Dan's Blog : SQL Server 2008

Perfect Conference Storm

dtjones — Sat, 19 Sep 2009 17:24:24 GMT

November is shaping up to be the perfect storm of conferences. In a time when budgets are extremely tight it amazes me that we have 4 conferences to cover in November.

SQL PASS

When: November 2-5, 2009
Where: Seattle, Washington
Site: http://summit2009.sqlpass.org/
Sessions: 2 focused on SQL Server 2008 R2/Visual Studio 2010

SQL Connections

When: November 9-12, 2009
Where: Las Vegas, Nevada
Site: http://www.devconnections.com/shows/FALL2009SQL
Sessions: 3 – 1 on SQL Server 2008 R2/Visual Studio 2010 and 2 on SQL Server 2008

TechEd Europe

When: November 9-13, 2009
Where: Berlin, Germany
Site: http://www.microsoft.com/europe/teched/
Sessions: 3 – 1 on SQL Server 2008 R2/Visual Studio 2010 and 2 on SQL Server 2008

When: November 17-19, 2009
Where: Los Angeles, California
Site: http://microsoftpdc.com/
Sessions: 1 focused on SQL Server 2008 R2/Visual Studio 2010

We have a tremendous amount of great content @ each of these conferences.

It’s been a while since I’ve presented so I’m seriously considering taking the SQL Connections gig. After all I love Vegas and my brother from another mother and SQL Server MVP Steve Jones will be there. What more could I ask for.

PBM – it’s like liquid awesome!

dtjones — Sun, 30 Aug 2009 20:20:10 GMT

Brent Ozar and Thomas LaRock do a 3 part video series on the power of Policy-based Management (PBM). Interestingly SCOM comes up several times in the discussion.

They cover the basics of PBM and Thomas goes into more depth on how he’s using it in his environment. The series is about 35 minutes in total.

Regardless of your PBM experience level I highly recommend watching these videos.

http://thomaslarock.com/2009/08/policy-based-management-podcast-part-1/

http://thomaslarock.com/2009/08/policy-based-management-podcast-part-2/

http://thomaslarock.com/2009/08/policy-based-management-podcast-part-3/

PCI Compliance and SQL Server 2008

dtjones — Mon, 20 Apr 2009 03:22:08 GMT

Today compliance is a huge focus area for IT. Since compliance is written by lawyers for lawyers it takes a fair amount of translation to get it to the point where IT can understand what processes, policies, and configuration to implement to meet the requirements for a given compliance spec.

If you are a SQL Server DBA and face PCI compliance this white paper is a must read and the accompanying TechNet webcast a much watch.

Using PowerShell to Get Data & Log File Sizes

dtjones — Sun, 22 Mar 2009 00:42:06 GMT

I was playing around with PowerShell today (yes I’m was geeking out on Saturday afternoon) to learn how to access Performance Counters using PowerShell. My mission was to write a script that would return the Data File and Log File sizes for each database and a total for the instance. The good news is SQL Server already provides a bunch of fun performance counters and PowerShell has built-in support for accessing performance counters. I thought this was pretty cool so I decided to share it. In addition, I hadn’t blogged in a while and I was feeling bad about that.

There are two groups of scripts and sample output below: 1) Data File Space and 2) Log File Space.

Sorry about the formatting of the output I’m just being lazy today – after all it is Saturday.

Have fun with it!

Oh, btw: I was doing this on a Win7 machine (which I also need to give an update on) so I had to launch PS as administrator so it had access to the perf counters (same should be true on Vista and Windows Server 2K8). If you don’t launch as admin you’ll get an error that looks like this:

Get-Counter : The specified instance is not present.
At line:1 char:26
+ $DBDataFile = Get-Counter <<<< -Counter '\MSSQL$SQL2K8_01:Databases(*)\Data File(s) Size (KB)' -MaxSamples 1
+ CategoryInfo : InvalidResult: (:) [Get-Counter], Exception
+ FullyQualifiedErrorId : CounterApiError,Microsoft.PowerShell.Commands.GetCounterCommand

Data File Space

#Displays the total Data File size for each database (including system databases)

$DBDataFile = Get-Counter -Counter '\MSSQL$SQL2K8_01:Databases(*)\Data File(s) Size (KB)' -MaxSamples 1

$DBDataFile.CounterSamples | where-object -FilterScript {($_.InstanceName -ne "_total")} | sort-object -Property InstanceName | format-table @{Label = "Database"; Expression={$_.InstanceName}}, @{Label = "Total Data File(s) Size (MB)"; Express={$_.CookedValue/1000}} -AutoSize

Database Total Data File(s) Size (MB)

-------- ----------------------------

ansicheck 2.304

master 4.096

model 1.28

msdb 10.752

mssqlsystemresource 61.696

pbm-prototype 2.048

tempdb 8.192

We can also just grab the total Data File space used by the instance:

$DBDataFile = Get-Counter -Counter '\MSSQL$SQL2K8_01:Databases(*)\Data File(s) Size (KB)' -MaxSamples 1

$DBDataFile.CounterSamples | where-object -FilterScript {($_.InstanceName -eq "_total")} | sort-object -Property InstanceName | format-table @{Label = "Database"; Expression={$_.InstanceName}}, @{Label = "Total Data File(s) Size (MB)"; Express={$_.CookedValue/1000}} -AutoSize

Database Total Data File(s) Size (MB)

-------- ----------------------------

_total 90.368

Log File Space

#Displays the total Log File size for each database (including system databases)

$DBLogFile = Get-Counter -Counter '\MSSQL$SQL2K8_01:Databases(*)\Log File(s) Size (KB)' -MaxSamples 1

$DBLogFile.CounterSamples | where-object -FilterScript {($_.InstanceName -ne "_total")} | sort-object -Property InstanceName | format-table @{Label = "Database"; Expression={$_.InstanceName}}, @{Label = "Total Log File(s) Size (MB)"; Express={$_.CookedValue/1000}} -AutoSize

Database Total Log File(s) Size (MB)

-------- ---------------------------

ansicheck 0.496

master 1.016

model 0.504

msdb 0.504

mssqlsystemresource 0.504

pbm-prototype 1.016

tempdb 0.504

We can also just grab the total Log File space used by the instance:

$DBLogFile = Get-Counter -Counter '\MSSQL$SQL2K8_01:Databases(*)\Log File(s) Size (KB)' -MaxSamples 1

$DBLogFile.CounterSamples | where-object -FilterScript {($_.InstanceName -eq "_total")} | sort-object -Property InstanceName | format-table @{Label = "Database"; Expression={$_.InstanceName}}, @{Label = "Total Log File(s) Size (MB)"; Express={$_.CookedValue/1000}} –AutoSize

Database Total Log File(s) Size (MB)

-------- ---------------------------

_total 4.544

www.sqlserverenergy.com

dtjones — Sun, 26 Oct 2008 23:32:00 GMT

Check out the new SQL Server site!

Addressing Compliance with SQL Server 2008

dtjones — Sun, 26 Oct 2008 23:01:17 GMT

JC Cannon, Senior PM on the SQL Server team, presents this webcast where he explores some of the new features in SQL Server 2008 and describes how they can be used to help an organization address their compliance needs.

In the webcast JC covers:

Identity management
Data protection
Separation of duties
Auditing and reporting
Policy-based management

SQL Server 2008 Books Online Update

dtjones — Wed, 01 Oct 2008 02:48:13 GMT

The latest update to BOL is now available. The update is listed as August 2008 as that as the lock down for content changes. You can download it here.

SQL Server 2008 Experience

dtjones — Tue, 30 Sep 2008 18:16:03 GMT

I'm not super big on promoting marketing related stuff. But in this case I think there's a bunch of value for everyone. Our marketing team just launched a new site http://www.sqlserverexperience.com. Yes, it contains the normal marketing blah blah - sorry SQL Server Marketing team. But it also contains a bunch of video interviews with folks from the development team. These are cool and absolutely worth navigating the marketing-speaking to get at.

Refreshing the IntelliSense Cache

dtjones — Fri, 12 Sep 2008 06:17:17 GMT

There are cases where the local cache used by IntelliSense becomes stale. Refreshing the cache is easy but not necessarily obvious. There are two ways to refresh the cache: 1) Go to Edit -> IntelliSense -> Refresh Local Cache and 2) Hit Ctrl+Shift+R

SQL Server 2008 Webchat on Sept 23, 2008

dtjones — Thu, 11 Sep 2008 02:13:54 GMT

The SQL Server Product Group is holding a Webchat on Sept 23 from 10 am to 12 noon PST to get feedback from the community, customers, partners about their experience on SQL Server 2008 after over a month of RTM. You can find the details and participate through

Microsoft.com Community Calendar : http://www.microsoft.com/communities/chats/default.mspx

TechNet: http://technet.microsoft.com/en-us/cc719754.aspx

MSDN: http://msdn.microsoft.com/en-us/chats/default.aspx

We look forward to chatting with you.

PowerShell vs. T-SQL or Why Did We Add PowerShell Support in SQL2K8

dtjones — Fri, 29 Aug 2008 19:29:33 GMT

The following write-up was originally an email I sent to Allan White (SQL Server MVP) in the context of a discussion on PowerShell and some of the PowerShell community postings questioning why SQL Server needed PowerShell support @ all given the strength of T-SQL. Allan and others encouraged me to post the response. I realize there are times when the SQL Server product team appears opaque. In general this is not intentional. There are situations when we're working on very new stuff that cannot be discussed publicly (we function in a very competitive market), but generally we want to engage with the community and share the reasons we make certain decisions. I hope you find this posting informative and interesting. Note: I edited the original email just a little so that it made sense without the additional back and forth threads.

Within Microsoft we have a concept called Common Engineering Criteria (CEC). The goal of CEC is to establish consistency across IT products (Windows Server, Exchange, SQL Server, BizTalk, System Center, etc). Some of the requirements of CEC include items like virtualization support, security requirements, System Center support, localization requirements, and installation and patching requirements. I’m sure I don’t need to go into detail on why this is a good thing for our customers.

One of the latest CEC requirements is support for PowerShell. The old requirement used to be generic support for scripting. Imagine a world where every product used a different scripting syntax and had different constructs for error handling and procedural execution. Yuck! Just the cost to develop and maintain this would be enormous. There would be no consistency and each team would reinvent the wheel. This all might be fine if users never used more than one product. But we know that’s not the case. Lines of responsibility are blurring and IT professionals have to learn and use multiple products. Developers also must use multiple products.

While DBAs have traditionally been a specialized bunch and typically work solely with SQL Server we see this changing. For example, think of SharePoint. I had a number of people come up to me @ TechEd this year asking questions about SQL Server from a SharePoint perspective. These people have never written Transact-SQL and have no clue what a foreign key relationship is. They have no interest in learning T-SQL, rather they have a business need to keep the environment up and running – that’s what they’re measured on. If both products (SharePoint and SQL Server) implement a similar scripting language this person is instantly more comfortable – there’s automatic orientation. Yes there are new nouns and verbs to learn but there isn’t new syntax to learn. The fact that both products leverage the same scripting language places power in the user’s hands. They don’t have to “shell out” of one environment to make something happen in another.

Transact-SQL is a good scripting language but it’s not model driven. We will continue to invest in and evolve T-SQL. But we’ll also continue to invest in and evolve the models over DDL and DML. The model over DDL is SMO. The model over DML is EDM. We built PBM over SMO. This gave us instant backward compatibility with SQL2K5 and SQL2K. If we had to develop our own model over T-SQL we would have had to build the backward compatibility on our own. Frankly we wouldn’t have done this as the cost would’ve been too great. We also expose PBM in PowerShell which opens up use cases that wouldn’t be possible without extending the T-SQL language (an expensive proposition) or forcing DBAs to write C# or VB to accomplish seemingly simple tasks.

Take a database backup as an example. As part of the backup script I may want to find a volume which has greater than a minimum amount disk space and back-up to that drive. How do I do this in T-SQL? Or take a script that needs to be aware of the OS or processor the instance is running on. How do I get that from T-SQL? I’ve had a number of requests for help building policies which we just can’t do because either we don’t have a facet to support it or T-SQL can’t get access to the information. But if PBM had PowerShell support we’d be golden. As an aside, I really wish we added PowerShell support to PBM in the form of an ExecutePowerShellScript() function. We didn’t because of time constraints. Yes we can always add this stuff to T-SQL but we’ll be in a perpetual state of catch up. That’s a frustrating place for us and our customers.

Both PowerShell and our implementation on PowerShell are v1 products. They will evolve and get better. SMO is a v2 product and it’s evolving and getting better. In SQL11 we’ll make investments in T-SQL. Our goal in SQL Server manageability is to get out of the business of delivery one off tools and move to a state where everything is model driven. This creates an environment that is far more consistent and predictable. It allows us to build capability once and expose it in many places. For example, we want to get to the point where all we have to do is add a new object to SMO and it’s automatically available in PowerShell, SSMS, PBM, System Center, etc without making changes to those environments.

APIs aren’t always user friendly. The implementation may be pure from a coding and relationship perspective but from a usability perspective it may be hard to work with. Take a table object in SMO. It takes a lot of “work” to construct a new table in SMO. It’s far easier in T-SQL. But if I don’t know T-SQL and I know PowerShell wouldn’t a create-table cmdlet be valuable to me? The same goes for backing up a database. This is where the power of models and abstractions are realized. After all, when was the last time you wrote assembly code?

Where the Heck is DMO?

dtjones — Fri, 29 Aug 2008 01:20:04 GMT

You've installed SQL Server 2008 on a clean machine and your vb script that uses DMO doesn't work. You're asking yourself: What the heck happened? Did those goofballs over @ MS hose me? Deprecation of DMO was announced in SQL Server 2005, but SQL Server 2005 still installed DMO. In SQL Server 2008 we've taken the next step down the deprecation path - we ship DMO but don't automatically install it. There are two places to get DMO:

SQL Server 2008 Installation Media: go to ..\x86\Setup\x86\SQLServer2005_BC.msi
SQL Server 2008 Feature Pack @ the Microsoft Download Center: http://www.microsoft.com/downloads/details.aspx?FamilyId=C6C3E9EF-BA29-4A43-8D69-A2BED18FE73C&displaylang=en

Generally it's best to go to http://www.microsoft.com/downloads, search for "SQL Server Feature Pack" and sort by Release Date to find the latest version.

With the addition of PowerShell support in SQL Server 2008 we highly encourage you to move off of DMO and start using SMO.

Inside The SQL Server 2008 Shiproom on RTM Day

dtjones — Mon, 25 Aug 2008 08:30:34 GMT

The Edge folks (namely Joey Snow) sat in on the SQL Server 2008 shiproom on the day we RTM'd. You can view the video here. A few interesting things to share.

During a release shiproom happens every day. Each team across SQL sends at least one representative. Depending on where we are in the release cycle the agenda and mood of shiproom is different. For example, in the end game we're in what we call "Ask Mode". This simply means that each and every fix that is checked in must get approval. That's right, every single fix. This probably sounds like a lot of overhead, but that's somewhat the point. SQL's a big product and it takes time to slow down the ship. Every code change carries risk. The purpose of Ask Mode is to slow down code churn and ensure we apply a consistent triage bar.

The final day, sign off day, is very fun and festive. I was there the day we signed off on Yukon. The champagne was flowing and people were hugging and shouting. It was a lot of fun. For Katmai I was in a conference room one floor up preparing for the press conference call publicly announcing SQL2K8 RTM. The shouts of joy coming from shiproom were deafening.

The sign off paper is ceremonial and doesn't carry any operational significance. Joe Wilcox thought it was funny that people were signing a piece of paper. The actual sign off on the release is digital and doesn't require each team to sign off. For SQL2K5 we simply went around the room and each team said "<Team Name> is a go!" It doesn't matter if it's a verbal sign off or a written signoff, it's fun and cool to do it.

SQL2K8 Express with Advanced Services and Express with Tools Now Available

dtjones — Sat, 23 Aug 2008 06:53:13 GMT

Two Express editions came on-line today. Express with Advanced Services and Express with Tools. You can download them from the Microsoft Download site. Here are the links:

SQL Server 2008 Express with Advanced Services (ENU): http://www.microsoft.com/downloads/details.aspx?FamilyId=B5D1B8C3-FDA5-4508-B0D0-1311D670E336&displaylang=en
SQL Server 2008 Express with Tools (ENU): http://www.microsoft.com/downloads/details.aspx?FamilyId=7522A683-4CB2-454E-B908-E805E9BD4E28&displaylang=en

Automating DBA Processes

dtjones — Mon, 18 Aug 2008 02:37:12 GMT

This year @ TechEd I had the privilege of sharing the stage with Kevin Kline, Steve Jones, Buck Woody, and Allen White to discuss automating DBA tasks/processes. We discussed why automation is important and some of the technologies you can use to automate those tasks. Click here to launch the video. Or go to the TechEd Online Library where you will find it in WMV (Low & Hi), MP3 and MP4 formats.