2026-04-06 Release by BrentOzar · Pull Request #3885 · BrentOzarULTD/SQL-Server-First-Responder-Kit

BrentOzar · 2026-04-06T23:26:10Z

No description provided.

…th BEGIN/END Co-authored-by: BrentOzar <245462+BrentOzar@users.noreply.github.com>

…egin-end Wrap all stored procedure bodies with BEGIN/END

…mpts for null plans Co-authored-by: BrentOzar <245462+BrentOzar@users.noreply.github.com>

…2 with null plans Co-authored-by: BrentOzar <245462+BrentOzar@users.noreply.github.com>

@ai

…-null-query-plan sp_BlitzCache: Skip AI calls for null query plans, allow @ai=2 prompts without plans

Converts the Brent Ozar Unlimited SQL Server Setup Checklist from Word format to Markdown, covering server provisioning, storage, Windows configuration, SQL Server installation, security, alerting, monitoring, and application deployment. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Add SQL Server Setup Checklist documentation

@askai

Update parameter name from @askai to @ai in README

sp_BlitzLock - fix for deadlock group query number issue #3859

@ai

Ensure ai_prompt is never NULL when @ai > 0 by removing the QueryPlan IS NOT NULL filter from prompt building, adding explicit messages for missing query text/plan, and overriding with a simple message when both are NULL. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

@ai

sp_BlitzCache: fix ai_prompt returning nulls when @ai > 0

New stored procedure that collects all running session data, recommends which queries to kill (with reasons), optionally executes the kills, frees plan cache entries, and logs everything to an output table. Designed for performance emergencies where the alternative is restarting the server. Defaults to read-only mode and requires explicit targeting filters before killing anything. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

- Add version/build gating for sys.dm_exec_query_statistics_xml to avoid assertion errors on problematic SQL Server builds - Escape single quotes in @ParametersUsed string building to handle logins/app names with apostrophes - Remove MAX_GRANT_PERCENT from OPTION hint (not supported on all versions), keep MAXDOP 1 - Fix open_transaction_count to fall back to session-level value for sleeping sessions that have no active request - Fix is_read_only detection to use corrected open_transaction_count - Remove @OmitLogin from safety check since it's an exclusion filter, not an inclusive targeting filter - Add validation that @RequestsOlderThanMinutes must be >= 0 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

…vars Move license/header comments inside the stored procedure body so they're visible when opening the proc to troubleshoot. Simplify the live query plan DMV check to just EXISTS (if the DMV exists, it's supported - no need to also version-gate). Remove unused variables (@platform, @productversion, @ProductVersionMajor, @ProductVersionMinor, @EngineEdition). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Allow @ExecuteKills = 'Y' with no filters to kill all user sessions except the caller's own SPID. The no-filter guard in the recommendation logic now only applies in display mode (@ExecuteKills = 'N'). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

@loginname

Rename #sp_kill_sessions to #hitlist. Replace sequential subtractive filter logic (mark all, then un-mark per filter) with a single UPDATE that combines all filters as AND conditions. This ensures that passing multiple filters like @loginname + @appName only kills sessions matching both, not sessions matching either. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

@readonly

…pport Change @LeadBlockers, @readonly, @SPIDState, @OmitLogin, and @HasOpenTran from string defaults ('N'/'') to NULL. This makes filter detection consistent: NULL = not specified, value = active. EXEC sp_kill @ExecuteKills = 'Y' now cleanly kills all user sessions (except system, rollbacks, and the caller's own SPID) with no filters needed. Display mode with no filters still just shows sessions. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Seconds granularity is more useful in production emergencies where you may need to kill all queries running longer than e.g. 10 seconds. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Add sp_kill stored procedure (#3864)

…fallback Fixes #3868 — when a plan exceeds 128 levels of XML nesting, populate the QueryPlan column with the text plan wrapped in an XML processing instruction (same technique as sp_QuickieStore) instead of leaving it NULL and telling the user to run dm_exec_text_query_plan manually. Also uses StatementStartOffset/StatementEndOffset for statement-level plan retrieval, and separates truly missing plans into their own UPDATE. Tested on SQL2016, SQL2017, SQL2019, SQL2022, SQL2025. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Not sure if that null can happen, but the coalesce looks like a harmless tweak, so taking it. Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Not sure this can happen either, but this also looks harmless, so taking it. Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

…ed-plans sp_BlitzCache: return oversized plans inline instead of manual query fallback

Raise the minimum SQL Server version from 2008 to 2016, remove dead compatibility branches, modernize syntax, and improve performance. - Raise version floor to SQL Server 2016 in both gate checks - Remove dead code for pre-2016 versions (~80 lines): NULL-substitution ELSE branches, trigger accuracy workaround, deprecated memory clerk columns (single_pages_kb), always-true version guards - Replace 25 IF OBJECT_ID/DROP TABLE patterns with DROP TABLE IF EXISTS - Cache repeated COUNT(*) on filter temp tables into variables - Replace 4 WHILE loops for CSV hex parsing with STRING_SPLIT (~120 lines) - Simplify function stats and row goals version checks Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Fix database ID reference in sp_BlitzCache

Cleaning up files before Claude starts work.

…claude_starts #3881 release prep

20260406 release prep

Replace the old build-only script with a single PowerShell script that automates the full release workflow: bump version numbers, build the Install-*.sql files, install on local SQL Server, run test calls, and draft a PR if everything passes. Test Calls.sql is removed — its content is now embedded in the script. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

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

#3881 Build better merge/test/release process

Copilot

Pull request overview

Release bundle updating First Responder Kit scripts to v8.31 (2026-04-06), modernizing code paths for SQL Server 2016+ features, and adding a new emergency session-killing utility (sp_kill) plus updated documentation and release tooling.

Changes:

Added new sp_kill stored procedure and wired it into uninstall + README navigation/docs.
Updated multiple procedures to v8.31 and introduced/standardized SQL Server 2016+ gating and DROP TABLE IF EXISTS patterns.
Refactored parts of sp_BlitzCache (plan cache materialization, AI prompt robustness) and enhanced deadlock plan matching in sp_BlitzLock.

Reviewed changes

Copilot reviewed 19 out of 21 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
Uninstall.sql	Adds `sp_kill` to the uninstall list (fixes UNION chaining).
sp_kill.sql	New procedure to recommend/execute KILL actions with optional logging and cache clearing.
sp_ineachdb.sql	Bumps version/date to 8.31/20260406.
sp_DatabaseRestore.sql	Bumps version/date and wraps proc body with BEGIN/END.
sp_BlitzWho.sql	Adds session SET options + SQL 2016+ gating; modernizes temp table drops and removes older-version branches.
sp_BlitzLock.sql	Adds stmt offsets for deadlock stack frames; adjusts ranking/ordering and tightens plan matching by statement offsets.
sp_BlitzIndex.sql	Adds SQL 2016+ gating, modernizes temp table drops, removes older-version branches, bumps version/date, adds BEGIN/END.
sp_BlitzFirst.sql	Bumps version/date to 8.31/20260406.
sp_BlitzCache.sql	Adds SQL 2016+ gating; materializes plan cache by db; improves AI prompt generation and huge-plan handling; bumps version/date.
sp_BlitzBackups.sql	Bumps version/date to 8.31/20260406.
sp_BlitzAnalysis.sql	Bumps version/date and wraps proc body with BEGIN/END.
sp_Blitz.sql	Adds SQL 2016+ gating; modernizes temp table drops; removes legacy branches; bumps version/date; adds BEGIN/END.
README.md	Adds `sp_kill` section + updates AI param naming references; updates navigation.
Documentation/Setup_Checklist_SQL_Server.md	Adds a new SQL Server setup checklist doc.
Documentation/Development/Test sp_DatabaseRestore.sql	Removed legacy test script.
Documentation/Development/ReleaseProcess.md	Removed legacy release process doc.
Documentation/Development/Merge Blitz.ps1	Removed legacy merge script.
Documentation/Development/Build Installation Scripts.ps1	Adds a new automated release/test/build script.
Documentation/Development/_TestBed.sql	Removed legacy testbed script.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-04-06T23:32:49Z

README.md

+
+/* Kill sleeping sessions with open transactions idle for 5+ minutes: */
+EXEC sp_kill @HasOpenTran = 'Y', @SPIDState = 'S',
+    @RequestsOlderThanMinutes = 5, @ExecuteKills = 'Y';


The sp_kill README example uses @RequestsOlderThanMinutes, but the stored procedure parameter is @RequestsOlderThanSeconds. As written, this example will fail with an unknown parameter error unless the proc adds a minutes parameter/alias.

Suggested change

@RequestsOlderThanMinutes = 5, @ExecuteKills = 'Y';

@RequestsOlderThanSeconds = 300, @ExecuteKills = 'Y';

Copilot · 2026-04-06T23:32:50Z

README.md

+Parameters for targeting sessions:
+
+* @ExecuteKills = 'Y' or 'N' (default 'N') - whether to actually kill, or just show recommendations. When set to 'Y', you must also specify at least one targeting filter to prevent accidentally killing everything.
+* @SPID - target a specific session ID.
+* @LoginName - kill sessions belonging to this login.
+* @AppName - kill sessions from this application (supports LIKE wildcards).
+* @DatabaseName - kill sessions using this database.
+* @HostName - kill sessions from this host.
+* @LeadBlockers = 'Y' - kill only lead blockers (sessions blocking others but not blocked themselves).
+* @ReadOnly = 'Y' - only kill read-only queries (SELECTs with no writes).
+* @SPIDState - 'S' for sleeping sessions only, 'R' for running only, empty string for both (default).
+* @HasOpenTran = 'Y' - only target sessions with open transactions. Combine with @SPIDState = 'S' to catch sleeping sessions with forgotten transactions.
+* @RequestsOlderThanMinutes - only target sessions whose last request started at least this many minutes ago.
+* @OmitLogin - exclude sessions belonging to this login from kill recommendations.
+* @OrderBy - sort order for kill recommendations: duration (default), cpu, reads, writes, tempdb, transactions. Useful when you want to kill the worst offenders one by one.


sp_kill parameter docs don’t match the implementation:

@SPIDState is documented as "empty string for both", but the proc validates @SPIDState as NULL/S/R ('' will raise an error).

Docs say @ExecuteKills='Y' requires at least one targeting filter, but the proc currently recommends/kills all sessions when no filters are provided.
Please align the docs with the actual behavior, or add enforcement in the proc to match the documented safety requirement.

Copilot · 2026-04-06T23:32:50Z

sp_kill.sql

+	/* 7c: Otherwise, apply all filters as AND conditions in a single pass.
+	   A session must match ALL specified filters to be recommended for kill.
+	   With no filters, @ExecuteKills = 'Y' kills everything;
+	   @ExecuteKills = 'N' just shows sessions without recommending. */
+	ELSE
+	BEGIN
+		/* When in display mode with no filters, just show sessions without recommending kills */
+		IF @ExecuteKills = 'N'
+			AND @LoginName IS NULL


With @ExecuteKills = 'Y' and no targeting filters, the code path in section 7c marks every non-rollback user session as KillRecommended = 1, which means the kill loop will attempt to kill everything. This contradicts the README’s stated safety requirement, and is risky for an emergency tool.
Consider adding a guard that refuses to execute kills unless at least one targeting filter (or an explicit "kill all" confirmation parameter) is supplied.

Copilot · 2026-04-06T23:32:50Z

sp_BlitzLock.sql

                        DENSE_RANK() OVER (ORDER BY dp.event_date),
                    qn =
-                        ROW_NUMBER() OVER (PARTITION BY dp.event_date ORDER BY dp.event_date) - 1,
+                        ROW_NUMBER() OVER (PARTITION BY dp.event_date ORDER BY dp.event_date),


qn is now computed as ROW_NUMBER() (1-based), but later logic still treats it as if it were 0-based (e.g., parallel deadlocks filter d.qn < 2 and the deadlock_group label has a WHEN d.qn = 0 branch). This changes behavior: for parallel deadlocks it will now exclude the second query row (qn=2) from results.
Either revert to ROW_NUMBER() ... - 1 or update the downstream filters/labeling to be consistently 1-based.

Suggested change

ROW_NUMBER() OVER (PARTITION BY dp.event_date ORDER BY dp.event_date),

ROW_NUMBER() OVER (PARTITION BY dp.event_date ORDER BY dp.event_date) - 1,

Addresses Copilot review feedback from PR #3885: - @RequestsOlderThanMinutes -> @RequestsOlderThanSeconds (param name) - @SPIDState default is NULL, not empty string - @ExecuteKills docs now accurately reflect that no filter = kill all Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Fix sp_kill README docs to match proc behavior

The qn column was changed from ROW_NUMBER() - 1 (0-based) to ROW_NUMBER() (1-based), but downstream code still expects 0-based values. This caused parallel deadlocks to silently drop the second query (d.qn < 2 only keeps 1 row instead of 2) and made the WHEN d.qn = 0 labeling branch dead code. Fixes issue identified by Copilot in PR #3885. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

The CASE expression mapped qn=0 to "Query #1" but then used qn directly for all other values, so qn=1 also displayed as "Query #1". Replace the CASE with a simple qn + 1 conversion so labels are always correct (qn=0 → "Query #1", qn=1 → "Query #2", etc). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Fix sp_BlitzLock parallel deadlock filtering regression

Copilot

Pull request overview

Copilot reviewed 19 out of 21 changed files in this pull request and generated 5 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-04-06T23:55:46Z

sp_BlitzWho.sql

 		        qmg.dop AS degree_of_parallelism ,						'
-		        + 
-		        CASE @EnhanceFlag
-				        WHEN 1 THEN N'query_stats.last_dop,
+		        + N'query_stats.last_dop,
        query_stats.min_dop,
        query_stats.max_dop,
        query_stats.last_grant_kb,


In ExpertMode, the dynamic SELECT now unconditionally references sys.dm_exec_query_stats columns like last_dop/min_dop/max_dop/last_grant_kb. Those columns aren’t available on all SQL Server 2016 builds (they were previously gated via a build check), so this can raise an "Invalid column name" error on 2016 RTM/SP1. Either restore a column-existence/build check and project NULLs when absent, or tighten the minimum supported build (e.g., 2016 SP2+).

Copilot · 2026-04-06T23:55:47Z

sp_BlitzCache.sql

+    INSERT INTO #only_sql_handles (sql_handle)
+    SELECT CAST('' AS XML).value('xs:hexBinary( substring(sql:column("v.val"), sql:column("t.pos")) )', 'varbinary(max)')
+    FROM STRING_SPLIT(@OnlySqlHandles, ',') AS ss
+    CROSS APPLY (SELECT LTRIM(RTRIM(ss.value))) AS v(val)
+    CROSS APPLY (SELECT CASE WHEN SUBSTRING(v.val, 1, 2) = '0x' THEN 3 ELSE 1 END) AS t(pos)


This new parsing logic relies on STRING_SPLIT, which requires database compatibility level 130+. README explicitly says the kit can be installed in any database, so installing/running sp_BlitzCache in a DB with lower compatibility will fail here. Consider either (a) retaining a non-STRING_SPLIT fallback splitter, or (b) explicitly validating compatibility_level >= 130 up front and raising a clear error message.

Copilot · 2026-04-06T23:55:47Z

sp_kill.sql

+		/* Create table if it doesn't exist */
+		SET @StringToExecute = N'USE ' + @OutputDatabaseName + N';
+			IF EXISTS(SELECT * FROM ' + @OutputDatabaseName + N'.INFORMATION_SCHEMA.SCHEMATA WHERE QUOTENAME(SCHEMA_NAME) = ''' + @OutputSchemaName + N''')
+			AND NOT EXISTS (SELECT * FROM ' + @OutputDatabaseName + N'.INFORMATION_SCHEMA.TABLES WHERE QUOTENAME(TABLE_SCHEMA) = ''' + @OutputSchemaName + N''' AND QUOTENAME(TABLE_NAME) = ''' + @OutputTableName + N''')


The persistent output-table dynamic SQL uses USE <db> (line 654). USE isn’t supported in Azure SQL Database, yet the proc detects Azure (@AzureSQLDB) and even defaults @OutputDatabaseName to the current DB. Consider skipping the USE statement when @AzureSQLDB = 1 (and avoid 3-part naming), or otherwise block output logging on Azure with a clear message.

Copilot · 2026-04-06T23:55:47Z

Documentation/Development/Build Installation Scripts.ps1

+$azureContent = Get-ChildItem -Path $RepoRoot -Filter "sp_Blitz*.sql" |
+    Where-Object { $_.Name -ne "sp_Blitz.sql" -and $_.Name -notlike "*BlitzBackups*" -and $_.Name -notlike "*DatabaseRestore*" -and $_.Name -notlike "*BlitzFirst*" } |
+    ForEach-Object { Get-Content $_.FullName -Raw }
+


The merged Install-*.sql contents depend on the enumeration order of Get-ChildItem, which isn’t guaranteed to be stable across platforms/filesystems. To make releases reproducible (and avoid ordering-dependent install failures), consider explicitly sorting the file list (e.g., Get-ChildItem ... | Sort-Object Name) before concatenating content for both Install-Azure.sql and Install-All-Scripts.sql.

Copilot · 2026-04-06T23:55:47Z

sp_BlitzIndex.sql

+BEGIN
+	DECLARE @msg VARCHAR(8000);
+	SELECT @msg = 'Sorry, sp_BlitzIndex doesn''t work on versions of SQL prior to 2016.' + REPLICATE(CHAR(13), 7933);
+	PRINT @msg;
+	RETURN;


The new version gate blocks SQL Server versions prior to 2016, but the procedure’s help text later in the file still lists older version limitations (e.g., mentions 2005/2000). To avoid confusing users, update the “Known limitations” help output to match the new minimum supported version (2016+).

- sp_BlitzWho: Add SQL Server 2016 SP2+ minimum build check (build 5026) - sp_BlitzCache: Check compat level 130+ up front for STRING_SPLIT support - sp_kill: Remove USE statement for Azure SQL DB compatibility, use 3-part naming - sp_BlitzIndex: Update help text to say "2014 and older" instead of "2005 and 2000" Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

- sp_kill: Branch on @AzureSQLDB for output table creation — use 2-part names on Azure SQL DB, 3-part names on on-prem/MI - sp_BlitzWho: Update help text to say "2016 SP2 and newer" to match the version gate Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

The post-kill UPDATE at line 853 was unconditionally rebuilding @ObjectFullName as a 3-part name, which would fail on Azure SQL DB. Apply the same @AzureSQLDB branching used in Section 10. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

#3888 Fix 4 compatibility issues from Copilot review

Adding Azure SQL DB tests

Tisit and others added 30 commits December 2, 2025 00:23

sp_BlitzLock - fix duplicate plans

dd8bc10

Correct data types in new code

caa1a87

Initial plan

035916c

Wrap sp_Blitz, sp_BlitzIndex, sp_BlitzAnalysis, sp_DatabaseRestore wi…

9282b02

…th BEGIN/END Co-authored-by: BrentOzar <245462+BrentOzar@users.noreply.github.com>

Merge pull request #3851 from BrentOzarULTD/copilot/wrap-procs-with-b…

24012b2

…egin-end Wrap all stored procedure bodies with BEGIN/END

Initial plan

6fb7640

sp_BlitzCache: don't call AI when query plan is null, allow @ai=2 pro…

36fac7b

…mpts for null plans Co-authored-by: BrentOzar <245462+BrentOzar@users.noreply.github.com>

sp_BlitzCache: keep null plan guard for @ai=1, allow prompts for @ai=…

b71cb3a

…2 with null plans Co-authored-by: BrentOzar <245462+BrentOzar@users.noreply.github.com>

Merge pull request #3854 from BrentOzarULTD/copilot/fix-sp-blitzcache…

3249e99

…-null-query-plan sp_BlitzCache: Skip AI calls for null query plans, allow @ai=2 prompts without plans

Merge pull request #3858 from BrentOzarULTD/3857_setup_checklist

c6cf256

Add SQL Server Setup Checklist documentation

fix for deadlock group query number issue #3859

59bb718

Update parameter name from @askai to @ai in README

5f14af8

Merge pull request #3861 from ipodaru/patch-1

ceac95c

Update parameter name from @askai to @ai in README

Merge pull request #3860 from VladDBA/sp_BlitzLock_DeadlockGroupFix

2bba26a

sp_BlitzLock - fix for deadlock group query number issue #3859

Merge pull request #3863 from BrentOzarULTD/3855_sp_BlitzCache_nulls

0bea51b

sp_BlitzCache: fix ai_prompt returning nulls when @ai > 0

sp_kill: rename @RequestsOlderThanMinutes to @RequestsOlderThanSeconds

8795706

Seconds granularity is more useful in production emergencies where you may need to kill all queries running longer than e.g. 10 seconds. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Merge pull request #3865 from BrentOzarULTD/3864_sp_kill

d4dc7e8

Add sp_kill stored procedure (#3864)

Update sp_BlitzCache.sql

bec0671

Not sure if that null can happen, but the coalesce looks like a harmless tweak, so taking it. Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update sp_BlitzCache.sql

118faa4

Not sure this can happen either, but this also looks harmless, so taking it. Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Merge pull request #3869 from erikdarlingdata/3868-blitzcache-oversiz…

4ea809c

…ed-plans sp_BlitzCache: return oversized plans inline instead of manual query fallback

BrentOzar and others added 8 commits April 6, 2026 12:21

Merge pull request #3880 from shekarkola/patch-1

455f24b

Fix database ID reference in sp_BlitzCache

#3881 release prep

acc5a37

Cleaning up files before Claude starts work.

Merge pull request #3882 from BrentOzarULTD/3881_release_prep_before_…

d59f126

…claude_starts #3881 release prep

20260406 release prep

7b9f24a

Merge pull request #3883 from BrentOzarULTD/20260406_release_prep

8fef37c

20260406 release prep

Resolve merge conflicts — keep UTF-8 encoding fix

2d5366e

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

Merge pull request #3884 from BrentOzarULTD/3881_release_process

bdd73ca

#3881 Build better merge/test/release process

Copilot AI review requested due to automatic review settings April 6, 2026 23:26

Copilot started reviewing on behalf of BrentOzar April 6, 2026 23:26 View session

Copilot AI reviewed Apr 6, 2026

View reviewed changes

BrentOzar mentioned this pull request Apr 6, 2026

Fix sp_kill README docs to match proc behavior #3886

Merged

2 tasks

BrentOzar and others added 2 commits April 6, 2026 23:40

Merge pull request #3886 from BrentOzarULTD/3885_fix_sp_kill_docs

63470a8

Fix sp_kill README docs to match proc behavior

BrentOzar mentioned this pull request Apr 6, 2026

Fix sp_BlitzLock parallel deadlock filtering regression #3887

Merged

3 tasks

BrentOzar and others added 2 commits April 6, 2026 16:47

Merge pull request #3887 from BrentOzarULTD/3885_fix_blitzlock_qn

22b3d4f

Fix sp_BlitzLock parallel deadlock filtering regression

BrentOzar requested a review from Copilot April 6, 2026 23:49

Copilot started reviewing on behalf of BrentOzar April 6, 2026 23:49 View session

Copilot AI reviewed Apr 6, 2026

View reviewed changes

BrentOzar mentioned this pull request Apr 7, 2026

Fix 4 compatibility issues from Copilot review of PR #3885 #3888

Closed

BrentOzar mentioned this pull request Apr 7, 2026

#3888 Fix 4 compatibility issues from Copilot review #3889

Merged

6 tasks

BrentOzar and others added 5 commits April 7, 2026 04:06

Merge pull request #3889 from BrentOzarULTD/3888_copilot_compat_fixes

c20b37f

#3888 Fix 4 compatibility issues from Copilot review

Adding Azure SQL DB tests

73f7d18

Merge pull request #3890 from BrentOzarULTD/azure_test_scripts

b7623dd

Adding Azure SQL DB tests

BrentOzar closed this Apr 7, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

2026-04-06 Release#3885

2026-04-06 Release#3885
BrentOzar wants to merge 68 commits intomainfrom
dev

BrentOzar commented Apr 6, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Copilot AI Apr 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

	@RequestsOlderThanMinutes = 5, @ExecuteKills = 'Y';
	@RequestsOlderThanSeconds = 300, @ExecuteKills = 'Y';

	ROW_NUMBER() OVER (PARTITION BY dp.event_date ORDER BY dp.event_date),
	ROW_NUMBER() OVER (PARTITION BY dp.event_date ORDER BY dp.event_date) - 1,

Conversation

BrentOzar commented Apr 6, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 6, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants