…ions

sp_QuickieStore: added sys.dm_db_tuning_recommendations to Expert Mode output

…items

@find_high_impact = 1 identifies the vital few queries consuming
disproportionate resources. Scores queries across 6 dimensions (CPU,
duration, reads, writes, memory, executions) using PERCENT_RANK,
shows share-of-total per metric, and surfaces diagnostic signals
(parameter sensitivity, plan instability, spills, wait time).
Tested on SQL Server 2016-2025.

Also fixes two items from Reece's tuning recommendations PR (#691):
- SUBSTRING parsing: added IIF fallback when CHARINDEX returns 0
  so field reordering in future SQL Server versions won't throw
  Msg 536 (invalid length parameter)
- OUTER APPLY for plan_force_recommendation_status: pulled into its
  own @new = 1 AND @expert_mode = 1 block so it doesn't run as
  wasted work when expert mode is off

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

…high_impact

Applied SWITCHOFFSET (or AT TIME ZONE when @Timezone is set) to the
start_date and end_date columns in the high impact output, matching
how the normal analysis displays execution times.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

… output

Renamed reads_share to physical_reads_share, and updated high_signals
and volatile_metrics to say "physical reads" instead of just "reads"
to avoid confusion with logical reads. Also updated help text.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

#692)

When indexes form a chain of key subsets (A ⊂ B ⊂ C), Rule 3 assigned
each subset to its nearest superset (A → B, B → C). Rule 6 only
collected includes from direct subsets of the final superset, so
transitive subsets' include columns were silently dropped from the
merge script.

Added chain resolution between Rule 3 and Rule 4 that flattens
A → B → C into A → C, so all include columns propagate correctly.

Also fixed superseded_by text to aggregate all subset names via
FOR XML PATH when multiple subsets target the same superset, instead
of non-deterministically picking one.

Tested on SQL Server 2016, 2017, 2019, 2022.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Add parallel efficiency metric using (cpu/elapsed - 1) / (dop - 1)
formula to score how effectively queries use parallelism. Also add
four new diagnostic rules:

- intermittent waits: duration volatile but CPU stable (external waits)
- rare but expensive: low execution count but high resource share
- adhoc bloat: many query_id variants per query_hash
- scan heavy: high physical reads per execution

Tested clean on sql2016-sql2025 across PerformanceMonitor,
StackOverflow2013, StackOverflow2010, and hammerdb_tpch databases.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

…act (#693)

New result set fires before the query details showing what percentage of
total resources the surfaced queries account for. Classifies workloads as
Concentrated (>= 50%), Moderate (25-49%), or Flat (< 25%) with actionable
recommendations for each profile. Helps users identify "death by a thousand
cuts" workloads where no single query dominates. Tested on sql2016, sql2017,
sql2022, sql2025.

Also adds documentation to the @help output explaining the new columns.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

SELECT INTO with OPTION(RECOMPILE) produces parallel insert plans that
cause massive LATCH_EX waits on tempdb allocation pages — even when zero
rows are produced (common with @warnings_only = 1).

Actual execution plans showed 253 seconds of cumulative LATCH_EX across
6 queries, with trivial CPU time (41-183ms). Each query spawned 8
parallel threads spending 7+ seconds each in latch coordination over
empty result sets.

Adding MAXDOP 1 eliminates the parallel insert while preserving
RECOMPILE for plan optimization. Tested on SQL Server 2022:

  Before: 50-54 seconds (avg 52,332ms)
  After:   4-5 seconds (avg 4,684ms)

~10x improvement. Results consistent at idle and under TPC-H load.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

…dop1-select-into

sp_HealthParser: add MAXDOP 1 to SELECT INTO queries

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

…ind_high_impact

INSERT...EXEC captures all result sets, not just the first. The batch
had three SELECTs (queries, plans, objects) so plan_ids and object_ids
were concatenated into #hi_id_staging_queries. Removed the extra two
SELECTs since they are already executed separately below.

Tested on SQL2022 against PerformanceMonitor and hammerdb_tpcc.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

…pact

Previously any query with >1 plan was flagged, which produced false
positives on stable workloads with many executions. Now requires
fewer than 5 executions per plan (frequent recompilation) and excludes
queries with RECOMPILE hints where multiple plans are expected.

Tested on SQL2022 against PerformanceMonitor and hammerdb_tpcc.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Converts bug report, feature request, and question templates from
freeform markdown to structured YAML forms with enforced required
fields. Adds config.yml to disable blank issues and link to Discussions.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Replace markdown issue templates with YAML issue forms

Version bumps (x.3 → x.4, date → 20260401):
- sp_HealthParser 3.4
- sp_HumanEvents 7.4
- sp_HumanEventsBlockViewer 5.4
- sp_IndexCleanup 2.4
- sp_LogHunter 3.4
- sp_PerfCheck 2.4
- sp_PressureDetector 6.4
- sp_QueryReproBuilder 1.4
- sp_QueryStoreCleanup 1.4
- sp_QuickieStore 6.4

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

…ease

Bump all SP versions for April 2026 release