Remote round-trip oracle coverage for schema objects

[timsehn]

Continues the schema-objects × VC oracle campaign (#1657 add, #1663 merge, #1665 checkout, #1666 revert/cherry-pick). Fills the remotes gap.

Coverage added (9 scenarios, live Dolt over file remotes)

• Clone: a repo with a view + index arrives with the view queryable and the index present; a cloned trigger fires on the consumer's first insert.
• Fast-forward pull: later commits adding a view, an index (usable over pulled data), or a trigger (fires post-pull) all land in the consumer.
• Pull merge (non-ff): remote view add over a local data commit merges cleanly with both preserved; remote view drop propagates; both sides dropped stays dropped through the pull-merge path (the Merge oracle coverage for schema objects; fix both-drop resurrection #1663 resurrection chokepoint exercised via dolt_pull).
• Dirty working set: both systems refuse the pull ("cannot pull/merge with uncommitted changes") and deliver nothing.
• Trigger scenarios run per-dialect scripts via a remote_flow_dual variant of the existing seed→clone→advance→consume harness.

No engine changes

All scenarios pass on master — probing found full parity across clone/push/pull with schema objects (one probe initially looked like a pull bug but was my own dirty-working-set setup error; the refusal itself matches Dolt and is now pinned as a scenario).

Suite total: 29/29. Siblings green: remotes oracle 34/34, push-default-branch 19/19, remotesrv HTTP 28/28.

🤖 Generated with Claude Code

[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	81	200	PASS
status_dirty_many_tables	84	200	PASS
diff_regular_working_one_table	77	150	PASS
diff_regular_working_many_tables	90	200	PASS
diff_stat_working_many_tables	89	200	PASS
diff_schema_working_many_tables	89	200	PASS
branch_list_many_branches	21	100	PASS
branch_create_delete	24	100	PASS
checkout_branch_clean	54	200	PASS
merge_data_no_conflicts	28	150	PASS
merge_schema_no_conflicts	21	100	PASS
merge_data_conflicts	125	250	PASS
merge_data_conflicts_with_resolve	125	250	PASS

[github-actions]

Sysbench-Style Benchmark: Doltlite vs SQLite

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	22,412	28,581	1.28
oltp_range_select	8,952	12,553	1.40
oltp_sum_range	8,564	12,009	1.40
oltp_order_range	2,363	2,865	1.21
oltp_distinct_range	3,369	3,992	1.18
oltp_index_scan	3,637	5,176	1.42
select_random_points	8,557	10,621	1.24
select_random_ranges	2,704	3,878	1.43
covering_index_scan	4,147	4,091	0.99
groupby_scan	29,017	34,229	1.18
index_join	5,766	7,888	1.37
index_join_scan	3,047	4,362	1.43
types_table_scan	1,056,240	1,252,512	1.19
table_scan	1,174,628	1,345,592	1.15
oltp_read_only	97,036	120,617	1.24
Average			1.27

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	181,768	256,592	1.41
oltp_insert	15,475	28,663	1.85
oltp_update_index	48,200	87,452	1.81
oltp_update_non_index	31,988	57,520	1.80
oltp_delete_insert	42,753	70,526	1.65
oltp_write_only	20,451	43,818	2.14
types_delete_insert	23,314	39,467	1.69
oltp_read_write	63,854	106,555	1.67
Average			1.75

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	91,001	52,705	0.58
oltp_range_select	16,329	15,189	0.93
oltp_sum_range	15,706	14,861	0.95
oltp_order_range	3,315	3,197	0.96
oltp_distinct_range	4,285	4,369	1.02
oltp_index_scan	10,578	7,982	0.75
select_random_points	15,914	13,403	0.84
select_random_ranges	9,560	6,392	0.67
covering_index_scan	11,075	6,780	0.61
groupby_scan	29,654	34,809	1.17
index_join	9,753	9,714	1.00
index_join_scan	4,119	4,826	1.17
types_table_scan	1,055,282	1,264,599	1.20
table_scan	1,175,378	1,347,248	1.15
oltp_read_only	198,002	158,248	0.80
Average			0.92

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	193,438	276,022	1.43
oltp_insert	21,463	35,912	1.67
oltp_update_index	73,002	112,356	1.54
oltp_update_non_index	54,474	79,977	1.47
oltp_delete_insert	62,589	92,217	1.47
oltp_write_only	42,340	62,943	1.49
types_delete_insert	38,004	52,039	1.37
oltp_read_write	86,023	126,202	1.47
Average			1.49

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	45,479	53,239	1.17
oltp_range_select	11,808	15,207	1.29
oltp_sum_range	11,266	14,764	1.31
oltp_order_range	2,736	3,182	1.16
oltp_distinct_range	3,724	4,401	1.18
oltp_index_scan	6,181	8,015	1.30
select_random_points	11,810	13,684	1.16
select_random_ranges	5,111	6,425	1.26
covering_index_scan	6,524	6,717	1.03
groupby_scan	29,145	34,519	1.18
index_join	7,354	9,574	1.30
index_join_scan	3,667	4,751	1.30
types_table_scan	1,052,239	1,265,787	1.20
table_scan	1,177,451	1,343,687	1.14
oltp_read_only	131,663	156,686	1.19
Average			1.21

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	20,666	76,935	3.72
oltp_insert_ac	24,851	95,489	3.84
oltp_update_index_ac	25,654	108,392	4.23
oltp_update_non_index_ac	21,950	87,312	3.98
oltp_delete_insert_ac	23,327	101,371	4.35
oltp_write_only_ac	23,962	99,714	4.16
types_delete_insert_ac	21,545	96,987	4.50
oltp_read_write_ac	28,427	104,210	3.67
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 (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	32,906	40,825	1.24
oltp_range_select	19,172	21,993	1.15
oltp_sum_range	17,924	21,581	1.20
oltp_order_range	3,475	3,901	1.12
oltp_distinct_range	4,577	5,117	1.12
oltp_index_scan	4,569	6,029	1.32
select_random_points	28,153	31,883	1.13
select_random_ranges	7,672	9,047	1.18
covering_index_scan	4,221	4,180	0.99
groupby_scan	35,591	40,774	1.15
index_join	8,173	10,587	1.30
index_join_scan	4,041	5,433	1.34
types_table_scan	1,054,856	1,232,851	1.17
table_scan	1,214,603	1,333,556	1.10
oltp_read_only	146,274	173,243	1.18
Average			1.18

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	248,986	360,739	1.45
oltp_insert	19,176	37,015	1.93
oltp_update_index	67,770	115,780	1.71
oltp_update_non_index	50,899	84,138	1.65
oltp_delete_insert	49,250	96,700	1.96
oltp_write_only	27,793	57,973	2.09
types_delete_insert	33,873	55,236	1.63
oltp_read_write	101,925	158,225	1.55
Average			1.75

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	103,266	64,188	0.62
oltp_range_select	25,956	24,614	0.95
oltp_sum_range	25,679	24,371	0.95
oltp_order_range	4,351	4,258	0.98
oltp_distinct_range	5,484	5,549	1.01
oltp_index_scan	11,897	8,946	0.75
select_random_points	37,046	36,030	0.97
select_random_ranges	15,202	11,918	0.78
covering_index_scan	11,391	7,068	0.62
groupby_scan	37,043	41,460	1.12
index_join	12,409	12,706	1.02
index_join_scan	5,020	5,848	1.16
types_table_scan	1,037,891	1,232,146	1.19
table_scan	1,206,281	1,328,786	1.10
oltp_read_only	255,527	211,249	0.83
Average			0.94

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	261,150	380,606	1.46
oltp_insert	26,322	46,962	1.78
oltp_update_index	98,984	142,371	1.44
oltp_update_non_index	76,702	104,791	1.37
oltp_delete_insert	77,358	119,734	1.55
oltp_write_only	51,830	77,332	1.49
types_delete_insert	50,552	69,385	1.37
oltp_read_write	129,364	175,583	1.36
Average			1.48

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	57,124	65,138	1.14
oltp_range_select	21,959	24,756	1.13
oltp_sum_range	20,424	24,600	1.20
oltp_order_range	3,887	4,207	1.08
oltp_distinct_range	4,887	5,456	1.12
oltp_index_scan	7,296	8,854	1.21
select_random_points	31,319	35,845	1.14
select_random_ranges	10,264	11,913	1.16
covering_index_scan	6,833	6,982	1.02
groupby_scan	35,777	41,025	1.15
index_join	9,655	12,988	1.35
index_join_scan	4,451	5,880	1.32
types_table_scan	1,046,396	1,233,034	1.18
table_scan	1,201,798	1,328,491	1.11
oltp_read_only	181,925	211,436	1.16
Average			1.16

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	25,149	84,519	3.36
oltp_insert_ac	24,443	104,706	4.28
oltp_update_index_ac	31,262	115,381	3.69
oltp_update_non_index_ac	25,120	93,885	3.74
oltp_delete_insert_ac	25,541	108,599	4.25
oltp_write_only_ac	26,535	105,415	3.97
types_delete_insert_ac	22,588	96,437	4.27
oltp_read_write_ac	32,117	116,417	3.62
Average			3.90

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	22,318	26,829	1.20
oltp_range_select	10,750	11,111	1.03
oltp_sum_range	10,131	10,232	1.01
oltp_order_range	2,401	2,608	1.09
oltp_distinct_range	3,182	3,470	1.09
oltp_index_scan	3,531	4,607	1.30
select_random_points	14,342	16,941	1.18
select_random_ranges	3,006	4,089	1.36
covering_index_scan	3,103	3,324	1.07
groupby_scan	27,197	27,815	1.02
index_join	5,465	7,196	1.32
index_join_scan	3,302	4,747	1.44
types_table_scan	887,486	990,592	1.12
table_scan	1,047,494	1,087,334	1.04
oltp_read_only	90,740	103,537	1.14
Average			1.16

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	161,045	236,564	1.47
oltp_insert	14,388	27,487	1.91
oltp_update_index	50,831	99,784	1.96
oltp_update_non_index	36,254	62,474	1.72
oltp_delete_insert	36,257	74,254	2.05
oltp_write_only	20,606	43,995	2.14
types_delete_insert	23,500	39,885	1.70
oltp_read_write	59,935	98,937	1.65
Average			1.82

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	47,052	34,938	0.74
oltp_range_select	13,641	12,087	0.89
oltp_sum_range	13,471	11,779	0.87
oltp_order_range	2,843	2,833	1.00
oltp_distinct_range	3,649	3,601	0.99
oltp_index_scan	6,410	5,915	0.92
select_random_points	17,141	18,206	1.06
select_random_ranges	5,749	5,044	0.88
covering_index_scan	6,119	4,585	0.75
groupby_scan	27,705	28,097	1.01
index_join	7,098	8,738	1.23
index_join_scan	3,760	4,987	1.33
types_table_scan	893,602	991,284	1.11
table_scan	1,049,005	1,084,272	1.03
oltp_read_only	125,523	115,787	0.92
Average			0.98

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	221,451	318,984	1.44
oltp_insert	42,811	56,295	1.31
oltp_update_index	168,109	191,284	1.14
oltp_update_non_index	137,188	119,626	0.87
oltp_delete_insert	131,370	139,835	1.06
oltp_write_only	93,212	89,901	0.96
types_delete_insert	89,819	80,282	0.89
oltp_read_write	191,409	178,210	0.93
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	30,671	34,828	1.14
oltp_range_select	12,015	12,176	1.01
oltp_sum_range	11,635	11,557	0.99
oltp_order_range	2,694	2,790	1.04
oltp_distinct_range	3,437	3,600	1.05
oltp_index_scan	4,871	5,804	1.19
select_random_points	16,099	18,396	1.14
select_random_ranges	4,141	4,988	1.20
covering_index_scan	4,569	4,626	1.01
groupby_scan	27,341	28,041	1.03
index_join	6,538	8,645	1.32
index_join_scan	3,713	5,020	1.35
types_table_scan	887,327	996,507	1.12
table_scan	1,047,522	1,085,096	1.04
oltp_read_only	102,430	115,696	1.13
Average			1.12

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	32,951	88,192	2.68
oltp_insert_ac	36,973	108,431	2.93
oltp_update_index_ac	39,188	110,741	2.83
oltp_update_non_index_ac	34,581	109,802	3.18
oltp_delete_insert_ac	37,908	117,107	3.09
oltp_write_only_ac	35,893	110,753	3.09
types_delete_insert_ac	32,577	96,191	2.95
oltp_read_write_ac	44,201	108,405	2.45
Average			2.90

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 (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	29,329	34,847	1.19
oltp_range_select	13,663	13,709	1.00
oltp_sum_range	12,126	13,472	1.11
oltp_order_range	3,197	3,238	1.01
oltp_distinct_range	4,205	4,375	1.04
oltp_index_scan	4,445	5,789	1.30
select_random_points	17,524	20,104	1.15
select_random_ranges	4,111	5,328	1.30
covering_index_scan	4,751	4,419	0.93
groupby_scan	33,802	37,073	1.10
index_join	6,807	8,785	1.29
index_join_scan	4,550	5,555	1.22
types_table_scan	1,152,471	1,263,483	1.10
table_scan	1,356,184	1,362,647	1.00
oltp_read_only	119,219	132,638	1.11
Average			1.12

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	230,436	347,409	1.51
oltp_insert	21,950	40,116	1.83
oltp_update_index	72,149	136,922	1.90
oltp_update_non_index	50,870	89,188	1.75
oltp_delete_insert	51,712	107,011	2.07
oltp_write_only	29,950	63,822	2.13
types_delete_insert	33,093	54,596	1.65
oltp_read_write	83,145	136,222	1.64
Average			1.81

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	112,381	64,055	0.57
oltp_range_select	22,339	16,669	0.75
oltp_sum_range	20,086	16,496	0.82
oltp_order_range	4,178	3,564	0.85
oltp_distinct_range	4,939	4,744	0.96
oltp_index_scan	13,209	9,053	0.69
select_random_points	26,386	23,401	0.89
select_random_ranges	12,246	8,176	0.67
covering_index_scan	13,453	7,776	0.58
groupby_scan	35,112	37,745	1.07
index_join	12,015	11,128	0.93
index_join_scan	5,806	6,215	1.07
types_table_scan	1,155,615	1,255,484	1.09
table_scan	1,376,394	1,371,002	1.00
oltp_read_only	239,497	174,179	0.73
Average			0.84

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	247,463	381,647	1.54
oltp_insert	55,062	52,668	0.96
oltp_update_index	112,636	172,908	1.54
oltp_update_non_index	101,217	114,842	1.13
oltp_delete_insert	87,535	137,699	1.57
oltp_write_only	96,605	87,769	0.91
types_delete_insert	53,394	74,858	1.40
oltp_read_write	139,301	158,743	1.14
Average			1.27

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	57,130	63,661	1.11
oltp_range_select	17,900	16,734	0.93
oltp_sum_range	15,865	16,528	1.04
oltp_order_range	3,678	3,595	0.98
oltp_distinct_range	4,605	4,732	1.03
oltp_index_scan	7,815	9,066	1.16
select_random_points	21,062	24,295	1.15
select_random_ranges	7,129	8,221	1.15
covering_index_scan	8,919	7,779	0.87
groupby_scan	34,484	37,588	1.09
index_join	9,161	11,368	1.24
index_join_scan	5,162	6,199	1.20
types_table_scan	1,160,800	1,264,034	1.09
table_scan	1,371,918	1,366,068	1.00
oltp_read_only	159,243	174,487	1.10
Average			1.08

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	16,406	60,118	3.66
oltp_insert_ac	18,223	76,218	4.18
oltp_update_index_ac	20,655	93,530	4.53
oltp_update_non_index_ac	16,782	72,730	4.33
oltp_delete_insert_ac	18,021	86,514	4.80
oltp_write_only_ac	18,491	84,797	4.59
types_delete_insert_ac	15,299	73,171	4.78
oltp_read_write_ac	24,689	94,405	3.82
Average			4.34

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.