Do not perform another pass of Query Automaton Minimization

[atris]

Native text engine minimises the query automaton post construction using Hopcroft's algorithm. This can get expensive for large query automatons, and does not yield much improvement anyways since the query automaton is build once and use once.

Post this change, performance numbers using BenchmarkNativeAndLuceneBasedLike:

Benchmark (_fstType) (_intBaseValue) (_numBlocks) (_numRows) (_query) Mode Cnt Score Error Units
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 0 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 40.436 ± 8.662 us/op
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 0 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 50.320 ± 4.254 us/op
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 1 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 42.378 ± 2.669 us/op
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 1 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 53.890 ± 2.951 us/op
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 10 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 47.751 ± 1.149 us/op
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 10 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 60.890 ± 1.949 us/op
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 100 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 93.937 ± 8.493 us/op
BenchmarkNativeAndLuceneBasedLike.query LUCENE 1000 100 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 129.687 ± 16.903 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 0 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 55.362 ± 10.320 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 0 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 16.610 ± 1.297 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 1 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 54.800 ± 1.501 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 1 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 18.417 ± 0.696 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 10 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 60.187 ± 3.858 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 10 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 25.549 ± 1.694 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 100 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%' avgt 5 106.765 ± 13.996 us/op
BenchmarkNativeAndLuceneBasedLike.query NATIVE 1000 100 2500000 SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%' avgt 5 99.888 ± 1.029 us/op

Note that for generic match queries '%domain%', Lucene and Native FST are at parity from 0 blocks to 100 blocks. For prefix queries, Native FST is 4x faster on 0 and 10 blocks, and 33% faster on 100 blocks.

BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000             0     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   50.320 ±  4.254  us/op

BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000             0     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   16.610 ±  1.297  us/op

BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000           100     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5  129.687 ± 16.903  us/op

BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000           100     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   99.888 ±  1.029  us/op

This behaviour was observed over multiple runs of the benchmark. Detailed results at:

https://docs.google.com/document/d/1Jd-Oe0F9gx9WAB1sa5YdW7KZ_EsPOJdcHK1bsON9JHM/edit?usp=sharing

[codecov-commenter]

Codecov Report

Merging #8237 (d7cc6d7) into master (4f17ede) will decrease coverage by 6.95%.
The diff coverage is 78.43%.

❗ Current head d7cc6d7 differs from pull request most recent head 95554f2. Consider uploading reports for the commit 95554f2 to get more accurate results

@@             Coverage Diff              @@
##             master    #8237      +/-   ##
============================================
- Coverage     70.96%   64.00%   -6.96%     
+ Complexity     4320     4239      -81     
============================================
  Files          1626     1584      -42     
  Lines         85081    83250    -1831     
  Branches      12803    12608     -195     
============================================
- Hits          60377    53288    -7089     
- Misses        20545    26127    +5582     
+ Partials       4159     3835     -324     

Flag	Coverage Δ
integration1	?
integration2	?
unittests1	66.95% <78.43%> (-0.40%)	⬇️
unittests2	14.10% <0.00%> (-0.02%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
...ery/optimizer/filter/NumericalFilterOptimizer.java	83.22% <ø> (-2.49%)	⬇️
...inot/core/operator/filter/FilterOperatorUtils.java	80.00% <42.85%> (-7.68%)	⬇️
...ot/common/request/context/RequestContextUtils.java	72.77% <50.00%> (-1.60%)	⬇️
.../pinot/core/operator/filter/NotFilterOperator.java	55.55% <55.55%> (ø)
...ava/org/apache/pinot/core/plan/FilterPlanNode.java	87.85% <75.00%> (-1.37%)	⬇️
...core/operator/dociditerators/NotDocIdIterator.java	95.00% <95.00%> (ø)
...he/pinot/common/request/context/FilterContext.java	78.37% <100.00%> (ø)
.../apache/pinot/pql/parsers/pql2/ast/FilterKind.java	100.00% <100.00%> (ø)
...che/pinot/core/operator/docidsets/NotDocIdSet.java	100.00% <100.00%> (ø)
...egment/local/utils/nativefst/automaton/RegExp.java	42.96% <100.00%> (-1.49%)	⬇️
... and 389 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 4f17ede...95554f2. Read the comment docs.

[richardstartin]

I observe a good speed up with this change.

[richardstartin]

I got these numbers with JDK11 (coretto) on my MacBook Pro with the CLI args below:

java -jar pinot-perf/target/benchmarks.jar -wi 5 -i 5 -r 1 -w 2 -f 1 -bm avgt -jvmArgsPrepend "-ms4G -mx4G -XX:+AlwaysPreTouch -XX:+UseParallelGC" BenchmarkNativeAndLuceneBasedLike

Benchmark                                (_fstType)  (_intBaseValue)  (_numBlocks)  (_numRows)                                                                    (_query)  Mode  Cnt    Score   Error  Units
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000             0     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5   47.434 ± 0.878  us/op
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000             0     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   49.754 ± 1.126  us/op
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000             1     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5   51.999 ± 0.557  us/op
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000             1     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   53.120 ± 0.825  us/op
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000            10     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5   59.283 ± 1.135  us/op
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000            10     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   62.024 ± 1.437  us/op
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000           100     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5  117.280 ± 0.580  us/op
BenchmarkNativeAndLuceneBasedLike.query      LUCENE             1000           100     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5  164.621 ± 8.522  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000             0     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5   44.811 ± 1.573  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000             0     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   14.112 ± 0.155  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000             1     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5   49.487 ± 0.414  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000             1     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   19.014 ± 0.583  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000            10     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5   54.869 ± 0.861  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000            10     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5   29.343 ± 0.206  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000           100     2500000     SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE '%domain%'  avgt    5  115.629 ± 5.838  us/op
BenchmarkNativeAndLuceneBasedLike.query      NATIVE             1000           100     2500000  SELECT INT_COL, URL_COL FROM MyTable WHERE DOMAIN_NAMES LIKE 'www.domain%'  avgt    5  128.661 ± 3.260  us/op

So no integer multiple differences for unanchored prefixes in this run, anchored prefixes are much faster than lucene, but the native implementation appears to warm up faster. I can run this on some more stable machines, but we wouldn't see this kind of improvement by accident.