GODRIVER-3026: Optimize Collection.insert()

[zhouselena]

GODRIVER-3026

Summary

Use switch instead of reflect to increase efficiency for InsertMany.

Background & Motivation

[mongodb-drivers-pr-bot]

API Change Report

No changes found!

[mongodb-drivers-pr-bot]

🧪 Performance Results

Commit SHA: f44f99f

The following benchmark tests for version 68a4be6570184300074e4d3a had statistically significant changes (i.e., |z-score| > 1.96):

Benchmark	Measurement	% Change	Patch Value	Stable Region	H-Score	Z-Score
BenchmarkMultiInsertLargeDocument	allocs_per_op	190216.5915	94413.0000	Avg: 49.6084 Med: 50.0000 Stdev: 0.5178	1.0000	182245.4417
BenchmarkMultiInsertLargeDocument	total_mem_allocs	119331.5475	2077090.0000	Avg: 1739.1469 Med: 1713.0000 Stdev: 95.4929	1.0000	21733.0301
BenchmarkMultiInsertSmallDocument	allocated_bytes_per_op	95.4616	5420.0000	Avg: 2772.9231 Med: 2792.0000 Stdev: 62.8574	0.9876	42.1124
BenchmarkMultiInsertLargeDocument	allocated_bytes_per_op	90.2425	53228285.0000	Avg: 27979174.5594 Med: 27920433.0000 Stdev: 656919.1737	0.9858	38.4356
BenchmarkMultiInsertLargeDocument	ns_per_op	63.1493	52706198.0000	Avg: 32305494.0000 Med: 32069582.0000 Stdev: 1943833.2923	0.9525	10.4951
BenchmarkMultiInsertSmallDocument	ns_per_op	45.8606	9613.0000	Avg: 6590.5405 Med: 6581.5000 Stdev: 153.3645	0.9721	19.7077
BenchmarkMultiInsertSmallDocument	total_mem_allocs	39.6344	3164893.0000	Avg: 2266556.5397 Med: 2237722.0000 Stdev: 113017.3801	0.9374	7.9487
BenchmarkMultiInsertSmallDocument	total_bytes_allocated	37.3741	612671776.0000	Avg: 445987875.1329 Med: 445380176.0000 Stdev: 13676059.4929	0.9570	12.1880
BenchmarkMultiInsertLargeDocument	total_bytes_allocated	24.6786	1171022280.0000	Avg: 939232665.0000 Med: 952990156.0000 Stdev: 48082774.7367	0.9001	4.8206
BenchmarkMultiFindMany	total_time_seconds	24.3939	1.3224	Avg: 1.0631 Med: 1.0301 Stdev: 0.0977	0.8407	2.6547
BenchmarkBSONDeepDocumentDecoding	ops_per_second_min	-18.6908	1607.3708	Avg: 1976.8619 Med: 1985.1235 Stdev: 120.7662	0.8188	-3.0596
BenchmarkSingleRunCommand	total_bytes_allocated	-12.4565	86532288.0000	Avg: 98844890.4270 Med: 99683088.0000 Stdev: 5079422.7369	0.7969	-2.4240
BenchmarkSingleFindOneByID	total_mem_allocs	-12.4149	1456249.0000	Avg: 1662668.3862 Med: 1671894.0000 Stdev: 67693.5237	0.8146	-3.0493
BenchmarkSingleRunCommand	total_mem_allocs	-12.0313	938500.0000	Avg: 1066857.1798 Med: 1077155.0000 Stdev: 54320.8845	0.7913	-2.3629
BenchmarkSingleRunCommand	ops_per_second_med	-11.7857	6726.8948	Avg: 7625.6258 Med: 7663.1289 Stdev: 297.7930	0.8292	-3.0180
BenchmarkSingleFindOneByID	total_bytes_allocated	-11.6975	90610696.0000	Avg: 102613970.0741 Med: 103118528.0000 Stdev: 4174203.9270	0.8034	-2.8756
BenchmarkBSONFullDocumentEncoding	total_bytes_allocated	-11.2986	215087968.0000	Avg: 242485478.8315 Med: 242696304.0000 Stdev: 3700011.4168	0.9241	-7.4047
BenchmarkBSONFullDocumentEncoding	total_mem_allocs	-11.2119	1243435.0000	Avg: 1400451.6854 Med: 1401964.0000 Stdev: 21551.3194	0.9229	-7.2857
BenchmarkBSONFullDocumentEncoding	total_time_seconds	-10.9709	1.0633	Avg: 1.1943 Med: 1.1938 Stdev: 0.0082	0.9654	-15.9618
BenchmarkLargeDocInsertOne	ops_per_second_med	-9.6938	5155.8883	Avg: 5709.3435 Med: 5716.8665 Stdev: 153.0671	0.8536	-3.6158
BenchmarkLargeDocInsertOne	total_bytes_allocated	-9.2551	32812136.0000	Avg: 36158654.0529 Med: 36253384.0000 Stdev: 994613.0921	0.8417	-3.3646
BenchmarkSingleFindOneByID	ops_per_second_med	-6.7546	3838.5800	Avg: 4116.6412 Med: 4138.0623 Stdev: 116.0625	0.7851	-2.3958
BenchmarkBSONFullDocumentEncoding	allocated_bytes_per_op	-0.1045	5197.0000	Avg: 5202.4382 Med: 5202.0000 Stdev: 2.0447	0.7948	-2.6597

For a comprehensive view of all microbenchmark results for this PR's commit, please check out the Evergreen perf task for this patch.