I have been off the blogs for a while and recently changed roles in Microsoft to be an Industry Technical Strategist in our Worldwide Financial Services practice. I guess I picked a great time to move into our Financial Services practice. :)

One of my passions in my previous roles was around SQL Server and specifically SQL Server in Line of business applications.

Let’s give a little history 

Previous to working at Microsoft, I worked with SQL Server, Sybase and Oracle back in the SQL 4.2 and 6.0 days. Anyway, back then SQL was a database that fit in between Jet(Microsoft Access and Foxpro) and Oracle. It was mainly focused at department level internal applications while Oracle and Sybase were focused at larger workloads. SQL Server was typically easier to use, manage and install than its counterparts. 

Starting with SQL 7.0, we focused on larger workloads and starting competing in TPC and other benchmarks and expanded this through SQL 2000. We started catching the competition and still had the ease of use and deployment.

Then in SQL 2005 and 2008, we really expanded and are able to handle pretty much any large business workload and the largest most scalable applications. We added reporting, OLAP, and enhanced integration capabilities and moved into data warehousing in a big way. At the same time, we still kept the ease of use, manageability, deployment and installation.

A funny thing happened along the way. We learned some big lessons along the way and added features to handle a world where servers need to be online 24X7. Features such as Replication, Clustering, mirroring were added to address availability. Table partitioning and larger 64-bit servers were added to handle the normal mainframe workloads.

Today

In the past few years, almost every customer I talk to still has a leftover view of whether or not SQL Server can handle their LOB solutions. I think this is still a leftover view of SQL's early days where it wasn't viewed as a mission critical database. So as my first post in a couple of years, I wanted to start adressing these questions in Financial Services

SQL Server in Mission Critical Workloads

Here are a few proof points you may want to view when looking at SQL Server in Financial Services.

What about the really high throughput workloads like Stock quotes and lookups. Can SQL handle thousands of updates per second?

Townsend Analytics runs on SQL 2005 and handles 60,000 quotes per second inserted into a DB and 100 events per second.

http://download.microsoft.com/download/d/d/7/dd79c390-5b75-45fb-a0f5-2359a6d20643/GEMS4641/Townsend_Analytics_SQL_Server_2005_Launch%20Case_Study.doc

Citi handles 200,000updates per second

 http://www.microsoft.com/casestudies/casestudy.aspx?casestudyid=4000001199

My point is that Microsoft SQL Server is the backend of exchanges, ticker plants, and Market research data.

What about Line of business applications with thousands of concurrent users?

Yes, we handle those as well.  How about Progressive who moved their Policy Admin System to SQL 2005 with ~27,000 concurrent users? (http://www.microsoft.com/casestudies/casestudy.aspx?casestudyid=4000002133)

My point here is that we work handle the workloads of things like teller solutions in banking, Insurance (claims and policy management), and capital markets (Trading stations) we handle workloads of thousands of concurrent users who are always logged on.

 What about online sites? Online Banking? Online investments? Insurance quotes?

How about Marketwatch?

http://www.microsoft.com/casestudies/casestudy.aspx?casestudyid=200969

We handle many of the online banking portals, securities portals and Insurance portals. Anyway, we have too many case studies to list, but today we handle the largest workloads with thousands of updates per second and thousands of concurrent users.

What’s next?

My next update will talk about reporting and data warehousing and finish off with how do you build solutions with SQL Server that are scalable and highly available.

I have been away from blogging for a while.  A little too much travel to keep up with multiple blog sites. In the past 12 months, I have been to Thailand, Japan, Seattle, San Fran, LA, New York 15 time, Orlando, Dallas, Tampa and various other cities. Basically around 120,000 airline miles.

 This week I am at the PDC, and I am really excited about the changes coming.

I will come back to my BizTalk blogs next week, but I wanted to talk about 4 great technologies that we are releasing soon or have already released.

Do you know the message you get when a program crashes and it asks you about sending information to Microsoft?

Do you know that when you clik yes it actually sends this information to a server that is used to be a gathering point for debugging crashes.

 
The server is actually winqual.
 
https://winqual.microsoft.com/

Did you know that as a developer, you can subscribe to this server and register your applications?

 
If you do, and if your application crashes and the end user chooses to send the information to Microsoft, you can go to this site and download the crash and debug it. It will also show you a list of crashes and provide you with the ability to actually send a message back with a link to a hotfix for your application after you fix it.

 
The LINQ project

The Language Integrated Query framework (the LINQ Project) is a set of language extensions to C# and VB and a unified programming model that extends the .NET Framework to offer integrated querying for objects, databases and XML.

LINQ’s query operators can be applied to any .NET array or collection of objects, effectively providing SQL- and XQuery-like capabilities for in-memory data. Essentially query-capabilities become a first class citizen in .NET programming languages.  In addition, LINQ provides two critical APIs code-named:

·         DLinq: for SQL relational data access
·         XLinq: for XML hierarchical data access

Windows Workflow Foundation or Workflow in general

This is bringing BizTalk like Workflow to the developer community without having to have a copy of BizTalk Server. This is like a developer version of BizTalk without requiring XML documents. There is actual c# or VB.NET code integrated. Events, code breaks, etc.. I would highly recommend checking this out at the site below.

http://www.msdn.microsoft.com/windowsvista/building/workflow/
 
There is also a series of hands on labs and documentation.
 

AJAX or Project Atlas

Yet another new tool to get excited about. This is a set of JavaScript libraries to build client side code that will work in conjunction with ASP.NET. Basically it gives a Client-side Jscript designer so that people don't have to build the libraries to do stuff like populate a table with data from a web services.

The bits are on ASP.NET. There is a whitepaper and hands on labs as well.

http://beta.asp.net/default.aspx?tabindex=7&tabid=47

I have been away for a week in New York again and just got back in the office and cleaned up a little of my emails. In parsing them, I wanted to pass along some information that on Side by Side assembly deployments with BizTalk and how to get this to work.

If you create another version of an already deployed Biztalk assmbly and you would like to start the new version, here is what you need to do to get this to work.

1. Configure any pipeline components to use a specific version of the schema. 
2. If the schema is redeployed under a different version and this is not done it will fail because it looks up by target name space / root element not by assembly.
3. Deploy the new version
4. Bind the orchestrations. 
5. The Manual steps are below
   1. Disable the receive location
   2. Unenlist the old version using the Explorer (do not stop it), this will prevent new instances from starting  while allowing old ones to finish
   3. Enlist the new version and start it, all new instances will be created using this one.
   4. Enable the receive location

A demo I sometimes show during various BizTalk presentations is one that shows how to compensate for problems with using long running transactions in BizTalk Server. This sample was originally developed by Erik Leasburgh and  Raman Mahal. I created the walkthrough so we can show how to create the sample and hopefully provide details to show how it works so others can use it as a template for their designs. The problem we are trying to solve is that the atomic transaction shape in orchestration doesn’t really provide true atomic transactional semantics when used with a transactional adapter like SQL, FTP etc.

Let’s say I have an orchestration which has a send port which is bound to SQL Adapter Send port for sending the message to the SQL Database. For transactional purposes the developer has enclosed this send shape in an atomic transaction. Also in the same transaction there’s another send shape which is sending messages to another system. Both these sends should commit totally or should rollback.

Now during runtime say Send1 to SQL Server succeeds, but Send2 to the other system fails and because of that Send1 has to be rolled back. A user would think that since the 2 sends are in atomic scope in the orchestration, the orchestration will take care of the rollback process. This is not the case, you will need to compensate for this in your Orchestration. The sample and walkthrough shows you how to do this.  This is a version 1.0 of my docs, so let me know if this works or if you see any problems.

I am prepping for a trip to New York and pulling together some demos and other miscellaneous information. One of the pieces of information I usually give out is more information on using HAT to debug problems. I normally pass along a whitepaper written originally by Guy Lau which I have updated. Please feel free to use to give you examples on how to use HAT in different scenarios.

Now on to the important stuff like pizza in New York!

Since I am frequently traveling in New York, I thought it might be nice to pass along a couple of places to eat and drink if you ever find yourself in the city. I normally am either in the financial district or midtown so locations are convenient to there. 

I can't mention going to New York without mentioning Pizza. If you have the time to go to Brooklyn, Grimaldi's is probably the best pizza in the world. Check it out if you have time. http://www.grimaldis.com/. If you need something more local to Manhattan John's pizza is also really good or at least a close second. There are multiple locations. http://www.newyorkmetro.com/pages/details/8512.htm. The original place is in the village off of Bleeker. I think it is slightly better than, the Times Sq location, but Times Sq is a better sit down place.

Once you have tried Pizza, try SeraFina's http://www.serafinarestaurant.com/. The 61st street restaurant is a great place. Great Italian outside of Little Italy.

Anyway, just a couple of tidbits. There are plenty of other places to check out, but these are a couple that I always have to stop by.

I hope you find the HAT walkthrough useful and if there is anything missing please let me know.

I wanted to focus this message on passing on some documentation on creating a deployment machanism to move from development to testing without having to run the deployment wizard over and over. Since I seem to be constantly modifying my samples and demos and redeploying them, I thought it might be nice to find a way to create a batch file to reuse my samples on different machines and even save them to a CD to take with me anywhere. Along the way I started looking into the way the BizTalk SDK samples were created and decided to copy them.

Once I got this figured out, I basically documented it and posted it here. Please use this as a guide to help you create your own mechanism to go from your development machine to QA. Will keep you form constantly have to run the deployment Wizard to remove it and add it back in. 

I also have found a tool written by Erik Leasburg and Mariusz Borsa that can be used to modifying the binding files. The source and tool is located here. Usage instructions will show up when you run it without arguments form the command prompt.

I hop you find this helpful and let me know what you think.