Thursday, September 8, 2016

Collect SQL Server Registry information using PowerShell

Do you use third party tools to document state of your SQL Server?
If not, that script is for you!

At first, you will know what is your SQL Server is up to.
At second, that might be your baseline document, to which you can compare a current SQL Server state over the time.
At third, that is a priceless piece of documentation!!! (I mean FREE!!!) which you can put in a folder and report to your boss. 



So, use it with a little disclaimer:
1. You have to run this script as an administrator. To do this: Right click on PowerShell Icon and choose: "Run As administrator":

2. ONLY YOU - who is responsible to run this script. If accidentally it will format your hard drive or delete all your backups - you will be guilty! So, see what it is doing and decide if it is safe for you to run.

3. I've tested it on SQL Servers 2014 & 2016. It also supposed to work with SQL 2012. Versions 2008 & 2005 might have some slide problems.

4. I'm SQL guy and know that this script is not perfect. It might miss some data or even go with an error in some special situation. So, see #2 and make changes/adjustments yourself and if you can't, leave your angry note here and I might help.

5. The script collects info ONLY regarding SQL Engine. If I see the popularity of it I might extend it to SSAS/SSIS/SSRS/SQL Agent etc.




# PowerShell script to retrieve SQL Server Information from the registry
$m = "HKLM:\SOFTWARE\Microsoft\Microsoft SQL Server"
$inst = (get-itemproperty $m).InstalledInstances
foreach ($i in $inst)
{
       $in = (Get-ItemProperty "$m\Instance Names\SQL").$i
       $p = $m+"\"+$in
       "-------------------------------------------------------------------`nServer Properties: "
       If ($i -Eq "MSSQLSERVER") {$r=" (Default)"} Else {$r=""}
       "`tInstance Name : $i$r"
       "`tVersion       : "+(Get-ItemProperty "$p\Setup").Version
       "`tSP            : "+(Get-ItemProperty "$p\Setup").SP
       "`tEdition       : "+(Get-ItemProperty "$p\Setup").Edition
       "`tLanguage      : "+(Get-ItemProperty "$p\Setup").Language
       "`tCollation     : "+(Get-ItemProperty "$p\Setup").Collation
       "`tSqlProgramDir : "+(Get-ItemProperty "$p\Setup").SqlProgramDir
"`tSQLPath       : "+(Get-ItemProperty "$p\Setup").SQLPath
       "`tSQLBinRoot    : "+(Get-ItemProperty "$p\Setup").SQLBinRoot
       "`tSQLDataRoot   : "+(Get-ItemProperty "$p\Setup").SQLDataRoot
"`tErrorDumpDir  : "+(Get-ItemProperty "$p\CPE").ErrorDumpDir
       $p="$p\MSSQLServer"
       "`tBackupDir     : "+(Get-ItemProperty "$p\").BackupDirectory
       If ((Get-ItemProperty "$p\").LoginMode -Eq 1) {$r="Windows"} `
       ElseIf ((Get-ItemProperty "$p\").LoginMode -Eq 2) {$r="Mixed"} Else {$r="Other"}
       "`tLoginMode     : $r Authentication Mode"
       If ((Get-ItemProperty "$p\HADR").HADR_Enabled) {$r="Enabled"} Else {$r="Disabled"}
       "`tHADR          : $r"
       "`nStartup Parameters: "
       (Get-ItemProperty "$p\Parameters" | Select SQLArg*  | Format-List | Out-String ).trim() -replace "SQLArg","`tSQLArg"

       $p="$p\SuperSocketNetLib"
       If ((Get-ItemProperty "$p").ForceEncryption) {$r="Yes"} Else {$r="No"}
       "`nProtocol Properties: "
       "`tForceEncryption: $r"
       If ((Get-ItemProperty "$p").HideInstance) {$r="Yes"} Else {$r="No"}
       "`tHideInstance: $r"
       $r="`t"+(Get-ItemProperty "$p\AdminConnection").DisplayName+": "
       $r=$r+(Get-ItemProperty "$p\AdminConnection\Tcp").DisplayName+" "
       $r+(Get-ItemProperty "$p\AdminConnection\Tcp").TcpDynamicPorts
       "`nNetwork Protocols: "
       If ((Get-ItemProperty "$p\Np").Enabled) {$r="Enabled"} Else {$r="Disabled"}; "`t"+((Get-ItemProperty "$p\Np").DisplayName)+": $r";
       If ((Get-ItemProperty "$p\Sm").Enabled) {$r="Enabled"} Else {$r="Disabled"}; "`t"+((Get-ItemProperty "$p\Sm").DisplayName)+": $r"
       If ((Get-ItemProperty "$p\Via").Enabled) {$r="Enabled"} Else {$r="Disabled"};"`t"+((Get-ItemProperty "$p\Via").DisplayName)+": $r"
       If ((Get-ItemProperty "$p\Tcp").Enabled) {$r="Enabled"} Else {$r="Disabled"};"`t"+((Get-ItemProperty "$p\Tcp").DisplayName)+": $r"
       If ((Get-ItemProperty "$p\Tcp").ListenOnAllIPs) {$r="Yes"} Else {$r="No"}; "`t"+"TCP ListenOnAllIPs:$r "

       Get-ChildItem "$p\Tcp" | ForEach-Object {Get-ItemProperty $_.pspath} `
       | Format-Table -Autosize -Property PSChildName, Enabled, Active, TcpPort, TcpDynamicPorts, IpAddress, DisplayName
}

Just in case you have an idea what to expect from this script, here is an output from my test system:

-------------------------------------------------------------------
Instance Name: Test2016

Version       : 13.0.1601.5
SP            : 0
Edition       : Developer Edition
Language      : 1033
Collation     : SQL_Latin1_General_CP1_CI_AS
SqlProgramDir : C:\SQLServer2016\
SQLPath       : C:\SQLServer2016\MSSQL13.Test2016\MSSQL
SQLBinRoot    : C:\SQLServer2016\MSSQL13.Test2016\MSSQL\Binn
SQLDataRoot   : C:\SQLServer2016\MSSQL13.Test2016\MSSQL
FeatureList   : SQL_Engine_Core_Inst=3 SQL_DataFiles_Core_Inst=3 SQL_ENGINE_DB_CNI=3 SQL_CMDLINETOOLS_CNI=3 SQL_DUMPER_CNI=3
                SQL_Engine_CNI=3 SQL_ENGINE_CORE_CNI=3 SQL_LEGACYTOOLS_CNI=3 SQL_REPL_ENGINE_SUPPORT_CNI=3 SQL_SLP_ENGINE_SUPPORT_CNI=3
                MPT_AGENT_CORE_CNI=3 MPT_AGENT_CORE_MNGD_CNI=3 SQL_AGENT_FNI=3 SQL_DATA_COLLECTOR_FNI=3 SQL_LEGACYTOOLS_FNI=3
                SQL_MAIL_FNI=3 SQL_REPL_ENGINE_SUPPORT_FNI=3 SQL_UPGRADESCRIPTS_FNI=3
Properties:
ErrorDumpDir: C:\SQLServer2016\MSSQL13.Test2016\MSSQL\LOG\
BackupDirectory: C:\SQLServer2016\MSSQL13.Test2016\MSSQL\Backup
LoginMode: Mixed Authentication Mode
HADR: Disabled
ForceEncryption: No
HideInstance: No
Dedicated Administrative Connection: TCP/IP 56212

Startup Parameters:
SQLArg0 : -dC:\SQLServer2016\MSSQL13.Test2016\MSSQL\DATA\master.mdf
SQLArg1 : -eC:\SQLServer2016\MSSQL13.Test2016\MSSQL\Log\ERRORLOG
SQLArg2 : -lC:\SQLServer2016\MSSQL13.Test2016\MSSQL\DATA\mastlog.ldf

Network Protocols:
Named Pipes: Disabled
Shared Memory: Disabled
VIA: Disabled
TCP/IP: Disabled
TCP ListenOnAllIPs: Yes
PSChildName Enabled Active TcpPort TcpDynamicPorts IpAddress                    DisplayName
----------- ------- ------ ------- --------------- ---------                    -----------
IP1               0      1         0               fe80::30d2:a78a:a5f2:62d8%97 Specific IP Address
IP10              0      1         0               192.168.56.1                 Specific IP Address
IP11              0      1         0               fe80::7d86:4ce4:724a:743d%5  Specific IP Address
IP12              0      1         0               169.254.116.61               Specific IP Address
IP13              0      1         0               ::1                          Specific IP Address
IP14              0      1         0               127.0.0.1                    Specific IP Address
IP15              0      1         0               fe80::5efe:192.168.58.1%9    Specific IP Address
IP16              0      1         0               fe80::5efe:192.168.57.1%21   Specific IP Address
IP17              0      1         0               fe80::5efe:192.168.56.1%11   Specific IP Address
IP18              0      1         0               fe80::5efe:192.168.1.141%101 Specific IP Address
IP2               0      1         0               192.168.57.1                 Specific IP Address
IP3               0      1         0               fe80::65c2:c76:2b5c:a889%93  Specific IP Address
IP4               0      1         0               192.168.58.1                 Specific IP Address
IP5               0      1         0               fe80::7944:dba3:1931:86e2%36 Specific IP Address
IP6               0      1         0               192.168.1.141                Specific IP Address
IP7               0      1         0               fe80::d156:33ec:a379:337a%34 Specific IP Address
IP8               0      1         0               169.254.51.122               Specific IP Address
IP9               0      1         0               fe80::b882:c715:7b1e:4a75%20 Specific IP Address
IPAll                              0                                            Any IP Address

Tuesday, September 6, 2016

Auto-Parameterization in SQL Server

I've known about Auto-Parameterization in SQL Server for a long time, but just recently got into a situation when it magically worked or did not work for me.

I've done a research and got some very interesting facts on how SQL Server determines to auto-parameterize a query or not.

I've used following script to create and populate a test table:
DROP TABLE IF EXISTS tbl_TestPlanCache;
GO
CREATE TABLE tbl_TestPlanCache(
    SearchKey INT,
    SearchValue NVARCHAR(100)
);
GO
INSERT INTO tbl_TestPlanCache(SearchKey, SearchValue)
SELECT message_id, LEFT(text,10) FROM sys.messages
WHERE language_id = 1033;
GO

I used following list of cases to come up with different indexing scenarios:

-- Cases
-- 1
-- No Indexes
-- 2
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey)
-- 3
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey)
--CREATE INDEX IX_TestPlanCache_SearchValue ON tbl_TestPlanCache(SearchValue)
-- 4
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey) INCLUDE (SearchValue)
-- 5
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey) INCLUDE (ID)
-- 6
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey) INCLUDE (ID, SearchValue)
-- 7
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey, SearchValue)
-- 8
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey, ID)
-- 9
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey, ID, SearchValue)
-- 10
--CREATE INDEX IX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey, ID, SearchValue)
--CREATE INDEX IX_TestPlanCache_SearchValue ON tbl_TestPlanCache(SearchValue)
-- 11
--CREATE UNIQUE INDEX UIX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey)
-- 12
--ALTER TABLE tbl_TestPlanCache ADD CONSTRAINT UQE_TestPlanCache_SearchKey UNIQUE (SearchKey)
-- 13
--CREATE CLUSTERED INDEX CLIX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey)
-- 14
--CREATE CLUSTERED INDEX CLIX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey)
--CREATE INDEX IX_TestPlanCache_SearchValue ON tbl_TestPlanCache(SearchValue)
-- 15
--CREATE CLUSTERED INDEX CLIX_TestPlanCache_SearchKey ON tbl_TestPlanCache(SearchKey)
--CREATE INDEX IX_TestPlanCache_SearchValue ON tbl_TestPlanCache(SearchValue) INCLUDE (ID)


Then I used following script to generate and capture generated plans:
(DO NOT TRY THIS ON YOUR PRODUCTION SERVER)



DBCC FREEPROCCACHE;
GO
SELECT * FROM dbo.tbl_TestPlanCache WHERE SearchValue = N'The select';
GO
SELECT * FROM dbo.tbl_TestPlanCache WHERE SearchKey = 126;
GO
;WITH XMLNAMESPACES ('http://schemas.microsoft.com/sqlserver/2004/07/showplan' as ns)
SELECT cp.objtype
    , RIGHT(st.TEXT, LEN(st.TEXT)-PATINDEX('%WHERE%',st.TEXT)-5) AS Query
    , CAST(qp.query_plan AS XML).value('max(//ns:StmtSimple/@StatementSubTreeCost)','float') AS StatementCost
    , CAST(qp.query_plan AS XML)
FROM sys.dm_exec_query_stats AS qs WITH (NOLOCK)
LEFT JOIN sys.dm_exec_cached_plans AS cp WITH (NOLOCK)
    ON qs.[plan_handle] = cp.[plan_handle]
CROSS APPLY sys.dm_exec_sql_text(qs.[plan_handle]) AS st
CROSS APPLY sys.dm_exec_text_query_plan(qs.[plan_handle], qs.statement_start_offset,qs.statement_end_offset) AS qp
WHERE st.TEXT like '%tbl_TestPlanCache%'
ORDER BY qs.creation_time OPTION (RECOMPILE);
GO

And here is the result of these test cases:

Based on the results I could make following assumptions for a single table query with single equality column in a WHERE clause:

SQL Server will Auto-Parameterize a query in following cases:
1. If Clustered Index Seek is performed.
2. If Clustered Index Scan is performed.
3. If Index Seek is performed for covering index (No RID or Key Lookup).
4. If Index Seek is performed for Unique Index or UniqueConstraint
5. If Table Scan is performed for a column, which included in any other index.
6. If Table Scan is performed with no indexes in a table at all.

SQL Server will NOT Auto-Parameterize a query in following cases:
1. If Index Seek is performed for not covering and not unique index resulting RID or Key Lookup.
2. If Table Scan is performed for a column, which not included in any index and other indexes exist.


As you can see, Auto-Parameterization in SQL Server, does not depend on query cost at all and it is very hard to tell if there are any benefits for it in some instances. So, I'd strongly recommend to explicitly parameterize/prepare all your production queries.

If you are really sure you want ALL your queries must be parameterized and you can't change the application code you can use "FORCE" parametrization setting for your database like this:

ALTER DATABASE <YourDatabase>
SET PARAMETERIZATION FORCED;



Thursday, September 1, 2016

SQL Saturday in Pittsburgh, PA. Sessions I want to visit.

I've started the series of SQL Saturday pre-reviews from SQL Saturday in Charlotte, NC last Saturday.
Now I want to continue with overview of SQL Saturday #545 in Pittsburgh, PA, which is sceduled exactly one month from now:

Event date: October 1st, 2016.
Session schedule: http://www.sqlsaturday.com/545/Sessions/Schedule.aspx

Event Home

Slot #1. 08:45 AM - 09:45 AM

That is probably the hardest slot to make the decision with.
There are two "performance" sessions: Chris Bell's "Would You Just Load Already?! Maximizing Your SSIS Data Load" and Jeff Moden's "The Secret to High Performance Code...The Devil’s in the Data".
There are two "High Availability" sessions: Adam Belebczuk's "Availability Groups 101: An Introduction" and Drew Furgiuele's "SQL Server Transactional Replication: A Deep Dive"
Even though my favorite so far is Jeremy Frye's "Speed Up Your SSAS Cube's Data Refresh With Dynamic Partition Processing" I might make my mind at the last moment or run between the sessions and capture the best of all.


Slot #2. 10:00 AM - 11:00 AM

That is the easiest slot for me. The session I'd like to visit in that slot would be Kevin Kline's "Configuration Option and Trace Flag Secrets: Surprising Behavior at the Flick of a Switch".
However, I'll miss that one because I must attend the best one: "What treasures are hidden in the Plan Cache?" which will be delivered by myself. Although, it will be my first time with that session, I hope that will be the turning point for all attendees, as it was for me few years ago when I attended similar topic delivered by Christina Leo. That topic should open your eyes on what the Plan Cache is, how it works, why we need it and what extra benefits we can get from it.

Slot #3. 11:15 AM - 12:15 PM

Despite of all other good sessions I want to visit Jonathan Stewart's "An introduction to the magical world of BIML!". I already know what it is and even tried it once, but I want to nail it down by asking questions and start using BIML from this point on daily basis.

Slot #4. 01:30 PM - 02:30 PM

In My previous post I've said that I will visit Rick Heiges' session "Winning with Columnstore - A Customer Story" in Pittsburgh, but I might review that decision and visit it in Charlotte. In Pittsburgh Rick will compete with Thomas Grohser's "Designing SQL Server HA/DR Infrastructure to meet the SLA", which I'm not going to miss that day.

Slot #5. 02:40 PM - 03:40 PM

That slot looks more relaxing. Different interests covered and I'll visit Robert Davis's "Exploiting SQL Server Security Holes".

Slot #6. 03:50 PM - 04:50 PM

The last slot and the less confusing one: Steve Hood's session "Buffer Pool Memory Management" is the winner in that slot for me.

So, if you are not registered yet for SQL Saturday in Pittsburgh, do it ASAP and dig into the schedule to find the topic you want.

Wednesday, August 31, 2016

Parsing "DBCC MEMORYSTATUS" without using PowerShell

When I wanted to research memory problem on a server and started to dig deeper into "DBCC MEMORYSTATUS" command.Very useful links to understand that command were from Microsoft:
https://support.microsoft.com/en-us/kb/271624
https://support.microsoft.com/en-us/kb/907877

During the research I've faced two problems:
1. I had to wait several seconds to get the full result set.
2. I had to scroll down and shuffle 115 different data sets to find the counter I want.

To eliminate both these problem all these different counters have to be in the same table/data set.
That will make research easier and data can be stored or compared with the base line.

So in order to achieve that we have to parse results of "DBCC MEMORYSTATUS".

I've did a quick search over the internet and found several parsing solutions, but all of them are based on PowerShell.
There is no problem with PowerShell for me, but the whole Idea seems wrong.
It is like if I'm asking for driving direction I can get the answer only in Chinese.

Being in frustration, I've decided to make it available in native SQL language.
So, here is the script:
SET NOCOUNT ON
GO
IF OBJECT_ID('tempdb..#tbl_MemoryStatusDump') IS NOT NULL
DROP TABLE #tbl_MemoryStatusDump;
GO
IF OBJECT_ID('tempdb..#tbl_MemoryStatus') IS NOT NULL
DROP TABLE #tbl_MemoryStatus;
GO
CREATE TABLE #tbl_MemoryStatusDump(
 ID INT IDENTITY(1,1) PRIMARY KEY
 , Dump VARCHAR(100));
GO
CREATE TABLE #tbl_MemoryStatus(
 ID INT IDENTITY(1,1), 
 [DataSet] VARCHAR(100), 
 [Measure] VARCHAR(20), 
 [Counter] VARCHAR(100), 
 [Value] MONEY);
GO
INSERT INTO #tbl_MemoryStatusDump(Dump)
EXEC ('xp_cmdshell ''sqlcmd -E -S localhost -Q "DBCC MEMORYSTATUS" ''');
GO
DECLARE @f BIT = 1
 , @i SMALLINT = 1
 , @m SMALLINT = (SELECT MAX(ID) FROM #tbl_MemoryStatusDump)
 , @CurSet VARCHAR(100)
 , @CurMeasure VARCHAR(20)
 , @Divider TINYINT
 , @CurCounter VARCHAR(100)
 , @CurValue VARCHAR(20);

WHILE @i < @m
BEGIN
 SELECT @Divider = PATINDEX('% %',REVERSE(RTRIM(Dump)))
  , @CurCounter = LEFT(Dump, LEN(Dump) - @Divider)
  , @CurValue = RIGHT(RTRIM(Dump), @Divider - 1)
 FROM #tbl_MemoryStatusDump WHERE ID = @i;

 IF @f = 1 
  SELECT @CurSet = @CurCounter, @CurMeasure = @CurValue, @f = 0 
  FROM #tbl_MemoryStatusDump WHERE ID = @i;
 ELSE IF LEFT(@CurCounter,1) = '(' SET @f = 1;
 ELSE IF @CurCounter != 'NULL' and LEFT(@CurCounter,1) != '-'
  INSERT INTO #tbl_MemoryStatus([DataSet], [Measure], [Counter], [Value])
  SELECT @CurSet, @CurMeasure, @CurCounter, CAST(@CurValue as MONEY)
  FROM #tbl_MemoryStatusDump WHERE ID = @i;
 SET @i += 1;
END
GO
SELECT * FROM #tbl_MemoryStatus
GO


Side notes:


1. Script uses "xp_cmdshell" command, which can be disabled on your server. In case you've got to the point when you are using "DBCC MEMORYSTATUS" command you have to have enough privileges to turn it on/off. Also, there are plenty of resources in the Internet on "how to enable xp_cmdshell".

2. Within "xp_cmdshell" command I use trusted connection "-E" and default local server: "-S localhost". These parameters can be changed adjusting for your instance/environment.

3. I've tested it on SQL Server 2014, but I'm pretty sure the script is supposed to work from 2008 to 2016. For SQL Server 2005 it would require some adjustments.

4. For the Value column I've used "MONEY" data type, which is limiting amount of "Available Virtual Memory" by 838 TB. There is possibility to use BIGINT, but "Memory Broker Clerk (Buffer Pool)" returns decimal values and that might be the problem. If you'd like you can use value data type as "DECIMAL(38,6)", but that I think would be the extreme case, which you might like to catch.

Saturday, August 27, 2016

SQL Saturday in Charlotte, NC. Sessions I want to visit.

Usually, I do my blog posts after visiting SQL Saturday events, but I want to change that tradition and talk about them before the actual event.

In this post I want to talk about sessions I have personal interest in and I understand that it might drive people from other sessions. However, it will definitely drive some audience in and I hope those people will thank me for an additional information/suggestions/guidance.

So, will start:
SQL Saturday #560, Charlotte, NC

Event date: September 17th, 2016.
Session schedule: http://www.sqlsaturday.com/560/Sessions/Schedule.aspx

Slot #1. 9:00 A.M. - 10:00 A.M.

In that slot there is ONLY ONE session I'm going to visit. That is my own session "Inside of Indexes". I'll be talking about:
- What indexes are;
- How SQL Server is using different types of indexes to retrieve data in the best possible way;
- What are Page Split and how they impact performance;
and some more.
I hope my session will be very interesting, but I'm afraid that half of the people, interested in indexing, will split between my session and Jeffrey Garbus' session "Indexing for performance".

If I didn't have my session in the very first slot I'd definitely go to Aaron Nelson's session "Achieve More in Less Time by Using the New SQL PowerShell".

Slot #2. 10:15 A.M. - 11:15 A.M.

In the second slot, the winner, without a doubt is Adam Machanic's "SQL Server 2016 "Hidden Gems". The room has to be big enough to accommodate all attendees. If you desperately want to visit his session, I'd suggest to come to the previous session of Angela Henry's "Lightning Talks" to secure your space.

Even though Adam Machanic will be the shining star in that slot I won't go for his session because I've already attended it on another event.

The second session in the priority is  Rick Heiges' "Winning with Columnstore - A Customer Story". Since Microsoft nailed down Columnstores in SQL Server 2016 and because of Rick's HIP (High Influential Power) - that would be the session to visit.

Despite on all advantages of Rick's presentation I won't go there. The trick is: I want to visit his session two weeks later on SQL Saturday #545 in in Pittsburgh.
So, I'll have a tough  choice between Rushabh Mehta's "Mastering and Enhancing Company Data" and William Assaf's "SQL Server Permissions and Security Principals".

Slot #3. 11:30 A.M. - 12:30 P.M.

The Best and "Must Go" in the third slot is Geoff Hiten's session "What’s your Imperative doing in my Declarative?". The name is not very descriptive, but that session can do a Inventor's Light bulb  effect on any beginner SQL Developer, who just graduated college with most of the experience in Java and C#. I guarantee, there will be the "AHA! moment" during that session, which will shift a paradigm of your programming experience for life.

So, because I already know all of that I won't go to that session.
The next in the line is Tom Norman's "Encrypting Data within Sql Server". I'm from DC region and "Security" here is always #1 priority.

Slot #4. 1:45 P.M. - 2:45 P.M.

After lunch session. There will be a lot of other performance tuning topics I like, but my choice will be Brian Carrig's session "Managing Very Large Databases with SQL Server".

Slot #5. 3:00 P.M. - 4:00 P.M.


Very good session is supposed to be of Kevin Feasel's "APPLY Yourself" about the art of using "Apply" operator.
Wonderful session of Jonathan Buller: "Tips and Tricks for the SSIS Catalog" about new abilities of managing SSIS packages directly from SQL. That  topic is not fully covered in SQL community yet and I'm waiting the time when we will discuss parallel executions of multiple queries using SSIS catalog and SSIS packages addressing their own scheduling manipulating SSISDB.

Because it will be the last slot of sessions and everybody, including me, will fill pretty tired, I'd probably will go an relax listening to "Replication v. AlwaysOn, the great debate", hosted by Chris Skorlinski and Curt Mathews. They encourage to bring your own questions to that discussion, so, I assume there might be several of very good blog posts delivered only based on that conversation.

See you in Charlotte!
And not hesitate to show your smile when I'll make your picture.