstd_parquet_2 1.00 5.0±0.02ms 24.9 MB/sec 1.01 5.1±0.03ms 24.6 MB/sec
bool_ree/bloom_filter 1.00 28.0±0.30ms 7.6 MB/sec 1.00 28.1±0.34ms 7.6 MB/sec
bool_ree/cdc 1.01 28.8±0.24ms 7.4 MB/sec 1.00 28.5±0.33ms 7.5 MB/sec
bool_ree/default 1.00 25.2±0.29ms 8.4 MB/sec 1.00 25.2±0.29ms 8.4 MB/sec
bool_ree/number_distinct_values 1.00 37.2±0.15ms 5.7 MB/sec 1.00 37.2±0.27ms 5.7 MB/sec
bool_ree/parquet_2 1.00 26.1±0.41ms 8.1 MB/sec 1.00 26.1±0.29ms 8.1 MB/sec
bool_ree/zstd 1.00 26.0±0.46ms 8.2 MB/sec 1.00 26.0±0.32ms 8.2 MB/sec
bool_ree/zstd_parquet_2 1.00 26.3±0.27ms 8.1 MB/sec 1.00 26.4±0.27ms 8.1 MB/sec
decimal/bloom_filter 1.06 140.1±5.63ms 428.2 MB/sec 1.00 132.4±5.13ms 453.1 MB/sec
decimal/cdc 1.00 109.7±4.08ms 547.2 MB/sec 1.00 109.2±3.77ms 549.3 MB/sec
decimal/default 1.00 84.1±4.95ms 713.7 MB/sec 1.00 83.8±6.04ms 715.6 MB/sec
decimal/number_distinct_values 1.00 336.6±15.34ms 178.3 MB/sec 1.00 337.1±16.20ms 178.0 MB/sec
decimal/parquet_2 1.02 173.9±6.96ms 344.9 MB/sec 1.00 171.1±4.73ms 350.7 MB/sec
decimal/zstd 1.02 98.1±5.65ms 611.8 MB/sec 1.00 96.3±5.90ms 622.8 MB/sec
decimal/zstd_parquet_2 1.02 195.0±6.98ms 307.6 MB/sec 1.00 191.8±5.53ms 312.9 MB/sec
fixed_size_binary_ree/bloom_filter 1.00 64.6±1.09ms 15.6 MB/sec 1.03 66.3±3.93ms 15.2 MB/sec
fixed_size_binary_ree/cdc 1.01 75.9±1.90ms 13.3 MB/sec 1.00 75.2±1.70ms 13.4 MB/sec
fixed_size_binary_ree/default 1.01 63.6±3.38ms 15.8 MB/sec 1.00 63.2±3.54ms 15.9 MB/sec
fixed_size_binary_ree/number_distinct_values 1.00 84.3±3.34ms 11.9 MB/sec 1.00 84.3±3.53ms 12.0 MB/sec
fixed_size_binary_ree/parquet_2 1.00 63.5±2.34ms 15.9 MB/sec 1.00 63.6±2.24ms 15.8 MB/sec
fixed_size_binary_ree/zstd 1.00 67.0±3.62ms 15.0 MB/sec 1.00 66.9±4.06ms 15.1 MB/sec
fixed_size_binary_ree/zstd_parquet_2 1.00 64.1±0.64ms 15.7 MB/sec 1.00 64.2±1.04ms 15.7 MB/sec
float_with_nans/bloom_filter 1.03 82.0±2.44ms 170.7 MB/sec 1.00 79.7±1.82ms 175.7 MB/sec
float_with_nans/cdc 1.01 79.1±1.43ms 177.0 MB/sec 1.00 78.0±1.58ms 179.4 MB/sec
float_with_nans/default 1.00 71.3±1.92ms 196.5 MB/sec 1.01 72.2±2.38ms 193.8 MB/sec
float_with_nans/number_distinct_values 1.00 178.4±4.34ms 78.5 MB/sec 1.00 177.5±4.19ms 78.9 MB/sec
float_with_nans/parquet_2 1.02 71.5±2.44ms 195.8 MB/sec 1.00 70.4±1.73ms 198.9 MB/sec
float_with_nans/zstd 1.00 109.6±2.77ms 127.7 MB/sec 1.00 110.1±2.75ms 127.2 MB/sec
float_with_nans/zstd_parquet_2 1.01 109.2±2.50ms 128.2 MB/sec 1.00 107.9±2.51ms 129.7 MB/sec
fsb/bloom_filter 1.03 13.0±0.11ms 1241.4 MB/sec 1.00 12.6±0.11ms 1282.8 MB/sec
fsb/cdc 1.02 14.9±0.12ms 1082.2 MB/sec 1.00 14.5±0.09ms 1108.9 MB/sec
fsb/default 1.00 12.3±0.44ms 1316.0 MB/sec 1.03 12.6±0.43ms 1282.7 MB/sec
fsb/number_distinct_values 1.02 17.1±0.53ms 944.2 MB/sec 1.00 16.7±0.37ms 964.9 MB/sec
fsb/parquet_2 1.00 12.8±0.18ms 1258.4 MB/sec 1.00 12.8±0.15ms 1262.2 MB/sec
fsb/zstd 1.03 13.3±0.11ms 1215.2 MB/sec 1.00 12.9±0.09ms 1246.7 MB/sec
fsb/zstd_parquet_2 1.00 15.0±0.50ms 1071.9 MB/sec 1.02 15.3±0.53ms 1053.2 MB/sec
int32_ree/bloom_filter 1.01 34.6±0.26ms 11.8 MB/sec 1.00 34.3±0.31ms 11.9 MB/sec
int32_ree/cdc 1.01 38.3±0.42ms 10.6 MB/sec 1.00 38.0±0.36ms 10.7 MB/sec
int32_ree/default 1.01 34.0±0.25ms 12.0 MB/sec 1.00 33.7±0.16ms 12.1 MB/sec
int32_ree/number_distinct_values 1.00 50.8±0.34ms 8.0 MB/sec 1.00 50.8±0.30ms 8.0 MB/sec
int32_ree/parquet_2 1.01 33.9±0.28ms 12.0 MB/sec 1.00 33.7±0.12ms 12.1 MB/sec
int32_ree/zstd 1.01 34.9±0.32ms 11.7 MB/sec 1.00 34.5±0.52ms 11.8 MB/sec
int32_ree/zstd_parquet_2 1.01 34.6±0.40ms 11.7 MB/sec 1.00 34.3±0.24ms 11.9 MB/sec
int32_ree_95pct_null/bloom_filter 1.01 21.3±0.15ms 19.1 MB/sec 1.00 21.1±0.06ms 19.2 MB/sec
int32_ree_95pct_null/cdc 1.00 24.3±0.18ms 16.8 MB/sec 1.00 24.3±0.11ms 16.8 MB/sec
int32_ree_95pct_null/default 1.01 21.3±0.10ms 19.1 MB/sec 1.00 21.0±0.05ms 19.3 MB/sec
int32_ree_95pct_null/number_distinct_values 1.00 22.1±0.10ms 18.4 MB/sec 1.00 22.1±0.07ms 18.4 MB/sec
int32_ree_95pct_null/parquet_2 1.00 21.1±0.06ms 19.3 MB/sec 1.00 21.0±0.06ms 19.3 MB/sec
int32_ree_95pct_null/zstd 1.00 21.3±0.07ms 19.1 MB/sec 1.00 21.4±0.08ms 19.0 MB/sec
int32_ree_95pct_null/zstd_parquet_2 1.00 21.5±0.08ms 18.9 MB/sec 1.00 21.5±0.07ms 19.0 MB/sec
large_string_distinct/delta_byte_array 1.04 62.5±4.33ms 4.0 GB/sec 1.00 60.2±2.44ms 4.2 GB/sec
large_string_distinct/plain 1.00 66.8±7.03ms 3.7 GB/sec 1.01 67.7±5.73ms 3.7 GB/sec
large_string_distinct_nullable/delta_byte_array 1.04 58.5±3.77ms 4.3 GB/sec 1.00 56.2±2.86ms 4.5 GB/sec
large_string_non_null/bloom_filter 1.00 48.8±0.15ms 5.1 GB/sec 1.00 48.7±0.19ms 5.1 GB/sec
large_string_non_null/cdc 1.01 237.6±1.02ms 1077.7 MB/sec 1.00 235.8±0.97ms 1085.7 MB/sec
large_string_non_null/default 1.00 50.8±0.63ms 4.9 GB/sec 1.01 51.5±1.01ms 4.9 GB/sec
large_string_non_null/number_distinct_values 1.01 67.0±0.44ms 3.7 GB/sec 1.00 66.3±0.53ms 3.8 GB/sec
large_string_non_null/parquet_2 1.00 50.5±0.26ms 5.0 GB/sec 1.00 50.2±0.43ms 5.0 GB/sec
large_string_non_null/zstd 1.03 51.2±0.62ms 4.9 GB/sec 1.00 49.6±0.25ms 5.0 GB/sec
large_string_non_null/zstd_parquet_2 1.01 49.7±0.15ms 5.0 GB/sec 1.00 49.2±0.15ms 5.1 GB/sec
large_string_shared_prefix/delta_byte_array 1.02 48.0±1.80ms 5.2 GB/sec 1.00 47.2±2.10ms 5.3 GB/sec
large_string_shared_prefix/plain 1.03 107.9±7.37ms 2.3 GB/sec 1.00 104.7±7.45ms 2.4 GB/sec
large_string_shared_prefix_list/delta_byte_array 1.00 63.4±4.22ms 3.9 GB/sec 1.04 65.9±3.40ms 3.8 GB/sec
large_string_shared_prefix_nullable/delta_byte_array 1.00 39.7±0.29ms 6.3 GB/sec 1.00 39.6±0.31ms 6.3 GB/sec
large_string_shared_prefix_nullable/plain 1.00 77.5±8.56ms 3.2 GB/sec 1.04 80.8±8.19ms 3.1 GB/sec
large_string_shared_prefix_nullable_dense/delta_byte_array 1.03 22.1±0.25ms 5.6 GB/sec 1.00 21.6±0.17ms 5.8 GB/sec
large_string_shared_prefix_nullable_trailing/delta_byte_array 1.03 41.1±0.23ms 6.1 GB/sec 1.00 40.0±0.30ms 6.3 GB/sec
list_nested/bloom_filter 1.00 139.7±3.70ms 206.6 MB/sec 1.00 139.9±4.01ms 206.2 MB/sec
list_nested/cdc 1.00 152.8±3.10ms 188.9 MB/sec 1.00 152.4±2.82ms 189.4 MB/sec
list_nested/default 1.00 120.2±2.08ms 240.1 MB/sec 1.00 120.0±1.91ms 240.5 MB/sec
list_nested/number_distinct_values 1.01 269.9±3.87ms 106.9 MB/sec 1.00 267.4±3.30ms 107.9 MB/sec
list_nested/parquet_2 1.00 131.7±3.31ms 219.1 MB/sec 1.00 132.1±3.35ms 218.5 MB/sec
list_nested/zstd 1.00 130.6±2.81ms 220.9 MB/sec 1.00 130.3±2.69ms 221.5 MB/sec
list_nested/zstd_parquet_2 1.00 144.0±2.42ms 200.3 MB/sec 1.01 145.2±5.93ms 198.8 MB/sec
list_primitive/bloom_filter 1.04 284.6±13.12ms 1916.0 MB/sec 1.00 273.3±13.76ms 1995.7 MB/sec
list_primitive/cdc 1.00 329.5±20.70ms 1655.3 MB/sec 1.00 330.3±26.47ms 1651.2 MB/sec
list_primitive/default 1.00 215.1±20.64ms 2.5 GB/sec 1.01 216.7±20.46ms 2.5 GB/sec
list_primitive/number_distinct_values 1.02 402.0±22.21ms 1356.6 MB/sec 1.00 394.1±19.33ms 1383.9 MB/sec
list_primitive/parquet_2 1.00 207.4±5.05ms 2.6 GB/sec 1.01 209.4±10.83ms 2.5 GB/sec
list_primitive/zstd 1.01 474.7±20.56ms 1149.0 MB/sec 1.00 469.8±18.23ms 1160.9 MB/sec
list_primitive/zstd_parquet_2 1.01 440.1±15.29ms 1239.2 MB/sec 1.00 437.4±12.92ms 1246.8 MB/sec
list_primitive_non_null/bloom_filter 1.07 361.5±41.42ms 1505.5 MB/sec 1.00 339.3±29.42ms 1604.0 MB/sec
list_primitive_non_null/cdc 1.01 422.1±40.46ms 1289.4 MB/sec 1.00 419.2±34.50ms 1298.2 MB/sec
list_primitive_non_null/default 1.00 208.2±31.09ms 2.6 GB/sec 1.01 209.9±26.56ms 2.5 GB/sec
list_primitive_non_null/number_distinct_values 1.02 537.4±8.83ms 1012.7 MB/sec 1.00 528.9±14.06ms 1029.0 MB/sec
list_primitive_non_null/parquet_2 1.04 229.9±30.17ms 2.3 GB/sec 1.00 221.0±18.27ms 2.4 GB/sec
list_primitive_non_null/zstd 1.00 749.2±44.45ms 726.5 MB/sec 1.00 745.9±46.45ms 729.6 MB/sec
list_primitive_non_null/zstd_parquet_2 1.01 616.1±7.23ms 883.4 MB/sec 1.00 609.3±4.82ms 893.3 MB/sec
list_primitive_sparse_99pct_null/bloom_filter 1.00 12.7±0.10ms 2.9 GB/sec 1.00 12.7±0.13ms 2.9 GB/sec
list_primitive_sparse_99pct_null/cdc 1.00 24.0±0.21ms 1556.9 MB/sec 1.00 24.0±0.16ms 1554.2 MB/sec
list_primitive_sparse_99pct_null/default 1.00 12.3±0.15ms 3.0 GB/sec 1.00 12.3±0.11ms 3.0 GB/sec
list_primitive_sparse_99pct_null/number_distinct_values 1.00 12.5±0.09ms 2.9 GB/sec 1.00 12.4±0.12ms 2.9 GB/sec
list_primitive_sparse_99pct_null/parquet_2 1.00 13.0±0.22ms 2.8 GB/sec 1.00 13.1±0.21ms 2.8 GB/sec
list_primitive_sparse_99pct_null/zstd 1.00 14.2±0.15ms 2.6 GB/sec 1.00 14.1±0.15ms 2.6 GB/sec
list_primitive_sparse_99pct_null/zstd_parquet_2 1.00 12.5±0.13ms 2.9 GB/sec 1.00 12.6±0.12ms 2.9 GB/sec
list_struct_with_list/bloom_filter 1.00 269.2±6.47ms 169.4 MB/sec 1.00 269.4±6.10ms 169.3 MB/sec
list_struct_with_list/cdc 1.00 306.2±7.82ms 148.9 MB/sec 1.00 304.9±6.85ms 149.5 MB/sec
list_struct_with_list/default 1.00 238.3±7.57ms 191.3 MB/sec 1.00 237.6±7.67ms 191.9 MB/sec
list_struct_with_list/number_distinct_values 1.01 513.8±4.09ms 88.7 MB/sec 1.00 507.3±5.72ms 89.9 MB/sec
list_struct_with_list/parquet_2 1.00 258.7±5.10ms 176.3 MB/sec 1.00 257.6±9.02ms 177.0 MB/sec
list_struct_with_list/zstd 1.00 269.3±8.81ms 169.3 MB/sec 1.01 273.1±9.39ms 167.0 MB/sec
list_struct_with_list/zstd_parquet_2 1.00 277.2±7.16ms 164.5 MB/sec 1.00 276.1±9.19ms 165.2 MB/sec
medium_string_shared_prefix_nullable/delta_byte_array 1.03 38.2±0.16ms 6.5 GB/sec 1.00 37.3±0.21ms 6.7 GB/sec
ndv/bool/ndv_on 1.01 9.8±0.17ms 8.0 MB/sec 1.00 9.8±0.12ms 8.1 MB/sec
ndv/fsb16_high_cardinality/ndv_on 1.00 39.8±0.73ms 251.4 MB/sec 1.01 40.2±0.48ms 248.8 MB/sec
ndv/fsb16_low_cardinality/ndv_on 1.00 35.5±0.34ms 282.0 MB/sec 1.00 35.4±0.31ms 282.2 MB/sec
ndv/int32_high_cardinality/ndv_on 1.00 28.0±0.17ms 89.2 MB/sec 1.00 28.1±0.22ms 89.0 MB/sec
ndv/int32_low_cardinality/ndv_on 1.00 16.1±0.05ms 155.7 MB/sec 1.00 16.1±0.06ms 155.3 MB/sec
ndv/large_binary_high_cardinality/ndv_on 1.01 37.6±0.18ms 199.3 MB/sec 1.00 37.3±0.09ms 200.9 MB/sec
ndv/large_binary_low_cardinality/ndv_on 1.00 25.2±0.11ms 595.1 MB/sec 1.00 25.1±0.11ms 596.7 MB/sec
ndv/utf8_high_cardinality/ndv_on 1.01 68.0±6.21ms 2.3 GB/sec 1.00 67.7±5.38ms 2.3 GB/sec
ndv/utf8_low_cardinality/ndv_on 1.00 24.3±0.08ms 308.3 MB/sec 1.00 24.3±0.08ms 308.7 MB/sec
ndv/utf8_view_high_cardinality/ndv_on 1.01 38.0±0.20ms 263.1 MB/sec 1.00 37.8±0.19ms 264.4 MB/sec
ndv/utf8_view_low_cardinality/ndv_on 1.00 23.8±0.08ms 420.9 MB/sec 1.00 23.6±0.07ms 423.0 MB/sec
primitive/bloom_filter 1.00 140.7±2.16ms 319.0 MB/sec 1.00 141.3±2.43ms 317.6 MB/sec
primitive/cdc 1.04 152.3±3.36ms 294.6 MB/sec 1.00 147.0±2.44ms 305.3 MB/sec
primitive/default 1.03 108.8±1.72ms 412.5 MB/sec 1.00 105.2±2.07ms 426.7 MB/sec
primitive/number_distinct_values 1.01 325.0±4.10ms 138.1 MB/sec 1.00 320.9±4.31ms 139.8 MB/sec
primitive/parquet_2 1.00 117.0±2.01ms 383.7 MB/sec 1.01 118.2±2.19ms 379.8 MB/sec
primitive/zstd 1.03 142.4±3.68ms 315.2 MB/sec 1.00 138.1±2.70ms 325.0 MB/sec
primitive/zstd_parquet_2 1.03 154.1±2.33ms 291.3 MB/sec 1.00 150.2±2.12ms 298.8 MB/sec
primitive_all_null/bloom_filter 1.00 890.0±14.22µs 49.2 GB/sec 1.00 891.8±16.26µs 49.1 GB/sec
primitive_all_null/cdc 1.01 18.6±0.29ms 2.4 GB/sec 1.00 18.5±0.23ms 2.4 GB/sec
primitive_all_null/default 1.00 209.6±1.56µs 209.1 GB/sec 1.00 209.7±1.60µs 209.0 GB/sec
primitive_all_null/number_distinct_values 1.00 211.2±1.56µs 207.5 GB/sec 1.00 210.3±1.58µs 208.4 GB/sec
primitive_all_null/parquet_2 1.00 214.0±2.79µs 204.8 GB/sec 1.00 213.9±2.62µs 204.9 GB/sec
primitive_all_null/zstd 1.00 319.3±2.27µs 137.3 GB/sec 1.01 321.7±2.35µs 136.2 GB/sec
primitive_all_null/zstd_parquet_2 1.00 284.9±2.77µs 153.8 GB/sec 1.00 284.5±2.59µs 154.0 GB/sec
primitive_non_null/bloom_filter 1.00 98.3±1.88ms 447.7 MB/sec 1.04 102.2±1.77ms 430.5 MB/sec
primitive_non_null/cdc 1.00 80.5±1.95ms 546.4 MB/sec 1.02 82.5±2.12ms 533.1 MB/sec
primitive_non_null/default 1.04 59.1±1.32ms 744.1 MB/sec 1.00 57.0±1.68ms 772.2 MB/sec
primitive_non_null/number_distinct_values 1.01 423.0±3.63ms 104.0 MB/sec 1.00 417.8±5.53ms 105.3 MB/sec
primitive_non_null/parquet_2 1.00 73.9±1.49ms 595.5 MB/sec 1.01 75.0±1.71ms 587.0 MB/sec
primitive_non_null/zstd 1.02 91.0±2.65ms 483.3 MB/sec 1.00 89.4±2.09ms 492.2 MB/sec
primitive_non_null/zstd_parquet_2 1.01 108.6±2.28ms 405.2 MB/sec 1.00 107.4±2.31ms 409.6 MB/sec
primitive_sparse_99pct_null/bloom_filter 1.00 13.3±0.21ms 3.3 GB/sec 1.01 13.4±0.19ms 3.3 GB/sec
primitive_sparse_99pct_null/cdc 1.01 31.1±0.35ms 1442.4 MB/sec 1.00 30.9±0.37ms 1452.0 MB/sec
primitive_sparse_99pct_null/default 1.01 10.4±0.05ms 4.2 GB/sec 1.00 10.3±0.04ms 4.2 GB/sec
primitive_sparse_99pct_null/number_distinct_values 1.00 13.9±0.13ms 3.2 GB/sec 1.00 13.8±0.09ms 3.2 GB/sec
primitive_sparse_99pct_null/parquet_2 1.01 10.4±0.04ms 4.2 GB/sec 1.00 10.3±0.05ms 4.2 GB/sec
primitive_sparse_99pct_null/zstd 1.00 14.0±0.09ms 3.1 GB/sec 1.00 14.0±0.09ms 3.1 GB/sec
primitive_sparse_99pct_null/zstd_parquet_2 1.01 12.6±0.07ms 3.5 GB/sec 1.00 12.5±0.07ms 3.5 GB/sec
short_string_non_null/bloom_filter 1.02 29.1±0.35ms 412.1 MB/sec 1.00 28.7±0.36ms 418.4 MB/sec
short_string_non_null/cdc 1.00 20.2±0.23ms 592.9 MB/sec 1.00 20.2±0.21ms 593.1 MB/sec
short_string_non_null/default 1.00 16.4±0.20ms 730.4 MB/sec 1.01 16.6±0.21ms 724.1 MB/sec
short_string_non_null/number_distinct_values 1.00 74.4±2.17ms 161.3 MB/sec 1.00 74.2±2.39ms 161.8 MB/sec
short_string_non_null/parquet_2 1.00 25.8±0.21ms 465.0 MB/sec 1.00 25.8±0.22ms 464.5 MB/sec
short_string_non_null/zstd 1.01 36.6±0.24ms 327.8 MB/sec 1.00 36.3±0.41ms 330.8 MB/sec
short_string_non_null/zstd_parquet_2 1.00 28.7±0.19ms 418.5 MB/sec 1.01 28.8±0.20ms 416.0 MB/sec
small_string_distinct/delta_byte_array 1.01 936.2±10.96µs 8.4 GB/sec 1.00 926.7±8.15µs 8.5 GB/sec
small_string_distinct/plain 1.04 805.1±6.35µs 9.7 GB/sec 1.00 775.3±6.90µs 10.1 GB/sec
small_string_partial_prefix/delta_byte_array 1.02 993.1±5.67µs 7.9 GB/sec 1.00 969.4±6.29µs 8.1 GB/sec
small_string_partial_prefix/plain 1.05 1063.5±7.56µs 7.4 GB/sec 1.00 1009.1±6.40µs 7.8 GB/sec
small_string_shared_prefix/delta_byte_array 1.00 968.8±1.88µs 8.1 GB/sec 1.00 965.5±1.62µs 8.1 GB/sec
small_string_shared_prefix/plain 1.03 1163.5±14.73µs 6.7 GB/sec 1.00 1133.9±11.36µs 6.9 GB/sec
string/bloom_filter 1.00 181.9±15.04ms 2.8 GB/sec 1.01 183.3±18.59ms 2.8 GB/sec
string/cdc 1.00 219.5±18.57ms 2.3 GB/sec 1.00 220.6±23.95ms 2.3 GB/sec
string/default 1.03 89.6±10.54ms 5.7 GB/sec 1.00 87.1±8.07ms 5.9 GB/sec
string/number_distinct_values 1.00 213.9±23.10ms 2.4 GB/sec 1.03 220.6±21.79ms 2.3 GB/sec
string/parquet_2 1.01 95.1±6.18ms 5.4 GB/sec 1.00 93.7±4.69ms 5.5 GB/sec
string/zstd 1.01 468.9±27.67ms 1118.1 MB/sec 1.00 465.5±24.75ms 1126.1 MB/sec
string/zstd_parquet_2 1.01 403.9±24.16ms 1297.9 MB/sec 1.00 400.1±18.07ms 1310.2 MB/sec
string_and_binary_view/bloom_filter 1.03 65.2±0.91ms 494.5 MB/sec 1.00 63.2±0.57ms 510.5 MB/sec
string_and_binary_view/cdc 1.02 59.8±0.89ms 538.9 MB/sec 1.00 58.8±0.79ms 548.3 MB/sec
string_and_binary_view/default 1.02 48.9±0.55ms 660.1 MB/sec 1.00 48.0±0.34ms 672.3 MB/sec
string_and_binary_view/number_distinct_values 1.02 110.7±1.25ms 291.3 MB/sec 1.00 108.9±1.15ms 296.1 MB/sec
string_and_binary_view/parquet_2 1.01 61.8±1.86ms 522.1 MB/sec 1.00 61.2±1.53ms 527.1 MB/sec
string_and_binary_view/zstd 1.02 87.1±1.47ms 370.4 MB/sec 1.00 85.7±0.95ms 376.2 MB/sec
string_and_binary_view/zstd_parquet_2 1.00 76.0±1.48ms 424.1 MB/sec 1.01 76.5±1.54ms 421.7 MB/sec
string_dictionary/bloom_filter 1.05 87.2±5.92ms 3.0 GB/sec 1.00 83.0±4.88ms 3.1 GB/sec
string_dictionary/cdc 1.03 49.6±3.95ms 5.2 GB/sec 1.00 48.1±2.86ms 5.4 GB/sec
string_dictionary/default 1.02 43.1±1.76ms 6.0 GB/sec 1.00 42.4±2.36ms 6.1 GB/sec
string_dictionary/number_distinct_values 1.00 108.3±8.96ms 2.4 GB/sec 1.01 109.3±9.81ms 2.4 GB/sec
string_dictionary/parquet_2 1.01 51.0±7.83ms 5.1 GB/sec 1.00 50.7±4.68ms 5.1 GB/sec
string_dictionary/zstd 1.01 216.3±12.59ms 1221.1 MB/sec 1.00 214.1±11.71ms 1233.5 MB/sec
string_dictionary/zstd_parquet_2 1.01 195.8±8.14ms 1348.8 MB/sec 1.00 194.6±7.33ms 1357.2 MB/sec
string_dictionary_low_cardinality_100/bloom_filter 1.00 21.8±0.06ms 183.7 MB/sec 1.00 21.9±0.06ms 183.2 MB/sec
string_dictionary_low_cardinality_100/cdc 1.02 24.4±0.14ms 164.3 MB/sec 1.00 23.9±0.11ms 167.3 MB/sec
string_dictionary_low_cardinality_100/default 1.00 21.7±0.05ms 184.9 MB/sec 1.00 21.8±0.07ms 184.1 MB/sec
string_dictionary_low_cardinality_100/number_distinct_values 1.00 44.7±0.79ms 89.6 MB/sec 1.01 45.0±0.88ms 89.1 MB/sec
string_dictionary_low_cardinality_100/parquet_2 1.01 21.9±0.06ms 182.8 MB/sec 1.00 21.7±0.10ms 184.7 MB/sec
string_dictionary_low_cardinality_100/zstd 1.00 22.1±0.05ms 181.6 MB/sec 1.00 22.1±0.06ms 181.5 MB/sec
string_dictionary_low_cardinality_100/zstd_parquet_2 1.00 22.0±0.05ms 181.7 MB/sec 1.00 22.1±0.07ms 181.4 MB/sec
string_dictionary_low_cardinality_20/bloom_filter 1.00 21.0±0.06ms 190.3 MB/sec 1.03 21.6±0.14ms 185.5 MB/sec
string_dictionary_low_cardinality_20/cdc 1.00 23.4±0.12ms 171.4 MB/sec 1.02 23.8±0.12ms 168.5 MB/sec
string_dictionary_low_cardinality_20/default 1.00 21.0±0.10ms 191.1 MB/sec 1.02 21.4±0.11ms 186.8 MB/sec
string_dictionary_low_cardinality_20/number_distinct_values 1.00 44.1±0.78ms 90.8 MB/sec 1.01 44.6±0.88ms 89.9 MB/sec
string_dictionary_low_cardinality_20/parquet_2 1.03 21.7±0.13ms 184.9 MB/sec 1.00 21.1±0.10ms 189.9 MB/sec
string_dictionary_low_cardinality_20/zstd 1.03 21.8±0.13ms 183.6 MB/sec 1.00 21.2±0.10ms 188.6 MB/sec
string_dictionary_low_cardinality_20/zstd_parquet_2 1.03 21.8±0.14ms 183.7 MB/sec 1.00 21.2±0.10ms 188.9 MB/sec
string_dictionary_low_cardinality_400/bloom_filter 1.01 22.2±0.11ms 180.6 MB/sec 1.00 22.0±0.07ms 182.4 MB/sec
string_dictionary_low_cardinality_400/cdc 1.01 24.5±0.14ms 163.9 MB/sec 1.00 24.3±0.09ms 165.1 MB/sec
string_dictionary_low_cardinality_400/default 1.01 22.3±0.22ms 179.7 MB/sec 1.00 22.1±0.23ms 181.4 MB/sec
string_dictionary_low_cardinality_400/number_distinct_values 1.01 45.1±0.69ms 88.9 MB/sec 1.00 44.8±0.74ms 89.6 MB/sec
string_dictionary_low_cardinality_400/parquet_2 1.00 21.9±0.06ms 183.3 MB/sec 1.00 21.9±0.09ms 183.1 MB/sec
string_dictionary_low_cardinality_400/zstd 1.01 22.7±0.11ms 176.8 MB/sec 1.00 22.6±0.16ms 177.9 MB/sec
string_dictionary_low_cardinality_400/zstd_parquet_2 1.02 22.8±0.15ms 175.8 MB/sec 1.00 22.4±0.05ms 178.9 MB/sec
string_non_null/bloom_filter 1.05 212.2±14.16ms 2.4 GB/sec 1.00 201.4±14.25ms 2.5 GB/sec
string_non_null/cdc 1.04 266.1±39.44ms 1968.8 MB/sec 1.00 254.9±22.89ms 2.0 GB/sec
string_non_null/default 1.01 102.0±20.01ms 5.0 GB/sec 1.00 100.6±15.46ms 5.1 GB/sec
string_non_null/number_distinct_values 1.01 300.9±37.72ms 1741.5 MB/sec 1.00 297.2±33.95ms 1762.8 MB/sec
string_non_null/parquet_2 1.00 115.7±18.70ms 4.4 GB/sec 1.00 115.8±20.39ms 4.4 GB/sec
string_non_null/zstd 1.00 608.3±42.56ms 861.4 MB/sec 1.00 608.7±42.88ms 860.9 MB/sec
string_non_null/zstd_parquet_2 1.02 506.6±28.42ms 1034.3 MB/sec 1.00 494.5±11.55ms 1059.6 MB/sec
string_ree/bloom_filter 1.05 107.9±8.38ms 152.0 MB/sec 1.00 103.3±7.19ms 158.9 MB/sec
string_ree/cdc 1.00 193.1±15.33ms 85.0 MB/sec 1.01 194.4±21.30ms 84.4 MB/sec
string_ree/default 1.02 90.4±15.88ms 181.5 MB/sec 1.00 89.0±12.27ms 184.4 MB/sec
string_ree/number_distinct_values 1.01 127.2±11.78ms 129.0 MB/sec 1.00 126.0±12.90ms 130.2 MB/sec
string_ree/parquet_2 1.00 80.6±1.50ms 203.6 MB/sec 1.01 81.6±9.21ms 201.1 MB/sec
string_ree/zstd 1.02 138.3±1.92ms 118.6 MB/sec 1.00 135.4±2.19ms 121.1 MB/sec
string_ree/zstd_parquet_2 1.00 93.8±1.51ms 174.8 MB/sec 1.00 93.5±1.29ms 175.6 MB/sec
string_ree_95pct_null/bloom_filter 1.00 25.1±0.46ms 56.0 MB/sec 1.00 25.0±0.42ms 56.3 MB/sec
string_ree_95pct_null/cdc 1.00 33.3±0.44ms 42.3 MB/sec 1.00 33.3±0.34ms 42.3 MB/sec
string_ree_95pct_null/default 1.00 24.3±0.22ms 57.8 MB/sec 1.00 24.4±0.24ms 57.6 MB/sec
string_ree_95pct_null/number_distinct_values 1.00 27.4±0.41ms 51.4 MB/sec 1.00 27.5±0.45ms 51.2 MB/sec
string_ree_95pct_null/parquet_2 1.00 24.4±0.21ms 57.8 MB/sec 1.00 24.4±0.26ms 57.8 MB/sec
string_ree_95pct_null/zstd 1.00 25.5±0.41ms 55.2 MB/sec 1.01 25.7±0.46ms 54.8 MB/sec
string_ree_95pct_null/zstd_parquet_2 1.02 25.5±0.37ms 55.2 MB/sec 1.00 25.0±0.25ms 56.3 MB/sec
struct_all_null/bloom_filter 1.00 360.5±5.30µs 43.7 GB/sec 1.01 364.8±4.75µs 43.2 GB/sec
struct_all_null/cdc 1.00 7.5±0.06ms 2.1 GB/sec 1.00 7.5±0.06ms 2.1 GB/sec
struct_all_null/default 1.00 91.4±0.58µs 172.3 GB/sec 1.00 91.3±0.58µs 172.5 GB/sec
struct_all_null/number_distinct_values 1.00 91.6±0.62µs 172.0 GB/sec 1.00 91.5±0.62µs 172.1 GB/sec
struct_all_null/parquet_2 1.00 93.1±1.29µs 169.2 GB/sec 1.00 92.9±1.26µs 169.5 GB/sec
struct_all_null/zstd 1.00 137.3±0.85µs 114.7 GB/sec 1.00 137.6±1.03µs 114.4 GB/sec
struct_all_null/zstd_parquet_2 1.00 122.9±1.09µs 128.1 GB/sec 1.00 123.1±1.18µs 127.9 GB/sec
struct_non_null/bloom_filter 1.03 42.3±0.75ms 378.1 MB/sec 1.00 41.0±0.84ms 390.7 MB/sec
struct_non_null/cdc 1.00 40.5±0.97ms 394.9 MB/sec 1.01 41.0±0.95ms 390.5 MB/sec
struct_non_null/default 1.02 27.0±0.58ms 592.4 MB/sec 1.00 26.4±0.33ms 605.5 MB/sec
struct_non_null/number_distinct_values 1.00 201.8±6.68ms 79.3 MB/sec 1.01 203.4±6.96ms 78.7 MB/sec
struct_non_null/parquet_2 1.01 33.4±0.24ms 479.2 MB/sec 1.00 33.0±0.56ms 484.8 MB/sec
struct_non_null/zstd 1.01 36.2±0.95ms 441.7 MB/sec 1.00 36.0±0.64ms 444.0 MB/sec
struct_non_null/zstd_parquet_2 1.00 47.8±0.88ms 334.9 MB/sec 1.01 48.3±0.70ms 331.5 MB/sec
struct_sparse_99pct_null/bloom_filter 1.00 7.6±0.16ms 2.1 GB/sec 1.01 7.7±0.11ms 2.1 GB/sec
struct_sparse_99pct_null/cdc 1.00 14.6±0.15ms 1104.6 MB/sec 1.00 14.6±0.14ms 1107.8 MB/sec
struct_sparse_99pct_null/default 1.00 6.7±0.11ms 2.4 GB/sec 1.00 6.7±0.08ms 2.4 GB/sec
struct_sparse_99pct_null/number_distinct_values 1.00 7.5±0.09ms 2.1 GB/sec 1.00 7.5±0.09ms 2.1 GB/sec
struct_sparse_99pct_null/parquet_2 1.00 6.1±0.05ms 2.6 GB/sec 1.01 6.2±0.06ms 2.5 GB/sec
struct_sparse_99pct_null/zstd 1.00 7.5±0.10ms 2.1 GB/sec 1.01 7.6±0.07ms 2.1 GB/sec
struct_sparse_99pct_null/zstd_parquet_2 1.00 7.4±0.15ms 2.1 GB/sec 1.00 7.5±0.10ms 2.1 GB/sec
```
</details>
<details><summary>Per-runner information</summary>
<details><summary>arrow_writer — shard 1/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-963d8389-bvh2
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6069093342-3213-g9dxb 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 795.2s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.4 GiB |
| CPU user | 773.9s |
| CPU sys | 19.1s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 790.2s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.4 GiB |
| CPU user | 771.1s |
| CPU sys | 16.8s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 2/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-59aaa3df-xbl7
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6069093342-3214-66r9v 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 1155.3s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 1107.0s |
| CPU sys | 44.4s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 1155.3s |
| Peak memory | 3.1 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 1105.8s |
| CPU sys | 44.6s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 3/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-9ad7f920-rj49
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6069093342-3215-85b49 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 1200.3s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 1170.2s |
| CPU sys | 25.3s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 1185.3s |
| Peak memory | 3.0 GiB |
| Avg memory | 2.5 GiB |
| CPU user | 1158.3s |
| CPU sys | 22.1s |
| Peak spill | 0 B |
</details>
</details>
<details><summary>arrow_writer — shard 4/4</summary>
**Node:** gk3-benchmark-cluster-nap-101n84r4-9ad7f920-6m4m
**Instance:** c4a-highmem-16 (12 vCPU / 65 GiB)
**uname:**
```text
Linux bench-c6069093342-3216-wwqd4 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux
```
**BENCH_COMMAND:**
```sh
cargo bench --features=arrow,async,test_common,experimental,object_store --bench arrow_writer
```
<details><summary>CPU Details (lscpu)</summary>
```text
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Resource Usage</summary>
**base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 960.2s |
| Peak memory | 3.1 GiB |
| Avg memory | 2.3 GiB |
| CPU user | 942.8s |
| CPU sys | 12.4s |
| Peak spill | 0 B |
**branch**
| Metric | Value |
|--------|-------|
| Wall time | 950.2s |
| Peak memory | 3.1 GiB |
| Avg memory | 2.3 GiB |
| CPU user | 934.2s |
| CPU sys | 12.0s |
| Peak spill | 0 B |
</details>
</details>
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
<!-- benchmark-request:6069093342:finish:94c258e0c281c66493123f9a6c69da8d -->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6069093342
run benchmark arrow_writer
shards: 4
baseline:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
changed:
ref: b1c118bd52655725911d374aff4d1a0d0b85a4c9
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6068970219
Could not resolve immutable commits for sharding (https://github.com/apache/arrow-rs/pull/11460#issuecomment-6068969422): GitHub API GET error 422 Unprocessable Entity: {"message":"No commit found for SHA: https://github.com/apache/arrow-rs/pull/11460/commits/b1c118bd52655725911d374aff4d1a0d0b85a4c9","documentation_url":"https://docs.github.com/rest/commits/commits#get-a-commit","status":"422"}
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6173982110
@Jefffrey @Rich-T-kid PTAL
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl commented on PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460#issuecomment-6170401189
Our friends from @vortex-data seem to do the same https://github.com/vortex-data/vortex/pull/10483
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [PR] bench(parquet): try mimalloc for Arrow writer benchmarks [arrow-rs]
Posted by "sdf-jkl (via GitHub)" <gi...@apache.org>.
sdf-jkl merged PR #11460:
URL: https://github.com/apache/arrow-rs/pull/11460
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org