Add HTTP remote force push tests

[timsehn]

Summary

• add HTTP remotesrv coverage for stale push rejection followed by --force rewrite on main
• verify fresh clones and fetched tracking refs see the forced tip, not the displaced branch head
• add non-main branch force-push coverage over HTTP
• run the new script in non-Windows CI with the other HTTP remote tests

Test

• bash test/remotesrv_http_force_push_test.sh build/doltlite build/doltlite-remotesrv
• bash test/remotesrv_http_test.sh build/doltlite build/doltlite-remotesrv
• bash test/remotesrv_http_concurrency_test.sh build/doltlite build/doltlite-remotesrv
• bash test/remotesrv_http_partial_failure_test.sh build/doltlite build/doltlite-remotesrv

[github-actions]

Version-Control Performance Ceilings

Runs: median of 3 executions per benchmark, excluding fixture setup. The
working-set fixtures use 800 committed tables with 125 rows each
(100000 rows total); data-dirty cases update 1 row in 40
tables, and schema-dirty cases update 1 row plus add a column in 20 tables.
Branch tests use 300 branches, checkout uses a clean 400-table
branch switch over 100000 rows, and merge
tests use 100000-row tables with 2000 changed or conflicting
rows per side.

Benchmark	Median ms	Ceiling ms	Result
status_clean_many_tables	90	200	PASS
status_dirty_many_tables	91	200	PASS
diff_regular_working_one_table	85	150	PASS
diff_regular_working_many_tables	91	200	PASS
diff_stat_working_many_tables	91	200	PASS
diff_schema_working_many_tables	92	200	PASS
branch_list_many_branches	23	100	PASS
branch_create_delete	26	100	PASS
checkout_branch_clean	58	200	PASS
merge_data_no_conflicts	29	150	PASS
merge_schema_no_conflicts	22	100	PASS
merge_data_conflicts	128	250	PASS
merge_data_conflicts_with_resolve	127	250	PASS

[github-actions]

Sysbench-Style Benchmark (TEXT PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 32-char hex TEXT PRIMARY KEY (UUID-shaped).
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	25,696	31,513	1.23
oltp_range_select	11,621	12,327	1.06
oltp_sum_range	11,298	12,260	1.09
oltp_order_range	2,634	2,831	1.07
oltp_distinct_range	3,368	3,665	1.09
oltp_index_scan	3,888	5,510	1.42
select_random_points	16,323	19,784	1.21
select_random_ranges	3,412	4,548	1.33
covering_index_scan	4,147	4,104	0.99
groupby_scan	28,868	30,310	1.05
index_join	6,099	8,738	1.43
index_join_scan	3,814	5,009	1.31
types_table_scan	1,050,740	1,166,619	1.11
table_scan	1,242,306	1,255,248	1.01
oltp_read_only	104,681	119,713	1.14
Average			1.17

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	198,840	292,844	1.47
oltp_insert	18,794	33,257	1.77
oltp_update_index	62,301	119,254	1.91
oltp_update_non_index	43,067	72,720	1.69
oltp_delete_insert	44,177	88,554	2.00
oltp_write_only	24,613	51,372	2.09
types_delete_insert	28,104	46,548	1.66
oltp_read_write	73,610	118,540	1.61
Average			1.78

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	55,519	43,373	0.78
oltp_range_select	14,505	13,600	0.94
oltp_sum_range	14,712	13,666	0.93
oltp_order_range	3,053	3,017	0.99
oltp_distinct_range	3,776	3,822	1.01
oltp_index_scan	7,071	6,678	0.94
select_random_points	19,179	21,329	1.11
select_random_ranges	6,457	5,672	0.88
covering_index_scan	7,625	5,298	0.69
groupby_scan	29,174	30,480	1.04
index_join	7,965	9,595	1.20
index_join_scan	4,288	5,265	1.23
types_table_scan	1,017,923	1,164,284	1.14
table_scan	1,341,292	1,278,237	0.95
oltp_read_only	156,673	138,629	0.88
Average			0.98

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	211,833	310,858	1.47
oltp_insert	55,075	43,086	0.78
oltp_update_index	85,295	141,451	1.66
oltp_update_non_index	96,954	89,041	0.92
oltp_delete_insert	65,559	107,221	1.64
oltp_write_only	52,365	67,426	1.29
types_delete_insert	40,480	59,783	1.48
oltp_read_write	102,561	137,333	1.34
Average			1.32

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	35,276	43,620	1.24
oltp_range_select	12,651	13,806	1.09
oltp_sum_range	12,433	13,732	1.10
oltp_order_range	2,799	2,984	1.07
oltp_distinct_range	3,522	3,831	1.09
oltp_index_scan	5,251	6,794	1.29
select_random_points	17,780	21,203	1.19
select_random_ranges	4,433	5,715	1.29
covering_index_scan	5,631	5,359	0.95
groupby_scan	29,176	30,566	1.05
index_join	7,502	9,515	1.27
index_join_scan	4,162	5,249	1.26
types_table_scan	1,026,242	1,153,292	1.12
table_scan	1,267,304	1,259,298	0.99
oltp_read_only	120,019	137,708	1.15
Average			1.14

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	15,145	55,433	3.66
oltp_insert_ac	17,642	68,899	3.91
oltp_update_index_ac	19,226	80,680	4.20
oltp_update_non_index_ac	15,260	66,195	4.34
oltp_delete_insert_ac	18,360	72,744	3.96
oltp_write_only_ac	17,033	74,135	4.35
types_delete_insert_ac	14,694	63,051	4.29
oltp_read_write_ac	21,934	78,799	3.59
Average			4.04

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark (composite PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 2-column INTEGER PRIMARY KEY(a, b) WITHOUT ROWID.
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	34,235	38,208	1.12
oltp_range_select	19,929	21,303	1.07
oltp_sum_range	17,701	20,433	1.15
oltp_order_range	3,673	3,929	1.07
oltp_distinct_range	4,676	4,990	1.07
oltp_index_scan	4,641	5,957	1.28
select_random_points	28,165	31,097	1.10
select_random_ranges	7,546	8,350	1.11
covering_index_scan	4,302	4,383	1.02
groupby_scan	38,205	41,052	1.07
index_join	8,013	10,499	1.31
index_join_scan	4,169	5,736	1.38
types_table_scan	1,181,827	1,273,090	1.08
table_scan	1,456,697	1,423,114	0.98
oltp_read_only	155,503	166,062	1.07
Average			1.12

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	245,293	340,018	1.39
oltp_insert	19,351	36,389	1.88
oltp_update_index	71,611	121,108	1.69
oltp_update_non_index	53,197	84,343	1.59
oltp_delete_insert	51,355	97,905	1.91
oltp_write_only	27,948	60,597	2.17
types_delete_insert	33,007	54,334	1.65
oltp_read_write	102,478	151,298	1.48
Average			1.72

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	111,202	67,551	0.61
oltp_range_select	28,480	24,185	0.85
oltp_sum_range	26,255	23,307	0.89
oltp_order_range	4,467	4,276	0.96
oltp_distinct_range	5,563	5,376	0.97
oltp_index_scan	13,045	9,102	0.70
select_random_points	36,847	33,777	0.92
select_random_ranges	15,904	11,442	0.72
covering_index_scan	12,742	7,482	0.59
groupby_scan	39,091	41,731	1.07
index_join	12,269	12,508	1.02
index_join_scan	5,126	6,104	1.19
types_table_scan	1,165,380	1,271,316	1.09
table_scan	1,414,736	1,387,311	0.98
oltp_read_only	271,810	208,655	0.77
Average			0.89

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	257,171	362,051	1.41
oltp_insert	26,027	46,190	1.77
oltp_update_index	98,827	148,166	1.50
oltp_update_non_index	76,325	106,015	1.39
oltp_delete_insert	76,942	120,363	1.56
oltp_write_only	49,523	80,601	1.63
types_delete_insert	48,998	68,104	1.39
oltp_read_write	121,514	171,926	1.41
Average			1.51

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	60,448	67,768	1.12
oltp_range_select	22,873	24,482	1.07
oltp_sum_range	20,494	23,243	1.13
oltp_order_range	4,130	4,273	1.03
oltp_distinct_range	5,140	5,325	1.04
oltp_index_scan	7,742	9,118	1.18
select_random_points	31,186	34,398	1.10
select_random_ranges	10,377	11,434	1.10
covering_index_scan	7,602	7,535	0.99
groupby_scan	38,509	41,842	1.09
index_join	10,054	12,623	1.26
index_join_scan	4,720	6,115	1.30
types_table_scan	1,154,577	1,261,243	1.09
table_scan	1,428,695	1,397,080	0.98
oltp_read_only	191,834	209,072	1.09
Average			1.10

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	16,609	61,462	3.70
oltp_insert_ac	18,785	89,877	4.78
oltp_update_index_ac	20,697	97,972	4.73
oltp_update_non_index_ac	17,394	77,576	4.46
oltp_delete_insert_ac	18,794	85,177	4.53
oltp_write_only_ac	18,180	83,278	4.58
types_delete_insert_ac	16,124	73,609	4.57
oltp_read_write_ac	24,663	92,696	3.76
Average			4.39

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark: Doltlite vs SQLite

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	20,501	23,041	1.12
oltp_range_select	8,917	10,862	1.22
oltp_sum_range	8,266	10,127	1.23
oltp_order_range	2,383	2,715	1.14
oltp_distinct_range	3,272	3,625	1.11
oltp_index_scan	3,476	4,198	1.21
select_random_points	9,645	10,208	1.06
select_random_ranges	2,586	3,338	1.29
covering_index_scan	3,404	3,374	0.99
groupby_scan	28,158	30,455	1.08
index_join	5,149	7,050	1.37
index_join_scan	2,792	4,102	1.47
types_table_scan	1,016,159	1,165,298	1.15
table_scan	1,150,881	1,282,484	1.11
oltp_read_only	89,198	104,342	1.17
Average			1.18

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	140,517	189,826	1.35
oltp_insert	12,477	22,364	1.79
oltp_update_index	42,705	75,415	1.77
oltp_update_non_index	28,055	48,133	1.72
oltp_delete_insert	39,807	57,986	1.46
oltp_write_only	18,459	36,156	1.96
types_delete_insert	20,419	32,567	1.59
oltp_read_write	55,054	87,758	1.59
Average			1.65

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	50,010	33,557	0.67
oltp_range_select	12,388	12,364	1.00
oltp_sum_range	11,935	11,653	0.98
oltp_order_range	2,769	2,896	1.05
oltp_distinct_range	3,651	3,807	1.04
oltp_index_scan	6,747	5,712	0.85
select_random_points	13,390	11,956	0.89
select_random_ranges	5,791	4,479	0.77
covering_index_scan	6,633	4,834	0.73
groupby_scan	28,449	30,573	1.07
index_join	6,953	8,288	1.19
index_join_scan	3,247	4,363	1.34
types_table_scan	1,001,122	1,165,243	1.16
table_scan	1,157,998	1,285,077	1.11
oltp_read_only	133,153	120,044	0.90
Average			0.98

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	191,142	252,457	1.32
oltp_insert	26,760	40,071	1.50
oltp_update_index	147,514	152,126	1.03
oltp_update_non_index	119,286	105,603	0.89
oltp_delete_insert	128,809	119,445	0.93
oltp_write_only	94,415	87,337	0.93
types_delete_insert	80,803	81,024	1.00
oltp_read_write	131,311	138,865	1.06
Average			1.08

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	29,450	33,654	1.14
oltp_range_select	10,507	12,340	1.17
oltp_sum_range	9,946	11,857	1.19
oltp_order_range	2,556	2,912	1.14
oltp_distinct_range	3,541	3,820	1.08
oltp_index_scan	4,615	5,707	1.24
select_random_points	11,406	12,034	1.06
select_random_ranges	3,678	4,552	1.24
covering_index_scan	4,492	4,885	1.09
groupby_scan	28,516	30,666	1.08
index_join	6,007	8,244	1.37
index_join_scan	3,051	4,348	1.43
types_table_scan	1,000,723	1,167,215	1.17
table_scan	1,150,568	1,297,311	1.13
oltp_read_only	103,945	120,435	1.16
Average			1.18

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	23,577	86,663	3.68
oltp_insert_ac	26,864	99,914	3.72
oltp_update_index_ac	26,685	115,192	4.32
oltp_update_non_index_ac	22,835	89,795	3.93
oltp_delete_insert_ac	24,176	101,474	4.20
oltp_write_only_ac	24,521	110,339	4.50
types_delete_insert_ac	23,044	94,286	4.09
oltp_read_write_ac	27,705	112,335	4.05
Average			4.06

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark (BLOB PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 16-byte big-endian BLOB PRIMARY KEY.
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	30,491	36,578	1.20
oltp_range_select	12,758	14,836	1.16
oltp_sum_range	12,052	13,934	1.16
oltp_order_range	2,956	3,157	1.07
oltp_distinct_range	3,968	4,271	1.08
oltp_index_scan	4,572	6,282	1.37
select_random_points	18,565	20,183	1.09
select_random_ranges	4,152	5,292	1.27
covering_index_scan	4,322	4,649	1.08
groupby_scan	31,359	33,439	1.07
index_join	6,872	9,478	1.38
index_join_scan	4,330	5,530	1.28
types_table_scan	1,103,627	1,323,378	1.20
table_scan	1,340,310	1,455,300	1.09
oltp_read_only	123,531	136,361	1.10
Average			1.17

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	237,358	351,431	1.48
oltp_insert	19,728	38,957	1.97
oltp_update_index	70,885	130,428	1.84
oltp_update_non_index	49,339	84,671	1.72
oltp_delete_insert	50,112	102,851	2.05
oltp_write_only	27,737	61,898	2.23
types_delete_insert	32,519	53,932	1.66
oltp_read_write	86,089	139,469	1.62
Average			1.82

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	100,128	60,461	0.60
oltp_range_select	20,052	17,349	0.87
oltp_sum_range	19,496	16,648	0.85
oltp_order_range	3,804	3,615	0.95
oltp_distinct_range	4,851	4,640	0.96
oltp_index_scan	11,575	8,770	0.76
select_random_points	26,189	23,506	0.90
select_random_ranges	10,805	7,477	0.69
covering_index_scan	11,741	7,138	0.61
groupby_scan	32,770	33,865	1.03
index_join	10,812	11,154	1.03
index_join_scan	5,310	5,799	1.09
types_table_scan	1,075,791	1,317,763	1.22
table_scan	1,269,922	1,442,491	1.14
oltp_read_only	224,293	172,778	0.77
Average			0.90

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	250,921	369,943	1.47
oltp_insert	36,296	50,748	1.40
oltp_update_index	104,240	161,028	1.54
oltp_update_non_index	82,768	106,760	1.29
oltp_delete_insert	85,163	127,961	1.50
oltp_write_only	61,969	82,751	1.34
types_delete_insert	54,467	71,605	1.31
oltp_read_write	122,046	161,576	1.32
Average			1.40

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	53,271	61,373	1.15
oltp_range_select	15,603	17,359	1.11
oltp_sum_range	15,323	16,572	1.08
oltp_order_range	3,504	3,634	1.04
oltp_distinct_range	4,446	4,683	1.05
oltp_index_scan	7,214	8,854	1.23
select_random_points	21,424	24,211	1.13
select_random_ranges	6,757	7,705	1.14
covering_index_scan	7,875	7,293	0.93
groupby_scan	32,818	34,100	1.04
index_join	9,234	11,322	1.23
index_join_scan	4,976	5,898	1.19
types_table_scan	1,075,330	1,335,806	1.24
table_scan	1,387,797	1,454,125	1.05
oltp_read_only	156,769	173,060	1.10
Average			1.11

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	27,825	88,779	3.19
oltp_insert_ac	30,082	108,099	3.59
oltp_update_index_ac	31,542	126,260	4.00
oltp_update_non_index_ac	27,024	98,478	3.64
oltp_delete_insert_ac	28,723	108,818	3.79
oltp_write_only_ac	30,823	117,116	3.80
types_delete_insert_ac	26,405	103,941	3.94
oltp_read_write_ac	34,054	122,818	3.61
Average			3.70

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.