kkewwei opened a new pull request, #14397:
URL: https://github.com/apache/lucene/pull/14397
### Description
As mentioned in #14347, we use `LZ4WithPresetDictDecompressor` to
decompress, we will always read preset dict for every doc in non-merge
scenarios. If two consecutive documents fall into the same chunk, we can reuse
the same dictionary. This is a lossless optimization, the cached preset dict
dictionary does not take up additional memory.
Lucene benchmark: python3 src/python/localrun.py wikimediumall
Hardware used: linux ecs.t2-c1m2dev.8xlarge | 32 cores | 64G
In order to obtain the exact performance improvement, I tested it 6 times in
total, and the results are as follows:
Test1:
```
TaskQPS baseline StdDevQPS
my_modified_version StdDev Pct diff p-value
HighTermMonthSort 951.76 (10.5%) 897.70
(8.7%) -5.7% ( -22% - 15%) 0.061
BrowseDayOfYearSSDVFacets 3.41 (19.3%) 3.27
(17.9%) -4.2% ( -34% - 40%) 0.474
BrowseDateSSDVFacets 0.65 (15.0%) 0.62
(12.4%) -3.9% ( -27% - 27%) 0.372
OrHighNotHigh 181.95 (6.2%) 175.52
(5.1%) -3.5% ( -13% - 8%) 0.048
range 2164.58 (7.7%) 2098.62
(6.5%) -3.0% ( -16% - 12%) 0.178
OrHighHigh 18.62 (5.0%) 18.20
(5.7%) -2.3% ( -12% - 8%) 0.179
HighTermTitleBDVSort 2.15 (4.1%) 2.11
(4.0%) -2.0% ( -9% - 6%) 0.122
LowTerm 262.36 (7.1%) 257.53
(6.6%) -1.8% ( -14% - 12%) 0.397
HighIntervalsOrdered 0.42 (5.9%) 0.41
(5.6%) -1.4% ( -12% - 10%) 0.429
OrNotHighHigh 137.47 (4.0%) 135.52
(3.9%) -1.4% ( -8% - 6%) 0.257
HighSpanNear 5.50 (2.1%) 5.43
(2.2%) -1.4% ( -5% - 3%) 0.044
HighTermTitleSort 61.59 (5.9%) 60.76
(6.3%) -1.4% ( -12% - 11%) 0.485
OrHighLow 128.91 (4.9%) 127.20
(4.7%) -1.3% ( -10% - 8%) 0.382
OrHighNotMed 167.55 (6.0%) 165.47
(6.2%) -1.2% ( -12% - 11%) 0.521
HighTerm 170.81 (6.9%) 168.74
(5.9%) -1.2% ( -13% - 12%) 0.548
OrNotHighMed 124.94 (3.5%) 123.44
(5.3%) -1.2% ( -9% - 7%) 0.398
MedTermDayTaxoFacets 9.20 (2.8%) 9.10
(2.9%) -1.1% ( -6% - 4%) 0.230
OrNotHighLow 179.84 (4.5%) 177.97
(3.7%) -1.0% ( -8% - 7%) 0.427
OrHighMedDayTaxoFacets 2.43 (5.0%) 2.41
(3.5%) -1.0% ( -9% - 7%) 0.474
OrHighMed 40.19 (3.9%) 39.81
(2.4%) -1.0% ( -7% - 5%) 0.347
AndHighMed 26.18 (2.2%) 25.94
(3.0%) -0.9% ( -5% - 4%) 0.268
IntNRQ 17.99 (6.2%) 17.83
(6.0%) -0.9% ( -12% - 12%) 0.652
MedSpanNear 23.26 (3.0%) 23.08
(2.5%) -0.8% ( -6% - 4%) 0.372
Wildcard 28.13 (4.5%) 27.94
(4.2%) -0.7% ( -8% - 8%) 0.633
Prefix3 327.18 (9.4%) 325.10
(8.4%) -0.6% ( -16% - 18%) 0.821
OrHighNotLow 200.50 (5.6%) 199.53
(4.2%) -0.5% ( -9% - 9%) 0.754
AndHighLow 297.86 (6.7%) 296.43
(7.5%) -0.5% ( -13% - 14%) 0.832
TermDTSort 90.01 (6.6%) 89.67
(4.3%) -0.4% ( -10% - 11%) 0.831
LowIntervalsOrdered 2.83 (2.9%) 2.83
(3.0%) -0.2% ( -5% - 5%) 0.855
HighSloppyPhrase 4.48 (3.3%) 4.48
(4.0%) -0.1% ( -7% - 7%) 0.903
AndHighHigh 18.78 (2.9%) 18.77
(2.5%) -0.1% ( -5% - 5%) 0.919
LowSpanNear 32.35 (3.0%) 32.36
(2.8%) 0.0% ( -5% - 5%) 0.993
MedTerm 325.56 (5.8%) 325.60
(6.0%) 0.0% ( -11% - 12%) 0.994
LowPhrase 4.37 (3.8%) 4.38
(4.3%) 0.1% ( -7% - 8%) 0.933
AndHighHighDayTaxoFacets 3.50 (1.8%) 3.51
(2.6%) 0.2% ( -4% - 4%) 0.751
PKLookup 103.50 (4.6%) 103.98
(4.2%) 0.5% ( -7% - 9%) 0.742
HighPhrase 39.74 (5.6%) 39.95
(5.3%) 0.5% ( -9% - 12%) 0.760
LowSloppyPhrase 26.29 (2.6%) 26.47
(2.8%) 0.7% ( -4% - 6%) 0.419
MedPhrase 61.27 (5.9%) 61.71
(4.8%) 0.7% ( -9% - 12%) 0.673
AndHighMedDayTaxoFacets 12.94 (2.9%) 13.07
(2.3%) 0.9% ( -4% - 6%) 0.249
MedSloppyPhrase 35.28 (2.8%) 35.63
(2.4%) 1.0% ( -4% - 6%) 0.225
HighTermDayOfYearSort 108.74 (4.3%) 109.82
(4.5%) 1.0% ( -7% - 10%) 0.475
MedIntervalsOrdered 13.63 (5.9%) 13.79
(3.9%) 1.1% ( -8% - 11%) 0.475
BrowseRandomLabelSSDVFacets 1.78 (9.6%) 1.80
(8.6%) 1.2% ( -15% - 21%) 0.686
Fuzzy1 35.79 (4.1%) 36.21
(4.5%) 1.2% ( -7% - 10%) 0.380
Fuzzy2 31.10 (4.9%) 31.60
(3.8%) 1.6% ( -6% - 10%) 0.250
Respell 28.75 (5.1%) 29.23
(3.2%) 1.7% ( -6% - 10%) 0.213
BrowseMonthSSDVFacets 3.32 (11.0%) 3.39
(17.4%) 2.0% ( -23% - 34%) 0.661
BrowseDateTaxoFacets 4.14 (40.9%) 4.26
(46.1%) 2.9% ( -59% - 152%) 0.831
BrowseDayOfYearTaxoFacets 4.30 (41.2%) 4.44
(49.6%) 3.2% ( -61% - 159%) 0.823
BrowseRandomLabelTaxoFacets 2.57 (35.2%) 2.68
(38.5%) 4.4% ( -51% - 120%) 0.705
BrowseMonthTaxoFacets 4.08 (43.3%) 5.16
(65.5%) 26.6% ( -57% - 239%) 0.131
```
Test2:
```
TaskQPS baseline StdDevQPS
my_modified_version StdDev Pct diff p-value
PKLookup 107.53 (3.9%) 104.81
(4.5%) -2.5% ( -10% - 6%) 0.056
BrowseDateSSDVFacets 0.69 (16.5%) 0.67
(15.2%) -1.9% ( -28% - 35%) 0.708
Wildcard 27.49 (4.5%) 27.11
(4.0%) -1.4% ( -9% - 7%) 0.311
HighTermMonthSort 1057.35 (8.3%) 1044.09
(9.6%) -1.3% ( -17% - 18%) 0.659
HighIntervalsOrdered 5.42 (7.8%) 5.35
(7.5%) -1.2% ( -15% - 15%) 0.612
Fuzzy2 31.65 (2.5%) 31.29
(3.4%) -1.2% ( -6% - 4%) 0.219
IntNRQ 34.85 (3.5%) 34.47
(4.6%) -1.1% ( -8% - 7%) 0.387
LowSpanNear 3.88 (2.1%) 3.85
(2.7%) -0.9% ( -5% - 4%) 0.253
Respell 19.33 (2.6%) 19.18
(3.8%) -0.8% ( -6% - 5%) 0.435
OrHighMed 59.82 (2.5%) 59.42
(4.0%) -0.7% ( -7% - 6%) 0.531
AndHighMedDayTaxoFacets 15.30 (3.5%) 15.20
(3.4%) -0.7% ( -7% - 6%) 0.549
Fuzzy1 36.34 (3.6%) 36.12
(4.5%) -0.6% ( -8% - 7%) 0.647
Prefix3 63.78 (5.5%) 63.43
(5.3%) -0.6% ( -10% - 10%) 0.743
MedPhrase 36.74 (4.3%) 36.54
(3.6%) -0.5% ( -8% - 7%) 0.663
TermDTSort 88.62 (4.4%) 88.16
(4.4%) -0.5% ( -8% - 8%) 0.708
MedSloppyPhrase 2.14 (2.7%) 2.13
(3.2%) -0.5% ( -6% - 5%) 0.591
HighTermDayOfYearSort 114.76 (5.3%) 114.29
(4.8%) -0.4% ( -9% - 10%) 0.795
OrHighHigh 15.01 (4.6%) 14.95
(3.7%) -0.4% ( -8% - 8%) 0.788
OrHighNotHigh 141.36 (4.3%) 140.94
(4.1%) -0.3% ( -8% - 8%) 0.820
HighPhrase 34.33 (5.6%) 34.25
(6.1%) -0.2% ( -11% - 12%) 0.901
MedIntervalsOrdered 10.06 (5.8%) 10.04
(6.3%) -0.2% ( -11% - 12%) 0.924
OrNotHighLow 333.78 (5.1%) 333.23
(5.7%) -0.2% ( -10% - 11%) 0.922
MedTermDayTaxoFacets 8.98 (2.8%) 8.98
(4.1%) -0.1% ( -6% - 7%) 0.939
HighSloppyPhrase 7.60 (2.7%) 7.59
(3.4%) -0.0% ( -5% - 6%) 0.972
HighSpanNear 4.69 (4.1%) 4.69
(2.1%) -0.0% ( -5% - 6%) 0.974
MedSpanNear 4.58 (3.1%) 4.58
(2.8%) -0.0% ( -5% - 6%) 0.995
LowTerm 207.87 (7.7%) 207.88
(6.3%) 0.0% ( -13% - 15%) 0.998
MedTerm 245.93 (6.4%) 245.97
(3.1%) 0.0% ( -8% - 10%) 0.992
AndHighHigh 22.64 (3.1%) 22.65
(2.9%) 0.1% ( -5% - 6%) 0.948
AndHighHighDayTaxoFacets 5.66 (2.5%) 5.67
(2.4%) 0.2% ( -4% - 5%) 0.822
OrHighNotMed 222.59 (5.5%) 223.01
(4.5%) 0.2% ( -9% - 10%) 0.906
LowSloppyPhrase 23.25 (2.9%) 23.34
(3.0%) 0.4% ( -5% - 6%) 0.680
OrHighMedDayTaxoFacets 3.30 (3.6%) 3.31
(3.9%) 0.4% ( -6% - 8%) 0.741
LowIntervalsOrdered 11.12 (2.9%) 11.16
(3.3%) 0.4% ( -5% - 6%) 0.683
HighTermTitleSort 26.88 (5.6%) 26.98
(4.3%) 0.4% ( -9% - 10%) 0.797
AndHighMed 57.43 (4.5%) 57.74
(3.3%) 0.5% ( -7% - 8%) 0.670
OrNotHighHigh 131.23 (5.6%) 131.98
(4.7%) 0.6% ( -9% - 11%) 0.727
LowPhrase 50.84 (3.4%) 51.44
(4.2%) 1.2% ( -6% - 9%) 0.335
HighTermTitleBDVSort 4.59 (3.1%) 4.64
(3.5%) 1.2% ( -5% - 8%) 0.246
HighTerm 174.02 (6.6%) 176.17
(3.4%) 1.2% ( -8% - 12%) 0.459
OrHighNotLow 167.90 (5.0%) 170.06
(5.0%) 1.3% ( -8% - 11%) 0.418
OrNotHighMed 127.88 (3.7%) 129.60
(4.0%) 1.3% ( -6% - 9%) 0.274
OrHighLow 170.29 (4.5%) 172.60
(2.9%) 1.4% ( -5% - 9%) 0.253
BrowseDayOfYearSSDVFacets 3.33 (16.5%) 3.38
(16.2%) 1.5% ( -26% - 41%) 0.770
BrowseMonthSSDVFacets 3.28 (8.9%) 3.33
(8.6%) 1.6% ( -14% - 21%) 0.567
AndHighLow 323.35 (4.9%) 329.17
(5.2%) 1.8% ( -7% - 12%) 0.262
range 2173.60 (8.0%) 2219.00
(8.5%) 2.1% ( -13% - 20%) 0.425
BrowseRandomLabelSSDVFacets 1.75 (6.9%) 1.81
(9.4%) 3.1% ( -12% - 20%) 0.232
BrowseRandomLabelTaxoFacets 2.70 (33.1%) 2.95
(37.9%) 9.0% ( -46% - 119%) 0.422
BrowseDateTaxoFacets 4.31 (38.7%) 4.77
(47.8%) 10.7% ( -54% - 158%) 0.438
BrowseDayOfYearTaxoFacets 4.45 (39.4%) 4.93
(49.3%) 10.9% ( -55% - 164%) 0.438
BrowseMonthTaxoFacets 4.48 (47.7%) 5.80
(60.7%) 29.3% ( -53% - 263%) 0.090
```
Test3:
```
TaskQPS baseline StdDevQPS
my_modified_version StdDev Pct diff p-value
BrowseDayOfYearSSDVFacets 3.39 (14.2%) 3.18
(6.5%) -6.1% ( -23% - 17%) 0.083
BrowseMonthSSDVFacets 3.61 (17.0%) 3.41
(12.3%) -5.4% ( -29% - 28%) 0.250
range 2205.30 (8.2%) 2119.30
(11.4%) -3.9% ( -21% - 17%) 0.214
Fuzzy1 38.53 (3.6%) 37.93
(3.8%) -1.6% ( -8% - 6%) 0.177
MedIntervalsOrdered 15.14 (4.3%) 14.99
(4.6%) -1.0% ( -9% - 8%) 0.477
HighTermTitleSort 89.49 (4.2%) 88.63
(5.4%) -1.0% ( -10% - 9%) 0.530
OrHighNotHigh 199.21 (4.4%) 197.48
(3.9%) -0.9% ( -8% - 7%) 0.510
Respell 27.65 (3.8%) 27.42
(3.6%) -0.9% ( -7% - 6%) 0.460
HighTermDayOfYearSort 125.46 (6.1%) 124.48
(3.5%) -0.8% ( -9% - 9%) 0.622
HighSloppyPhrase 4.01 (3.2%) 3.98
(3.1%) -0.7% ( -6% - 5%) 0.457
MedTerm 227.57 (4.2%) 225.90
(3.2%) -0.7% ( -7% - 7%) 0.537
MedPhrase 23.76 (5.6%) 23.59
(4.1%) -0.7% ( -9% - 9%) 0.650
MedTermDayTaxoFacets 7.47 (2.3%) 7.43
(2.5%) -0.5% ( -5% - 4%) 0.488
OrHighMed 40.60 (2.8%) 40.40
(4.6%) -0.5% ( -7% - 7%) 0.684
AndHighHigh 16.67 (2.7%) 16.61
(3.0%) -0.3% ( -5% - 5%) 0.715
BrowseDateSSDVFacets 0.63 (14.5%) 0.63
(14.4%) -0.3% ( -25% - 33%) 0.949
OrHighMedDayTaxoFacets 1.26 (4.5%) 1.26
(3.4%) -0.3% ( -7% - 7%) 0.820
MedSloppyPhrase 2.91 (2.5%) 2.90
(2.5%) -0.2% ( -5% - 4%) 0.753
BrowseRandomLabelSSDVFacets 1.81 (8.3%) 1.81
(7.8%) -0.2% ( -15% - 17%) 0.952
HighPhrase 31.29 (8.0%) 31.25
(4.7%) -0.1% ( -11% - 13%) 0.944
AndHighHighDayTaxoFacets 3.85 (2.1%) 3.85
(2.3%) -0.1% ( -4% - 4%) 0.854
HighIntervalsOrdered 2.24 (5.2%) 2.24
(5.0%) -0.1% ( -9% - 10%) 0.974
TermDTSort 84.23 (5.2%) 84.25
(4.6%) 0.0% ( -9% - 10%) 0.990
Fuzzy2 35.64 (4.4%) 35.65
(3.5%) 0.0% ( -7% - 8%) 0.981
OrHighLow 148.60 (3.7%) 148.71
(5.2%) 0.1% ( -8% - 9%) 0.956
HighTermMonthSort 905.42 (6.4%) 906.40
(5.7%) 0.1% ( -11% - 12%) 0.955
LowSpanNear 7.78 (2.2%) 7.80
(2.2%) 0.2% ( -4% - 4%) 0.749
LowSloppyPhrase 12.79 (2.4%) 12.83
(1.2%) 0.2% ( -3% - 3%) 0.683
Wildcard 26.08 (3.3%) 26.16
(4.0%) 0.3% ( -6% - 7%) 0.794
AndHighMed 43.27 (3.8%) 43.44
(3.0%) 0.4% ( -6% - 7%) 0.718
Prefix3 89.79 (5.0%) 90.22
(4.8%) 0.5% ( -8% - 10%) 0.757
IntNRQ 78.48 (6.9%) 78.88
(5.7%) 0.5% ( -11% - 14%) 0.801
AndHighMedDayTaxoFacets 17.39 (2.7%) 17.49
(2.5%) 0.5% ( -4% - 5%) 0.519
MedSpanNear 9.92 (3.3%) 9.98
(2.1%) 0.6% ( -4% - 6%) 0.509
PKLookup 105.50 (5.4%) 106.12
(4.4%) 0.6% ( -8% - 10%) 0.706
OrHighNotLow 199.60 (5.6%) 200.99
(4.6%) 0.7% ( -8% - 11%) 0.667
HighSpanNear 7.14 (2.6%) 7.20
(2.0%) 0.8% ( -3% - 5%) 0.259
HighTermTitleBDVSort 2.49 (3.5%) 2.52
(3.8%) 0.9% ( -6% - 8%) 0.445
OrHighHigh 21.16 (4.2%) 21.36
(3.5%) 1.0% ( -6% - 9%) 0.438
LowPhrase 8.67 (3.2%) 8.76
(3.0%) 1.0% ( -4% - 7%) 0.299
OrNotHighMed 130.15 (3.5%) 131.48
(3.6%) 1.0% ( -5% - 8%) 0.366
LowIntervalsOrdered 4.90 (3.3%) 4.96
(2.4%) 1.2% ( -4% - 7%) 0.182
OrNotHighHigh 156.60 (2.6%) 158.68
(3.5%) 1.3% ( -4% - 7%) 0.169
HighTerm 224.01 (3.0%) 227.90
(4.8%) 1.7% ( -5% - 9%) 0.168
AndHighLow 211.74 (6.6%) 215.73
(4.6%) 1.9% ( -8% - 14%) 0.297
OrNotHighLow 262.16 (3.7%) 267.97
(3.2%) 2.2% ( -4% - 9%) 0.042
LowTerm 228.72 (4.8%) 233.79
(7.1%) 2.2% ( -9% - 14%) 0.248
OrHighNotMed 176.76 (4.8%) 182.29
(4.8%) 3.1% ( -6% - 13%) 0.039
BrowseRandomLabelTaxoFacets 2.38 (27.8%) 2.58
(31.1%) 8.5% ( -39% - 93%) 0.360
BrowseDayOfYearTaxoFacets 3.69 (30.6%) 4.08
(43.1%) 10.6% ( -48% - 121%) 0.372
BrowseDateTaxoFacets 3.59 (30.9%) 3.97
(44.5%) 10.6% ( -49% - 124%) 0.381
BrowseMonthTaxoFacets 3.75 (33.1%) 4.20
(46.7%) 12.0% ( -50% - 137%) 0.348
```
Test4
```
TaskQPS baseline StdDevQPS
my_modified_version StdDev Pct diff p-value
BrowseDayOfYearSSDVFacets 3.28 (7.8%) 3.20
(8.3%) -2.2% ( -17% - 15%) 0.383
BrowseRandomLabelSSDVFacets 1.92 (9.2%) 1.88
(8.8%) -2.1% ( -18% - 17%) 0.458
AndHighLow 373.59 (3.2%) 365.70
(4.0%) -2.1% ( -9% - 5%) 0.065
LowTerm 238.90 (5.9%) 234.19
(3.8%) -2.0% ( -11% - 8%) 0.209
OrHighNotHigh 160.60 (4.2%) 158.04
(3.6%) -1.6% ( -9% - 6%) 0.197
HighTermMonthSort 924.41 (7.5%) 909.94
(4.5%) -1.6% ( -12% - 11%) 0.425
LowSpanNear 8.91 (1.8%) 8.80
(2.1%) -1.3% ( -5% - 2%) 0.043
HighTermDayOfYearSort 119.25 (3.6%) 117.88
(4.2%) -1.1% ( -8% - 6%) 0.356
MedIntervalsOrdered 3.82 (6.1%) 3.78
(6.9%) -1.0% ( -13% - 12%) 0.619
BrowseDateTaxoFacets 3.80 (30.3%) 3.76
(30.3%) -1.0% ( -47% - 85%) 0.917
HighTermTitleSort 49.61 (7.6%) 49.13
(7.4%) -1.0% ( -14% - 15%) 0.684
HighSpanNear 1.40 (1.5%) 1.39
(2.0%) -0.9% ( -4% - 2%) 0.125
MedTerm 312.23 (3.1%) 309.55
(4.5%) -0.9% ( -8% - 6%) 0.478
AndHighHigh 15.47 (2.0%) 15.36
(2.8%) -0.7% ( -5% - 4%) 0.365
HighSloppyPhrase 3.06 (2.6%) 3.04
(2.6%) -0.7% ( -5% - 4%) 0.432
HighIntervalsOrdered 0.60 (5.5%) 0.60
(6.1%) -0.6% ( -11% - 11%) 0.742
AndHighMed 45.12 (2.1%) 44.84
(2.5%) -0.6% ( -5% - 4%) 0.416
OrNotHighMed 112.15 (3.8%) 111.49
(4.1%) -0.6% ( -8% - 7%) 0.635
MedTermDayTaxoFacets 9.67 (2.9%) 9.62
(2.4%) -0.6% ( -5% - 4%) 0.491
Prefix3 100.76 (6.8%) 100.20
(4.7%) -0.6% ( -11% - 11%) 0.764
OrHighNotMed 178.70 (3.5%) 177.85
(2.8%) -0.5% ( -6% - 6%) 0.636
OrHighHigh 13.29 (3.4%) 13.23
(3.0%) -0.4% ( -6% - 6%) 0.667
TermDTSort 82.39 (4.2%) 82.11
(3.4%) -0.3% ( -7% - 7%) 0.779
MedSloppyPhrase 14.51 (2.3%) 14.47
(1.9%) -0.3% ( -4% - 4%) 0.630
AndHighHighDayTaxoFacets 2.77 (1.8%) 2.76
(1.9%) -0.3% ( -3% - 3%) 0.582
OrHighMedDayTaxoFacets 3.66 (3.2%) 3.64
(2.8%) -0.3% ( -6% - 5%) 0.753
LowIntervalsOrdered 25.74 (3.4%) 25.69
(3.1%) -0.2% ( -6% - 6%) 0.842
LowSloppyPhrase 4.02 (2.8%) 4.01
(2.6%) -0.2% ( -5% - 5%) 0.836
OrHighLow 229.87 (5.1%) 229.52
(5.5%) -0.2% ( -10% - 11%) 0.927
BrowseDayOfYearTaxoFacets 3.85 (28.1%) 3.85
(32.0%) -0.1% ( -46% - 83%) 0.989
MedPhrase 11.60 (3.6%) 11.59
(3.2%) -0.1% ( -6% - 6%) 0.935
OrNotHighHigh 181.54 (2.8%) 181.48
(4.8%) -0.0% ( -7% - 7%) 0.978
HighPhrase 34.64 (4.8%) 34.64
(3.2%) 0.0% ( -7% - 8%) 0.998
HighTermTitleBDVSort 2.47 (4.2%) 2.48
(4.1%) 0.1% ( -7% - 8%) 0.956
AndHighMedDayTaxoFacets 38.29 (2.6%) 38.32
(1.8%) 0.1% ( -4% - 4%) 0.908
LowPhrase 18.49 (4.3%) 18.50
(3.7%) 0.1% ( -7% - 8%) 0.944
OrHighMed 26.44 (2.8%) 26.48
(3.5%) 0.2% ( -5% - 6%) 0.876
Wildcard 20.32 (3.8%) 20.37
(3.3%) 0.2% ( -6% - 7%) 0.832
MedSpanNear 17.47 (2.0%) 17.53
(1.8%) 0.3% ( -3% - 4%) 0.554
Fuzzy2 35.81 (2.2%) 36.00
(2.7%) 0.5% ( -4% - 5%) 0.502
Respell 28.51 (2.9%) 28.68
(2.8%) 0.6% ( -4% - 6%) 0.497
BrowseRandomLabelTaxoFacets 2.45 (27.6%) 2.47
(26.1%) 0.7% ( -41% - 75%) 0.939
OrHighNotLow 166.55 (3.1%) 167.93
(3.9%) 0.8% ( -5% - 8%) 0.455
HighTerm 268.60 (4.4%) 271.37
(4.8%) 1.0% ( -7% - 10%) 0.478
OrNotHighLow 341.54 (4.4%) 345.89
(4.9%) 1.3% ( -7% - 11%) 0.384
IntNRQ 19.69 (2.9%) 19.95
(3.5%) 1.3% ( -4% - 7%) 0.199
Fuzzy1 39.15 (2.3%) 39.87
(2.0%) 1.8% ( -2% - 6%) 0.007
BrowseMonthSSDVFacets 3.62 (18.0%) 3.69
(20.5%) 2.2% ( -30% - 49%) 0.723
range 2175.09 (10.7%) 2229.64
(8.3%) 2.5% ( -14% - 24%) 0.409
PKLookup 104.87 (3.2%) 107.93
(3.3%) 2.9% ( -3% - 9%) 0.004
BrowseDateSSDVFacets 0.66 (14.7%) 0.68
(17.1%) 3.1% ( -25% - 40%) 0.543
BrowseMonthTaxoFacets 4.18 (42.3%) 5.26
(64.0%) 25.7% ( -56% - 228%) 0.134
```
Test5
```
TaskQPS baseline StdDevQPS
my_modified_version StdDev Pct diff p-value
BrowseDayOfYearTaxoFacets 4.73 (45.7%) 4.07
(35.1%) -14.0% ( -65% - 122%) 0.276
BrowseDateTaxoFacets 4.57 (43.9%) 3.95
(36.2%) -13.6% ( -65% - 118%) 0.286
BrowseRandomLabelTaxoFacets 2.83 (37.0%) 2.60
(29.0%) -8.1% ( -54% - 91%) 0.439
range 2378.00 (11.9%) 2260.56
(8.5%) -4.9% ( -22% - 17%) 0.131
BrowseDayOfYearSSDVFacets 3.40 (10.3%) 3.32
(12.9%) -2.2% ( -23% - 23%) 0.551
BrowseRandomLabelSSDVFacets 1.96 (10.5%) 1.93
(13.9%) -2.0% ( -23% - 25%) 0.605
MedTermDayTaxoFacets 8.98 (3.4%) 8.85
(3.0%) -1.4% ( -7% - 5%) 0.174
Fuzzy1 42.66 (3.8%) 42.12
(4.1%) -1.3% ( -8% - 6%) 0.311
HighTermTitleSort 56.06 (5.1%) 55.36
(2.4%) -1.3% ( -8% - 6%) 0.317
AndHighHigh 14.14 (2.2%) 13.96
(3.1%) -1.2% ( -6% - 4%) 0.154
HighTermMonthSort 986.79 (7.8%) 975.74
(7.6%) -1.1% ( -15% - 15%) 0.646
HighSpanNear 2.60 (2.0%) 2.57
(2.3%) -1.1% ( -5% - 3%) 0.100
LowSloppyPhrase 21.13 (2.0%) 20.96
(2.9%) -0.8% ( -5% - 4%) 0.301
HighSloppyPhrase 2.23 (2.9%) 2.21
(2.6%) -0.8% ( -6% - 4%) 0.354
LowPhrase 36.07 (2.8%) 35.81
(3.1%) -0.7% ( -6% - 5%) 0.444
OrNotHighLow 338.46 (3.2%) 336.26
(4.9%) -0.6% ( -8% - 7%) 0.617
AndHighMed 61.39 (3.0%) 61.01
(2.7%) -0.6% ( -6% - 5%) 0.497
HighTerm 205.75 (3.1%) 204.50
(5.0%) -0.6% ( -8% - 7%) 0.645
Wildcard 24.56 (4.0%) 24.42
(3.0%) -0.6% ( -7% - 6%) 0.601
AndHighHighDayTaxoFacets 2.54 (2.4%) 2.53
(2.7%) -0.6% ( -5% - 4%) 0.476
Respell 27.74 (2.5%) 27.58
(3.4%) -0.6% ( -6% - 5%) 0.559
HighTermDayOfYearSort 116.66 (3.6%) 116.21
(4.0%) -0.4% ( -7% - 7%) 0.748
Prefix3 401.91 (5.9%) 401.13
(8.4%) -0.2% ( -13% - 15%) 0.933
MedSloppyPhrase 37.73 (2.9%) 37.67
(2.9%) -0.2% ( -5% - 5%) 0.866
HighPhrase 11.75 (4.9%) 11.75
(6.7%) -0.0% ( -11% - 12%) 0.999
MedSpanNear 5.55 (1.7%) 5.56
(2.3%) 0.2% ( -3% - 4%) 0.806
LowSpanNear 3.72 (2.2%) 3.72
(1.9%) 0.2% ( -3% - 4%) 0.771
AndHighMedDayTaxoFacets 21.81 (2.5%) 21.87
(2.3%) 0.3% ( -4% - 5%) 0.711
Fuzzy2 36.16 (4.5%) 36.27
(2.4%) 0.3% ( -6% - 7%) 0.780
IntNRQ 19.82 (5.7%) 19.88
(5.0%) 0.3% ( -9% - 11%) 0.843
OrHighLow 216.46 (4.2%) 217.19
(3.8%) 0.3% ( -7% - 8%) 0.790
OrNotHighHigh 138.32 (5.1%) 138.98
(4.2%) 0.5% ( -8% - 10%) 0.745
OrHighHigh 13.04 (2.7%) 13.10
(3.0%) 0.5% ( -5% - 6%) 0.588
HighTermTitleBDVSort 3.85 (3.5%) 3.87
(4.2%) 0.6% ( -6% - 8%) 0.632
MedIntervalsOrdered 10.81 (4.7%) 10.87
(5.0%) 0.6% ( -8% - 10%) 0.702
HighIntervalsOrdered 1.37 (6.7%) 1.38
(7.5%) 0.7% ( -12% - 15%) 0.750
LowIntervalsOrdered 9.17 (2.5%) 9.24
(2.7%) 0.8% ( -4% - 6%) 0.339
OrNotHighMed 207.30 (3.3%) 208.97
(4.1%) 0.8% ( -6% - 8%) 0.492
MedPhrase 12.82 (4.3%) 12.93
(2.7%) 0.8% ( -5% - 8%) 0.464
BrowseMonthSSDVFacets 3.48 (6.5%) 3.52
(15.2%) 0.9% ( -19% - 24%) 0.804
AndHighLow 285.41 (5.9%) 288.14
(5.0%) 1.0% ( -9% - 12%) 0.579
PKLookup 107.13 (4.1%) 108.20
(3.1%) 1.0% ( -5% - 8%) 0.380
TermDTSort 89.93 (5.0%) 90.88
(3.9%) 1.1% ( -7% - 10%) 0.449
OrHighMedDayTaxoFacets 3.44 (3.1%) 3.48
(2.6%) 1.1% ( -4% - 7%) 0.222
OrHighMed 42.67 (2.7%) 43.23
(2.4%) 1.3% ( -3% - 6%) 0.102
OrHighNotHigh 206.53 (6.2%) 209.43
(5.3%) 1.4% ( -9% - 13%) 0.441
OrHighNotMed 175.18 (3.3%) 178.12
(5.3%) 1.7% ( -6% - 10%) 0.230
MedTerm 241.21 (4.3%) 245.54
(5.0%) 1.8% ( -7% - 11%) 0.222
OrHighNotLow 160.53 (4.3%) 163.59
(4.8%) 1.9% ( -6% - 11%) 0.187
LowTerm 211.35 (3.0%) 216.15
(4.5%) 2.3% ( -5% - 10%) 0.060
BrowseDateSSDVFacets 0.65 (13.1%) 0.68
(16.5%) 5.3% ( -21% - 40%) 0.265
BrowseMonthTaxoFacets 4.67 (44.8%) 5.27
(56.3%) 13.0% ( -60% - 206%) 0.419
```
Test6
```
TaskQPS baseline StdDevQPS
my_modified_version StdDev Pct diff p-value
OrHighNotMed 171.74 (3.8%) 165.28
(6.1%) -3.8% ( -13% - 6%) 0.019
LowTerm 262.47 (6.8%) 254.38
(4.9%) -3.1% ( -13% - 9%) 0.099
OrHighNotLow 185.00 (4.5%) 179.79
(3.6%) -2.8% ( -10% - 5%) 0.029
OrNotHighHigh 136.66 (5.2%) 133.82
(3.8%) -2.1% ( -10% - 7%) 0.149
HighPhrase 66.26 (6.4%) 65.02
(4.1%) -1.9% ( -11% - 9%) 0.274
OrNotHighMed 200.38 (5.3%) 197.56
(6.1%) -1.4% ( -12% - 10%) 0.433
LowPhrase 10.09 (3.2%) 9.96
(3.3%) -1.3% ( -7% - 5%) 0.214
AndHighMed 70.37 (3.5%) 69.63
(4.4%) -1.0% ( -8% - 7%) 0.410
HighSloppyPhrase 4.63 (3.3%) 4.59
(3.3%) -0.9% ( -7% - 5%) 0.395
TermDTSort 60.46 (7.3%) 60.03
(7.2%) -0.7% ( -14% - 14%) 0.753
OrNotHighLow 225.20 (3.2%) 223.60
(4.0%) -0.7% ( -7% - 6%) 0.532
MedPhrase 50.43 (4.9%) 50.12
(5.3%) -0.6% ( -10% - 10%) 0.697
HighTermDayOfYearSort 111.28 (5.3%) 110.68
(4.1%) -0.5% ( -9% - 9%) 0.719
AndHighMedDayTaxoFacets 34.61 (3.2%) 34.49
(2.3%) -0.3% ( -5% - 5%) 0.694
OrHighLow 222.40 (4.8%) 221.63
(3.7%) -0.3% ( -8% - 8%) 0.797
AndHighHigh 9.56 (2.4%) 9.52
(2.3%) -0.3% ( -4% - 4%) 0.641
Respell 22.70 (3.0%) 22.65
(3.5%) -0.2% ( -6% - 6%) 0.827
OrHighNotHigh 150.87 (4.4%) 150.54
(5.3%) -0.2% ( -9% - 9%) 0.887
Fuzzy1 21.63 (3.5%) 21.58
(3.5%) -0.2% ( -6% - 7%) 0.844
HighTermTitleSort 46.88 (3.6%) 46.79
(3.2%) -0.2% ( -6% - 6%) 0.855
IntNRQ 16.16 (6.5%) 16.16
(6.2%) -0.0% ( -11% - 13%) 0.999
OrHighMedDayTaxoFacets 2.01 (3.9%) 2.01
(4.5%) 0.1% ( -7% - 8%) 0.917
OrHighMed 43.53 (2.9%) 43.59
(3.1%) 0.2% ( -5% - 6%) 0.869
LowSpanNear 3.64 (2.1%) 3.65
(1.9%) 0.3% ( -3% - 4%) 0.642
HighTermTitleBDVSort 2.92 (3.0%) 2.93
(3.4%) 0.4% ( -5% - 6%) 0.723
MedSloppyPhrase 30.86 (4.1%) 30.99
(3.4%) 0.4% ( -6% - 8%) 0.725
LowIntervalsOrdered 9.38 (2.7%) 9.42
(3.2%) 0.4% ( -5% - 6%) 0.646
HighSpanNear 0.61 (4.4%) 0.61
(3.7%) 0.4% ( -7% - 8%) 0.728
MedIntervalsOrdered 2.84 (3.5%) 2.85
(3.4%) 0.5% ( -6% - 7%) 0.633
LowSloppyPhrase 4.07 (3.3%) 4.10
(3.3%) 0.7% ( -5% - 7%) 0.496
Wildcard 27.11 (4.3%) 27.31
(5.2%) 0.8% ( -8% - 10%) 0.615
HighTerm 347.91 (5.7%) 350.57
(4.2%) 0.8% ( -8% - 11%) 0.630
BrowseRandomLabelSSDVFacets 1.83 (12.5%) 1.85
(11.4%) 0.8% ( -20% - 28%) 0.833
OrHighHigh 16.78 (2.4%) 16.92
(3.9%) 0.9% ( -5% - 7%) 0.397
MedTermDayTaxoFacets 5.69 (3.0%) 5.74
(3.2%) 0.9% ( -5% - 7%) 0.364
Fuzzy2 31.38 (3.7%) 31.66
(3.6%) 0.9% ( -6% - 8%) 0.432
AndHighHighDayTaxoFacets 5.60 (2.9%) 5.65
(2.4%) 1.0% ( -4% - 6%) 0.251
BrowseDateSSDVFacets 0.64 (15.2%) 0.65
(15.4%) 1.0% ( -25% - 37%) 0.830
MedSpanNear 4.94 (2.6%) 5.00
(1.7%) 1.2% ( -3% - 5%) 0.088
Prefix3 78.50 (4.9%) 79.47
(6.3%) 1.2% ( -9% - 13%) 0.491
PKLookup 103.02 (5.0%) 104.30
(3.9%) 1.2% ( -7% - 10%) 0.381
AndHighLow 222.95 (4.3%) 226.54
(4.4%) 1.6% ( -6% - 10%) 0.244
MedTerm 350.42 (5.9%) 356.74
(4.3%) 1.8% ( -7% - 12%) 0.269
BrowseMonthSSDVFacets 3.34 (14.4%) 3.40
(13.1%) 1.8% ( -22% - 34%) 0.672
HighIntervalsOrdered 1.22 (4.7%) 1.24
(4.5%) 2.0% ( -6% - 11%) 0.171
range 1161.32 (10.4%) 1187.32
(9.3%) 2.2% ( -15% - 24%) 0.472
BrowseDayOfYearSSDVFacets 3.09 (7.8%) 3.16
(7.1%) 2.3% ( -11% - 18%) 0.336
HighTermMonthSort 951.11 (8.8%) 990.93
(6.9%) 4.2% ( -10% - 21%) 0.094
BrowseMonthTaxoFacets 3.92 (31.4%) 4.11
(45.0%) 4.9% ( -54% - 118%) 0.687
BrowseDayOfYearTaxoFacets 3.92 (29.6%) 4.15
(36.8%) 5.7% ( -46% - 102%) 0.587
BrowseDateTaxoFacets 3.76 (28.7%) 4.04
(37.4%) 7.4% ( -45% - 103%) 0.482
BrowseRandomLabelTaxoFacets 2.36 (23.3%) 2.57
(35.0%) 8.8% ( -40% - 87%) 0.348
```
<!--
If this is your first contribution to Lucene, please make sure you have
reviewed the contribution guide.
https://github.com/apache/lucene/blob/main/CONTRIBUTING.md
-->
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]