- Ralph Waldo Emerson

After witnessing 3 SQL Server releases (Katmai, KJ and Denali “which is yet to be RTMed”) and resolving hundreds of SQL Server issue, I’ve now decide to move ahead to a new career chapter.

Yes it been a VERY tough decision for me since here @ MS, I just love my day-to-day work; our software products, our service commitments and most important, the IMPACT that we deliver to global customers. I’m part of a most knowledgeable and smartest set of colleagues and I have learned so much from them. However now having several years spent here, I've decided it is time for me to try something new – so my last day.

While I am leaving Microsoft, I continue to have a very positive view of the “my” company. No other company has the footprint that Microsoft does on the IT industry and as a company we’ll continue to “change the world” with our products. Needless to say that I have a great time @ MS and learned immensely from the smartest and most passionate set of people. Of course, there is never a good time to leave, but I feel a great sense of accomplishment supporting SQL 2000, SQL 2005, SQL 2008 / R2 and Denali….(yet to be RTM)

As I have been thinking about this transition, here is what I'm planning for the near future on my new blog:

• Get into shoes of a technical architect
• Explore the powers of “Power-Shell”
• Brush-up my skills on other RDBMS (beyond MS-SQL)
• While this MSDN blog will be frozen, I’ll maintain a new blog on wordpress that encompasses content on this blog as well as my new blog.

I want to say sincere thanks to all the folks with whom I've worked, I've learned so much from you guyz. Microsoft is filled with brilliant, hardworking people, and it is an honour to have worked with you. While here, I have had the chance to work “fastest growing database – SQL Server”.

Also sincere thanks to blog readers and fellow blogger; you have been a big inspiration throughout. I've started a new blog on new domain, http://sqltune.wordpress.com/. For uninterrupted blathering, please update your RSS readers to use the new address and RSS feed. My blog URL will be changing after this post. All the posts you find here will remain here and will still be search-able.

Current blog URL: http://blogs.msdn.com/b/varund/
New blog URL from this day forward: http://sqltune.wordpress.com/

You guys rock! And we’ll cross lines again.

- Varun

SQL Server 2005 Service Pack 4 (latest SP in SQL 2005) was publically released on 18-Dec-2010. This time the word RTM suffixed to this patch name.

This is 1st time that SQL Release team has added “RTM” to a SQL Service Pack release and to a degree this is created some confusion……..what is means RTM or No RTM? Let me attempt to clarify this.

Before RTM there’s pre-release called CTP – SQL Server 2005 Service Pack 4 CTP 

Basically,  per the process SQL teams currently are first releasing a CTP (Community Technology Preview) of Service Packs with an objective to release a patch with highest quality.  Refer below excerpt from Microsoft SQL Server Driver for PHP Team Blog

• a deeper engagement with the right partners earlier in the planning and stabilization phases to get the requirements, design, and quality right
• grouping the features into multiple "sets" that have a correct logical sequence as well as balance the effort across all the component teams to finish the "set" around the same time
• each "set" is called a Community Technology Preview (some use Customer Technology Preview)
• the release criteria for the CTP are the same as final release (some corner case bugs may be deferred to a later CTP)

Next comes RTM – SQL Server 2005 Service Pack 4 RTM

Once the new patch has reached as stage to be made generally available to public, it termed as RTM (Release To Manufacturing). So when SQL Server 2005 SP4 CTP time and objectives are completed the final product released is called RTM.

Hope this will dispel the confusion around….SQL Server 2005 SP4 RTM Release………why RTM

Cheers!!

Disclaimer: I work at Microsoft. Everything here, though, is my personal opinion and is not read or approved by Microsoft before it is posted. No warranties or other guarantees will be offered as to the quality of the opinions or anything else offered here.

It's has been couple of years for me working on SQL Server and I keep coming across various SQL Server misconceptions (or say a DBA Myth). In this specific post I’ll demystify a common DBA myth around log-shipping: "In a log shipping setup, to have smaller size T-log log backups being shipped to secondary, switch to "BULK_LOGGED" recovery model before performing any bulk operation"

SCENERIO:
- Customer have configured Log Shipping for a high OLTP database
- On Primary Server, to optimize the database performance, a daily RE-INDEXING job is scheduled
- Everyday, after the RE-INDEXING job is executed, T-log grows huge and size consecutive T-Log backup is large. It takes lot of time (Of course the Network Resources!!) to ship T-log backups to secondary server
- So, to resolve this issue, DBA came with below plan:

1. Change the recovery model of database to BULK_LOGGED
2. Perform RE-INDEXING (which qualifies to be minimally logged operation)
3. Change the recovery model back to FULL

- Customer believes, following these steps will reduce the size of T-log and thereby minimize network resource usage. This is "THE MYTH"

Read the complete latest POST on SQLServerFAQ and demystify this one

About SQLServerFAQ

SQLServerFAQ blog contains a plethora of information spread across various aspects of troubleshooting commonly encountered issues with SQL server. The posts are contributed by SQL PSS (Product Support) Engineering community.

You may have seen the announcement of release SQL Server Code-Named “Denali” in recent PASS summit (featured in Ted Kummert’s keynote). Even morem “Denali” Community Technology Preview 1 (CTP1) is available for download here 

“Denali” release have several new and improved components listed here. Being a SQL Engine guy, I particularly got excited with HADR or AlwaysON which is an Availability Enhancements. So here what I did…downloaded the CTP, Configured a 2 node Windows 2008R2 Failover Cluster on Hyper-V machines and then Installed the Denali Instances on Virtual machines.

Next, I configured HADR Availability Groups and did some play with failover.  I have blogged my experience in on team blogging site SQLServerFAQ as below

SQL Server Denali - AlwaysON (HADR): Step-by-Setup setup guide

About SQLServerFAQ

SQLServerFAQ blog contains a plethora of information spread across various aspects of troubleshooting commonly encountered issues with SQL server. The posts are contributed by SQL PSS (Product Support) Engineering community

Now, since it’s a new year is arriving, I want to share my new year resolution with you

“When ever in issue, I will ALWAYS read the documentation.….. just as soon as I can find it”

 Merry Christmas and a “Blessed” New Year

To begin with, lets first get the basics:

What is Atlanta?

Microsoft Codename Atlanta (http://beta.microsoftatlanta.com) is an online service that analyzes installations of Microsoft SQL Server 2008 (and later versions) and provides proactive alerts to help you avoid system downtime and follow best practices with regard to configuration and usage.

Do I (you) need Atlanta?

Today, database administrators (DBAs) have to spend a great deal of time and effort dealing with problems related to misconfiguration of Microsoft SQL Server. Microsoft Codename “Atlanta” is a configuration-monitoring cloud service that will help customers to reduce downtime and improve the performance of Microsoft SQL server deployments. This secure cloud service will help DBAs to:

• Proactively avoid problems with Microsoft SQL Server through ongoing monitoring and alerting of configuration from the cloud

• Reduce downtime and improve performance of Microsoft SQL Server through proactive scanning for known configuration issues and comparison with best practices

• Resolve issues faster by providing Microsoft, or your own support staff, the ability to get a current view of your configuration, changes that have been made, and use the most up to date solutions to solve the

Examples of the types of situations that Microsoft Codename Atlanta will alert customers on include:

• Degraded (Ex. Configuration/operational state matches a known “issue signature”)
• Misconfigured (Ex. Service waits for connections on a firewall-blocked port)
• Underpowered (Ex. HW configuration does not meet minimum/recommended specs)
• At risk of data loss (Ex. Event 825: Failed IO Request detected.)
• Unpatched (Ex. “Failure to Install Hotfix KB 962975 on Servers Using Dynamic Disks Could Result in Significant Data Loss in Certain Scenarios”)

The Architecture?

Is the RTM available?

No  , currently Atlanta CTP is available. So  to get your hands on the beta software, sign up immediately on https://www.microsoftatlanta.com/

I have signed up already, now how to Install/Configure/….and everything else?

Step 1. First Review the Atlanta System Requirements.

Step 2. Installation - download the security certificate

Note: If you get below error, while trying to download the registration certificate, this could due to fact that your machine is unable to locate a DC and user cannot sign in.

Solution: Try downloading security certificate from a different machine and then copy them back to the identified GATEWAY server

Step 3. Download Atlanta Gateway and Agent Setup

This involves you to first identify a machine as Gateway (a system connected to Internet) and then other servers running SQL Server 2008/R2 instances, where the Atlanta Agents has to be installed. Download Atlanta setup files to a target folder on Gateway server

Note: So if Gateway and Agent are both on same server, choose the appropriate options

Step 4. Give the path for Atlanta security certificate (as downloaded in Step 2.)

Step 5. Ready to proceed…

Step 6. Once Installation is completed successfully, you will be presented with Atlanta Home Page

Lets take a min here to understand now some terms frequently used here:

Alerts  View alerts to see any issues that have been detected.

Servers View and manage the gateways and agents deployed in your environment. You can view information about the last time the gateway uploaded data to the service or when the agent uploaded data to the gateway. You can also change which agents and gateways can communicate with the service.

Account View and change the details associated with your account. You can also close your account from this page

Configuration: Current Snapshot View current configuration information for the computers being analyzed.

Configuration: Change History View historical configuration information for the computers being analyzed

Step 7. By this time, Both Agent and Gateway are installed and running under default configuration, which means it will take some time for web sync to happen. Once this will happen you will be able to see your SQL Server Instances on the Atlanta console

By end of this step, you are now ready to drill down into the Alerts content i.e. ERROR and WARNINGS

Here’s what I got first on my server

Error # 1

A very useful fix related to queries with LIKE operator…..recommended

So you see, Atlanta is already into the action, suggesting you what not as per the best-practices in your SQL Server Inventory!!

Only thing in my wish list is ability to COPY the text from Alerts, which is not available in CTP.

Atlanta appears to be useful DBA tool and I recommend you too test this today!

Disclaimer: I work at Microsoft. Everything here, though, is my personal opinion and is not read or approved by Microsoft before it is posted. No warranties or other guarantees will be offered as to the quality of the opinions or anything else offered here. 

Hello All,

Sometime back, I have written blog post on ‘Backing-up SQL Server Database on a network shared drive’. This can be found here. I have received numerous queries on an another related topic i.e. ‘How to Create a SQL Server Database on a network shared drive’. It took sometime (a brand new SQL Release) before I could get the answer as follow:

1. Create a database on Network Share in SQL Server 2008 / SQL Server 2005

In SQL 2008 and SQL 2005, By default, you cannot create a new database on a network share drive.

This restriction is primarily due to fact that, On a network file share, there is always a risk on network errors compromising database integrity, along with I/O performance  issues which might partial or total data loss or corruption. Reference >> Microsoft KB # 304261

However, there’s a “workaround”, if in case you still want to go ahead. Follow below steps:

Step 1. Enable the Trace Flag 1807: (Bypasses the check and allows you to configure SQL Server with network-based database files)

DBCC TRACEON(1807, -1)

Step 2. Identify a file share, where SQL Server Service start-up account has FULL access

Step 3. Create the database

CREATE DATABASE [networked] ON  PRIMARY
( NAME = N'networked', FILENAME = N'\\varund-win7\ATOMNETWORKDB\networked.mdf' , SIZE = 3072KB , MAXSIZE = UNLIMITED, FILEGROWTH = 1024KB )
LOG ON
( NAME = N'networked_log', FILENAME = N'\\varund-win7\ATOMNETWORKDB\networked_log.ldf' , SIZE = 1024KB , MAXSIZE = 2048GB , FILEGROWTH = 10%)
GO

2. Create a database on Network Share in SQL Server 2008R2

Starting SQL Server 2008R2, you are now allowed to cerate database on network file share (UNC path), without the need to Trace 1807

Step 1. Identify a file share, where SQL Server 2008R2 Service start-up account has FULL access

Step 2. Create the database

CREATE DATABASE [networked_r2] ON  PRIMARY
( NAME = N'networked_r2', FILENAME = N'\\varund-win7\ATOMNETWORKDB\networked_r2.mdf' , SIZE = 3072KB , MAXSIZE = UNLIMITED, FILEGROWTH = 1024KB )
LOG ON
( NAME = N'networked_r2_log', FILENAME = N'\\varund-win7\ATOMNETWORKDB\networked_log_r2.ldf' , SIZE = 1024KB , MAXSIZE = 2048GB , FILEGROWTH = 10%)
GO

While the above options will surely help you to manage space used by database file across your networked servers, please be aware of the risk involved (as I discussed above). Once such known issue has been discussed in this CSS post. It is therefore advised to run SQLIOSIM.exe that ships with SQL Server 2008 R2, this has now been updated to allow testing against a UNC locations.

Microsoft recommends that you store database files either on Local Disk or on Storage Area Networks (SANs).

Enjoy and Thanks of reading!

Disclaimer: I work at Microsoft. Everything here, though, is my personal opinion and is not read or approved by Microsoft before it is posted. No warranties or other guarantees will be offered as to the quality of the opinions or anything else offered here.

Hi Guys

In a recent issue with a customer, I came across a strange SQL belief (or say a DBA Myth) around SQL Server Transaction Logs Backup/Restore. Let me explain you what was the scenario and then let’s try attempting to break the myth with real testing.

SCENERIO:
-    Customer is taking FULL and TRANSACTION LOG (T-Log) backup of a critical database
-    Backup files are maintained on disk in sequential order
-    Due to a hardware failure (let’s assume), the database got crashed and customer decided to recover from the backups
-    Due to an another issue (assuming again!!!) the  most recent FULL backup file is LOST!. So customer is now only left with 1st FULL backup and successive T-Log backups
-    Customer is now confident that he will not be able to perform point-in-time recovery. “THIS is the MYTH”

Read the complete latest POST on SQLServerFAQ and demystify this one

About SQLServerFAQ

SQLServerFAQ blog contains a plethora of information spread across various aspects of troubleshooting commonly encountered issues with SQL server. The posts are contributed by SQL PSS (Product Support) Engineering community.

He is trying to install SQL Server 2008 Enterprise on a Windows 2008 x64 box. As soon as we click ‘setup.exe’, we get below pop-up.

When we check from Control Panel –> Add/Remove Programs, I can clearly see Microsoft .NET Framework 3.5 SP1 installed.

Then why this error?

Started with the few basic check of looking into Setup Bootstrap logs, but there were none!!! means setup error has not even logged.

The came an interesting revelation: Customer informed that he was able to successfully install SQL Server 2008 32 bit, on different box, using same installer.

This gave me a clue to take a closer look at the SQL Installer. I started looking into x64 folder, comparing them to our own SQL installer library. And, my suspicions were validated as I could find several file/folders missing within x64 folder.

Very next thing I did was to list the missing folders, rename then at my end and then try simulating the error (In Microsoft Support, we love to re-produce the issue and find a solution for it) and I could get the same error, whoa! So this is a case of SQL media corruption.

Resolution then become simple as we just replaced the corrupt media with fresh one and then customer was able to install SQL Server 2008 successfully.

Thanks for reading!

Disclaimer: I work at Microsoft. Everything here, though, is my personal opinion and is not read or approved by Microsoft before it is posted. No warranties or other guarantees will be offered as to the quality of the opinions or anything else offered here.

SQL Server performance counters missing!!

Past few hours been crazy one, breaking my head on a ‘Missing Performance Counters’ and now since this is resolved, I want to now off-load every bit of this issue from my brain so thought blogging it :-)

Issue: A customer is having 24x7 PERFMON counter collection and suddenly they found that all SQL Server related perform counters are missing.

Cause: While working on this for few hours, we could determine that a known issue described in Microsoft KB # 812915. The issue is specific to an ‘ancient’ SQL built (SQL2K SP3), wherein if SQL Server Service is shutdown (re-started) in between a perform collection then, after re-start, SQL Server related PERFMON counters do not appear in Performance object list.

Resolution:

Quick one – Ensure to always close the Performance console or Stop collection before you restart SQL Server.

Long (and better) one – Consider applying SQL 2000 SP4

However the story goes beyond. While working on this issue, I hit across so many distinct issues due to which SQL Perfmon counter goes missing and how to fix/retrieve them. I have consolidated a list of those articles / blogs where these distinct issues have been documented

SQL Server 6.5 / 2000

• FIX: "Performance monitor shared memory setup failed: -1" error message when you start SQL Server
• PRB: SQL Performance Monitor Counters Missing
• PRB: SQL performance counters may be missing after MDAC installation on a cluster

SQL Server 2005 / 2008

• How It Works: Almost Everything You Wanted To Know About The SQL Server (2005, 2008) Performance Counter Collection Components
• What should we do when the sql server performance counters are missing
• How to rebuild the performance counters of a SQL Server 2005 or 2008 instance

Operating System

• How to manually rebuild Performance Counter Library values
• Missing Objects and Counters in Performance Monitor

Expect to find changes to this blog post, as I might come across more issue on ‘missing performance counter’, but as of today, so-far-so-good

Cheers!!

For more details on Slipstreaming and it’s benefits, I recommend reading Slipstream FAQ POST by PeterSad.

Over here, we will first create a merged (slipstream) drop containing SQL Server 2008 (RTM), SQL Server 2008 Service Pack1 (SP1) and SQL Server 2008 Service Pack1 Cumulative Update 4 (CU4)

So let’s get going!

The steps procedure describes how to create a new source media that will slipstream the below:

a. SQL Server 2008 original source media (RTM)

b. SQL Server 2008 Service Pack 1

c. SQL Server 2008 Service Pack 1 Cumulative Update 4.

Once this merged (slipstream) drop is created, we can install SQL Server 2008 SP1+CU4 in a single step.

IMP Note: These steps are for the SQL Server 2008x64. The steps will remain same of x84 and IA64 if you obtain the correct of service pack package.

Steps:

1. Copy the original SQL Server 2008 RTM source media to D:\SQL2008_Slipstrm_SP1CU4\

2. Download the Service Pack 1 package. The package names are as follows:

• SQLServer2008SP1-KB968369-x64-ENU.exe

3. Download the Service Pack 1 Cumulative Update 4 package. The package names are as follows:

• SQLServer2008-KB973602-x64.exe

At this point, the D:\SQL2008_Slipstrm_SP1CU4\ folder should look like this:

4. Extract the SP1 packages as follows:

• SQLServer2008SP1-KB968369-x64-ENU.exe /x: D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\PCU

5. Extract the CU4 packages as follows:

• SQLServer2008-KB973602-x64.exe /x: D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\CU

At this point, the D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM folder should look like this:

6. Run the following commands to copy the Setup.exe file and the Setup.rll file from the extracted location to the original source media location. I’m using ROBOCOPY** to achieve this.

robocopy D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\CU D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM Setup.exe

robocopy D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\CU D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM Setup.rll

Note: On the Copy Prompt, choose “Copy and Replace”

7. Run the following commands to copy all files (not the folders), except the Microsoft.SQL.Chainer.PackageData.dll file, From "D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\PCU\x64" To "D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\x64" to update the original files.

robocopy D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\PCU\x64 D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\x64 /XF Microsoft.SQL.Chainer.PackageData.dll

Note: On the Copy Prompt, choose “Copy and Replace”

Note If you accidentally copy the Microsoft.SQL.Chainer.PackageData.dll file, you may receive the following error message when you run the Setup.exe file.

SQL Server Setup has encountered the following error:
The specified action LandingPage is not supported for the SQL Server patching operation. Error code 0x84BF0007

If this issue occurs, restore the Microsoft.SQL.Chainer.PackageData.dll file back to the original version.

8. Determine if you have the Defaultsetup.ini file in the following folders:

• D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM\x64

If you have the Defaultsetup.ini file in the folders, open the Defaultsetup.ini file, and then add additional parameter in blue to the file as follows:

;SQLSERVER2008 Configuration File

[SQLSERVER2008]

PID="XXXXX-XXXXX-XXXXX-XXXXX-XXXXX"

PCUSOURCE=".\PCU"

CUSOURCE=".\CU"

9. Start the Setup program from below:

• D:\SQL2008_Slipstrm_SP1CU4\SQL2008DeveloperRTM

10. In the Ready to Install screen, you will notice below:

Screen # 1

Screen # 2 – on same screen going further down

11. You should see the PCUSource and CUSource being specified in Summary.log

12. Once the setup is completed successfully, Launch the SQL Server Management Studio (SSMS) and verify the SQL Server product Build:

13. The steps have been derived from Microsoft Knowledge Base # 955392

Happy Slipstreaming!

ROBOCOPY** - Robocopy, or "Robust File Copy", is a command-line directory replication command. It has been available as part of the Windows Resource Kit starting with Windows NT 4.0, and was introduced as a standard feature of Windows Vista, Windows 7 and Windows Server 2008. For Windows 2003, this comes included with “Windows Server 2003 Resource Kit Tools”

Disclaimer: I work at Microsoft. Everything here, though, is my personal opinion and is not read or approved by Microsoft before it is posted. No warranties or other guarantees will be offered as to the quality of the opinions or anything else offered here.