KalamDB Benchmark Report

Performance analysis — v0.4.2-rc1

http://127.0.0.1:8080 2026-04-11T16:19:37.175014+00:00 100 iters 5 warmup 10 concurrency

Total Benchmarks

Passed

Failed

298.54s

Total Duration

System Information

Hostname

MacBookPro.home

Machine Model

MacBookPro18,3

CPU Model

Apple M1 Pro

CPU Cores

10 logical / 10 physical

Total Memory

16.00 GiB

Available Memory

6.31 GiB

Used Memory

9.69 GiB

Memory Usage

60.6%

Darwin macOS 26.4

Kernel

25.4.0

Architecture

aarch64

Latency by Operation (µs)

Throughput (ops/sec)

Avg Latency by Category (µs)

Detailed Results

Status	Benchmark	Category	Description	Iters	Mean	P50	P95	P99	Min	Max	Ops/sec	Total	Verdict	vs Prev
PASS	create_table	DDL	CREATE TABLE with 3 columns	100	36.5ms	35.1ms	57.4ms	59.2ms	11.3ms	63.3ms	27	3.65s	🟢 Excellent	↑5% faster
PASS	drop_table	DDL	DROP TABLE on a previously created table	100	101.8ms	105.8ms	171.1ms	177.7ms	25.6ms	238.3ms	10	10.18s	🟡 Acceptable	↑6% faster
PASS	single_insert	Insert	INSERT a single row into a table	100	440µs	382µs	669µs	712µs	321µs	814µs	2.3K	44.0ms	🟢 Excellent	↑6% faster
PASS	bulk_insert	Insert	One INSERT statement with 50 rows (statement-scoped transaction)	100	1.4ms	1.4ms	1.6ms	1.7ms	1.2ms	2.0ms	725	137.8ms	🟢 Excellent	↓26% slower
PASS	transaction_multi_insert	Insert	Explicit BEGIN/COMMIT with 50 single-row INSERT statements	100	3.0ms	3.0ms	3.4ms	3.7ms	2.7ms	3.7ms	329	303.9ms	🟡 Acceptable	—
PASS	select_all	Select	SELECT * from a 200-row table	100	804µs	794µs	895µs	1.0ms	758µs	1.0ms	1.2K	80.4ms	🟢 Excellent	↑51% faster
PASS	select_by_filter	Select	SELECT with WHERE clause on a 200-row table	100	448µs	437µs	534µs	630µs	365µs	723µs	2.2K	44.8ms	🟢 Excellent	↑37% faster
PASS	select_count	Select	SELECT COUNT(*) on a 200-row table	100	632µs	614µs	753µs	831µs	586µs	898µs	1.6K	63.2ms	🟢 Excellent	↑51% faster
PASS	select_order_by_limit	Select	SELECT with ORDER BY + LIMIT 10 on a 200-row table	100	827µs	811µs	978µs	1.0ms	778µs	1.0ms	1.2K	82.7ms	🟢 Excellent	↑46% faster
PASS	single_update	Update	UPDATE a single row by filter condition	100	453µs	448µs	625µs	659µs	389µs	675µs	2.2K	45.3ms	🟢 Excellent	↑55% faster
PASS	single_delete	Delete	DELETE a single row by filter condition	100	396µs	398µs	498µs	581µs	330µs	658µs	2.5K	39.6ms	🟢 Excellent	↑53% faster
PASS	concurrent_insert	Concurrent	N concurrent INSERT operations in parallel (N = concurrency setting)	100	1.2ms	1.3ms	1.4ms	1.5ms	901µs	3.1ms	805	124.2ms	🟢 Excellent	↑43% faster
PASS	concurrent_select	Concurrent	N concurrent SELECT operations in parallel (N = concurrency setting)	100	1.3ms	1.3ms	1.5ms	1.6ms	955µs	1.7ms	773	129.4ms	🟢 Excellent	↑18% faster
PASS	point_lookup	Select	SELECT by primary key from a 10K-row table (single row lookup)	100	435µs	426µs	513µs	550µs	358µs	560µs	2.3K	43.5ms	🟢 Excellent	↑37% faster
PASS	aggregate_query	Select	GROUP BY + SUM/AVG/COUNT on a 10K-row table (analytical query performance)	100	28.6ms	28.2ms	33.7ms	36.1ms	26.7ms	36.9ms	35	2.86s	🔴 Slow	↑30% faster
PASS	multi_table_join	Select	SELECT with subquery across two tables (200 customers, 1000 orders)	100	2.9ms	2.9ms	3.2ms	3.3ms	2.7ms	3.5ms	345	289.8ms	🟢 Excellent	↑38% faster
PASS	large_payload_insert	Insert	INSERT rows with ~4KB TEXT payloads (serialization + storage throughput)	100	1.3ms	1.2ms	1.4ms	1.7ms	1.1ms	4.0ms	796	125.6ms	🟢 Excellent	↑43% faster
PASS	wide_column_insert	Insert	INSERT into a 20-column table (wide schema overhead)	100	1.8ms	1.7ms	2.1ms	4.3ms	1.6ms	7.2ms	544	183.8ms	🟢 Excellent	↓73% slower
PASS	bulk_delete	Delete	DELETE 100 rows at once with a range filter (bulk deletion)	100	139.4ms	136.5ms	163.9ms	207.9ms	115.4ms	281.5ms	7	13.94s	🟡 Acceptable	↑24% faster
PASS	sequential_crud	DML	INSERT → UPDATE → SELECT → DELETE full DML lifecycle per iteration	100	2.2ms	2.1ms	3.0ms	3.5ms	1.7ms	6.5ms	451	221.8ms	🟢 Excellent	↑26% faster
PASS	alter_table	DDL	ALTER TABLE ADD COLUMN + DROP COLUMN (schema evolution latency)	100	1.4ms	1.3ms	1.8ms	2.5ms	1.1ms	5.2ms	696	143.7ms	🟢 Excellent	↑55% faster
PASS	concurrent_update	Concurrent	N concurrent UPDATE operations on the same table (write contention test)	100	1.7ms	1.6ms	1.9ms	2.1ms	1.4ms	2.2ms	605	165.4ms	🟢 Excellent	↑60% faster
PASS	concurrent_mixed_dml	Concurrent	Concurrent INSERT + UPDATE + DELETE on the same table (multi-op contention)	100	3.0ms	3.0ms	3.7ms	4.2ms	2.2ms	5.9ms	330	303.0ms	🟢 Excellent	↑37% faster
PASS	namespace_isolation	Concurrent	Concurrent queries across 5 different namespaces (isolation test)	100	1.9ms	1.8ms	2.2ms	3.1ms	1.5ms	3.1ms	533	187.5ms	🟢 Excellent	↑57% faster
PASS	subscribe_initial_load	Subscribe	Subscribe to a 1000-row user table and receive the full initial data batch	100	7.4ms	7.2ms	8.7ms	13.8ms	6.7ms	14.8ms	135	742.9ms	🟢 Excellent	↑33% faster
PASS	subscribe_change_latency	Subscribe	Latency from INSERT to subscriber receiving the change notification	100	50.7ms	50.5ms	56.1ms	59.2ms	40.8ms	62.3ms	20	5.07s	🟡 Acceptable	↑99% faster
PASS	reconnect_subscribe	Subscribe	Disconnect and re-subscribe to a user table (reconnection overhead)	100	10.1ms	10.3ms	10.8ms	11.6ms	8.3ms	11.7ms	99	1.01s	🟢 Excellent	↑16% faster
PASS	flushed_parquet_query	Storage	SELECT from a shared table with 20 flushed Parquet files (200K rows)	100	150.4ms	173.6ms	246.1ms	261.7ms	37.7ms	282.7ms	7	15.04s	🟡 Acceptable	↑42% faster
PASS	concurrent_subscribers	Load	N WebSocket live-query subscribers receiving changes from concurrent writes	100	1.56s	1.57s	1.59s	1.60s	1.53s	1.60s	1	156.39s	🟢 Excellent	~ 1577599µs prior
PASS	concurrent_publishers	Load	N concurrent INSERTs into a topic-sourced table (measures publish overhead)	100	1.3ms	1.3ms	1.7ms	2.2ms	956µs	2.8ms	744	134.4ms	🟢 Excellent	↑38% faster
PASS	concurrent_consumers	Load	N concurrent topic CONSUME calls pulling messages in parallel	100	644µs	643µs	696µs	730µs	563µs	814µs	1.6K	64.4ms	🟢 Excellent	↓4% slower
PASS	sql_1k_concurrent	Load	1000 concurrent SQL SELECT queries at once (RPS degradation test)	100	164.7ms	162.2ms	180.2ms	194.1ms	153.5ms	214.9ms	6	16.47s	🟢 Excellent	↑62% faster
PASS	create_user	Load	CREATE USER (auth subsystem stress test)	100	1.2ms	1.2ms	1.3ms	1.4ms	1.2ms	1.5ms	822	121.6ms	🟢 Excellent	↑35% faster
PASS	drop_user	Load	DROP USER (auth subsystem teardown stress test)	100	296µs	283µs	395µs	493µs	239µs	548µs	3.4K	29.6ms	🟢 Excellent	↑46% faster
PASS	connection_storm	Load	N simultaneous login + SQL + cycles (connection setup overhead)	100	303.5ms	291.6ms	345.6ms	506.8ms	270.1ms	631.0ms	3	30.35s	🟢 Excellent	↓4811% slower
PASS	mixed_read_write	Load	50/50 concurrent reads + writes on same table (contention test)	100	2.7ms	2.6ms	3.7ms	6.0ms	1.9ms	6.1ms	366	273.3ms	🟢 Excellent	↑29% faster
PASS	wide_fanout_query	Load	N concurrent large-result-set SELECTs (serialization + memory pressure)	100	6.3ms	6.1ms	7.5ms	9.0ms	5.6ms	12.1ms	158	632.8ms	🟢 Excellent	↑67% faster
PASS	subscriber_scale	Scale	Progressive subscriber scale test (up to --max-subscribers, default 100K)	1	38.82s	38.82s	38.82s	38.82s	38.82s	38.82s	0	38.82s	🟢 Excellent	↓275% slower