CHORE: Add a multi-chunk LOB scenario to PR performance reports

[Gaurav Sharma (bewithgaurav)]

Work Item / Issue Reference

GitHub Issue: #554

Summary

Add one scenario that fetches a single 256 KiB VARCHAR(MAX) value through
fetchall(), bringing routine profiler coverage from 20 to 21 scenarios.

Validate the exact payload outside the timed fetch window and show available
diagnostic-call counts without treating missing instrumentation as zero.

No runtime driver changes, additional builds, platform changes, or regression
threshold changes.

[Copilot]

Copilot review overview

🟡 Changes recommended

Two moderate issues remain unresolved in profiler cleanup and diagnostic instrumentation/reporting.

Get a fresh assessment by requesting another Copilot review.

Review effort: Lite
Findings: 1

Open (1)

• Add instrumentation for raw SQLGetDiagRec calls · New

What changed in this PR

Adds a 256 KiB VARCHAR(MAX) fetchall() profiler scenario, updating reporting, tests, and documentation for 21 scenarios.

Changes:

• Adds and validates the multi-chunk LOB workload.
• Updates diagnostic reporting and scenario coverage.
• Documents measurement boundaries and limitations.

File	Summary
tests/​test_036_profiler_ci.py	Adds workload and reporting coverage.
eng/​profiler_benchmarks/​workloads.py	Adds the LOB workload; ctx.enable() should be protected by try/finally to prevent state contamination. (Moderate, 1 vote)
eng/​profiler_benchmarks/​report.py	Adds diagnostic-count handling; the SQLGetDiagRec branch is unreachable without instrumentation. (Moderate, 2 votes)
eng/​profiler_benchmarks/​README.md	Documents coverage and measurement behavior.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[github-actions]

PR Performance Report

✅ No regression detected

No consistent slowdowns detected across all 2 environments.

0 IMPROVEMENTS 0 SLOWDOWNS 2/2 ENVIRONMENTS

Coverage: 2 of 2 environments completed. Advisory result; does not block merging.

Performance diagnostics

Phase times are inclusive diagnostics and must not be added together. They identify where measured time changed, not why it changed.

No affected phases or call-count changes were recorded.

All database tasks and timings

Unix / SQL Server 2022

Database task	Before	After	Paired change	Result
Connection opening	10.921 ms	10.892 ms	+0.5%	no signal
SELECT queries	1.062 ms	1.021 ms	-3.8%	no signal
Row insertion	31.668 ms	31.551 ms	-0.6%	no signal
Executemany inserts	137.076 ms	136.176 ms	-0.2%	no signal
Fetch-all queries	124.638 ms	117.893 ms	-5.4%	no signal
Row-by-row fetching	12.637 ms	12.608 ms	-0.9%	no signal
Batched row fetching	111.081 ms	112.173 ms	+1.4%	no signal
Transaction commit and rollback	100.342 ms	104.437 ms	+4.7%	no signal
Arrow row fetching	89.892 ms	91.196 ms	+2.2%	no signal
100,000-row insertion	439.748 ms	413.443 ms	-3.8%	no signal
Row fetching in batches of 100	108.466 ms	109.691 ms	+1.1%	no signal
Row fetching in batches of 10,000	117.691 ms	134.682 ms	+16.3%	no signal
Repeated positional queries	30.183 ms	30.922 ms	+2.4%	no signal
Repeated named-parameter queries	32.463 ms	32.909 ms	+1.4%	no signal
Legacy 100,000-row insertion	307.231 ms	307.208 ms	-0.9%	no signal
Insertion with explicit input sizes	443.257 ms	432.540 ms	-1.4%	no signal
Joined aggregation queries	184.863 ms	184.924 ms	+0.2%	no signal
Large joined-result fetching	177.773 ms	178.564 ms	-0.5%	no signal
1.2-million-row fetching	3504.976 ms	3471.509 ms	-0.5%	no signal
Common table expression queries	5.591 ms	5.585 ms	-0.5%	no signal
256 KiB VARCHAR(MAX) / fetchall()	1.262 ms	1.279 ms	+0.2%	no signal

Unix / SQL Server 2025

Database task	Before	After	Paired change	Result
Connection opening	96.075 ms	96.496 ms	+0.5%	no signal
SELECT queries	1.083 ms	1.256 ms	+16.0%	no signal
Row insertion	31.841 ms	34.161 ms	+7.6%	no signal
Executemany inserts	131.452 ms	131.089 ms	-0.2%	no signal
Fetch-all queries	116.419 ms	115.088 ms	-1.7%	no signal
Row-by-row fetching	12.760 ms	12.905 ms	-0.1%	no signal
Batched row fetching	110.637 ms	109.642 ms	-0.9%	no signal
Transaction commit and rollback	103.474 ms	103.322 ms	-0.8%	no signal
Arrow row fetching	87.564 ms	88.469 ms	-0.4%	no signal
100,000-row insertion	389.020 ms	431.633 ms	+10.1%	no signal
Row fetching in batches of 100	106.091 ms	105.637 ms	-0.4%	no signal
Row fetching in batches of 10,000	117.761 ms	132.632 ms	-1.4%	no signal
Repeated positional queries	30.907 ms	30.295 ms	-1.5%	no signal
Repeated named-parameter queries	33.059 ms	33.135 ms	+0.1%	no signal
Legacy 100,000-row insertion	304.258 ms	338.725 ms	+3.1%	no signal
Insertion with explicit input sizes	432.499 ms	431.580 ms	+1.2%	no signal
Joined aggregation queries	169.941 ms	168.857 ms	-1.0%	no signal
Large joined-result fetching	179.175 ms	178.548 ms	+0.7%	no signal
1.2-million-row fetching	3516.432 ms	3560.276 ms	+2.0%	no signal
Common table expression queries	5.287 ms	5.265 ms	-0.7%	no signal
256 KiB VARCHAR(MAX) / fetchall()	1.388 ms	1.388 ms	+1.0%	no signal

Build and measurement details

ADO build 178002

PR head: 26075a47327ce07b7fb11097da7a72c90901a891
Base: 29fa5546eb7f24df0d5d42276549aa789937880d
Measured merge: d4522a39de1421ce5d8de560ac6981c4e2227b02

• Unix / SQL Server 2022: Python 3.12.3, x86_64, SQL 16.0.4295.3; 5 paired comparisons and 1 warmup.
• Unix / SQL Server 2025: Python 3.12.3, x86_64, SQL 17.0.5005.3; 5 paired comparisons and 1 warmup.

A consistent change requires more than 20% median paired movement, at least 1 ms between the median runtimes, and at least 80% of pairs exceeding the relative threshold in the same direction. A slowdown without enough pair agreement is reported as inconsistent.

The displayed change is the median of paired before-and-after ratios. It is not recalculated from the two displayed median runtimes.

Both revisions use profiling-enabled builds on the same agent and database, with alternating order and discarded warmups. Results are diagnostic and do not represent production-wheel latency.

Raw samples and logs are attached to the ADO run as profiler-* artifacts.

[Copilot]

Copilot review overview

🔵 Needs a closer look

A moderate test assertion does not verify the exact advertised SQL expression and repeat count.

Review effort: Lite
Findings: 1

Open (1)

• Add instrumentation for raw SQLGetDiagRec calls

[Copilot]

Copilot review overview

🟢 Approval recommended

No unresolved review issues remain.

Review effort: Lite
Findings: 1

Open (1)

• Add instrumentation for raw SQLGetDiagRec calls

[github-actions]

📊 Code Coverage Report

🔥 Diff Coverage 100%

🎯 Overall Coverage 84%

📈 Total Lines Covered: 9358 out of 11042 📁 Project: mssql-python

---

Diff Coverage

Diff: main...HEAD, staged and unstaged changes

No lines with coverage information in this diff.

---

📋 Files Needing Attention

📉 Files with overall lowest coverage (click to expand)

mssql_python.pybind.performance_counter.hpp: 0.7%
mssql_python.pybind.logger_bridge.cpp: 57.9%
mssql_python.pybind.ddbc_bindings.h: 62.6%
mssql_python.pybind.logger_bridge.hpp: 70.8%
mssql_python.pybind.ddbc_bindings.cpp: 79.1%
mssql_python.pybind.connection.connection_pool.cpp: 82.3%
mssql_python.pybind.connection.connection.cpp: 83.1%
mssql_python.logging.py: 86.2%
mssql_python.pooling.py: 90.1%
mssql_python.pybind.fetch_temporal.hpp: 92.1%

---

🔗 Quick Links

⚙️ Build Summary	📋 Coverage Details
View Azure DevOps Build	Browse Full Coverage Report

[Copilot]

Copilot review overview

🔵 Needs a closer look

The LOB test does not assert that the query requests the complete 262,144-byte payload.

Review effort: Lite
Findings: None

Resolved since last review (1)

• Add instrumentation for raw SQLGetDiagRec calls

[Copilot]

Copilot review overview

🟡 Changes recommended

Wrap the overlong assertion to satisfy the repository’s formatting check.

Get a fresh assessment by requesting another Copilot review.

Review effort: Lite
Findings: 1

Open (1)

• Wrap assertion to satisfy Black line length · New

[Copilot]

Copilot review overview

🟢 Approval recommended

No unresolved blocking issues were identified.

Review effort: Lite
Findings: None

Resolved since last review (1)

• Wrap assertion to satisfy Black line length

[Copilot]

Copilot review overview

🟢 Approval recommended

The reviewed changes are complete and introduce no unresolved approval blockers.

Review effort: Lite
Findings: None

[Copilot]

Copilot review overview

🟢 Approval recommended

No unresolved blocking issues were identified.

Review effort: Lite
Findings: None

[gargsaumya]

LGTM