SNESSUG 4/14/2010

April 14, 2010 at 7:32 pm (SNESSUG, Visual Studio) (, , , )

Tonight’s Southern New England SQL Server Users group is sponsored by Idera. Our presenter is Scott Abrants of Iron Mountain. He’s talking about deploying databases using Visual Studio Team System:Database Edition. We have a good turnout with 12 people (yeah, we’re small).

Scott’s presentation was a lot of fun and very informative. He’s very involved with automating his deployments to a fare-thee-well. He really has Visual Studio dancing and singing. It was a very thorough overview of the VSTS:DBE soltuion. Other user groups should be jealous that we got to see this presentation.

Permalink 3 Comments

Southern New England SQL Server Users Group

April 13, 2010 at 11:02 am (SNESSUG) (, , , , )

Tomorrow, Wednesday April 14th, is the next SNESSUG meeting. We’re going to get a great presentation from Scott Abrants on using Visual Studio Team System for database deployments. I saw Scott presenting this at SQL Saturday:Boston to a packed room. If you didn’t get to see it then, come on down to Rhode Island tomorrow evening. You won’t be sorry.

Permalink 1 Comment

Visual Studio Team System Article

June 24, 2009 at 8:06 am (Visual Studio) (, , )

I wrote up an article on how we’re configuring & deploying databases to disparate systems using a combination of database projects, server projects and compound projects in conjunction with configurations that has been published over at SQL Server Central. Please click over & read it.

Permalink Leave a Comment

MSDN Blog on Data Management in VSTS:DB

March 31, 2009 at 2:21 pm (Visual Studio) (, , , , )

Another excellent post over on Barclay Hill’s blog. This time he’s showing how to use pre- and post- deployment scripts to manage data when deployment might result in data loss. I am jazzed for part 2 of this set to see how what I should have been doing all along.

Permalink Leave a Comment

VSTS:DB Composite Projects

February 4, 2009 at 11:16 am (SQLServerPedia Syndication, Tools, Visual Studio) (, , , )

I’ve seen several searches go by from people who are looking for how to create composite projects. Here’s a quick run-down. In the example, I’m going to create a main database project called BaseDB, a server project called MyServer, and a composite project that combines the two called Comp. This project is just another database project as you can see in Figure 1:

New Database Project

New Database Project

Once the new project is created, you need to right click on the References object in the Solution Explorer. Figure 2:


From the pop-up menu select Add Database Reference. The window in Figure 3 will pop up:

Add Database Reference

Add Database Reference

You can select the other projects for a dynamic view into their changes, or for a static look you can select a .dbschema file. Unfortunately, the .dbschema file is very static. You won’t get changes as they’re made. Usually I use the projects.

You can try to affect Database Reference Variables for the project, but usually this isn’t necessary. If there are indirect references you can suppress them. Again, this isn’t something I’ve had to use so far.

Once you’ve created the references to other projects or .dbschema files, you can see these on the References tab of the project Properties as shown in Figure 4:


You now have a composite project. There are a number of other options you should consider. For example, how are the deployments done. One common approach that I’ve taken is to have the base database do a full rebuild of the database and then have the other two projects do incremental deployments. You’ll need to make sure that all the same Configurations are set up between the projects so that the stuff like the connection strings are the same between projects. You’ll want to make sure that the base project and the composite project are deploying to the same database name.

There’s more, but that should get you started.

Permalink 1 Comment

Visual Studio Team System Database Edition Best Practices

February 2, 2009 at 9:41 am (SQLServerPedia Syndication, Visual Studio) (, , )

Barclay Hill, Program Manager for Visual Studio Team System Database Edition at Microsoft has just launched a new blog that I’m excited about. He’s going to be blogging on best practices for VSTS:DB. His initial post outlines the topics he hopes to address and it’s an impressive list. I responded immediately with an offer to help because I’m using the tool and struggling a bit to get it to do everything we need and because I really like all that VSTS:DB has done for me and my team already. Any one else interested in pitching in should go over there and get involved.

Here’s what I sent as an initial message:

I saw your blog post and request for interest and participation and decided to pitch in. The company where I work has been using VSTS:DB through several iterations for about two years now. We have several automated mechanisms for deployments and some standards in place for how best to use & configure the tool. But we need more, so I’m pretty excited about your project. The three most interesting topics are the three that we’ve wrestled with the most, Build Automation and Build Management, Deployment Automation and Continuous Integration and Targeting Multiple Database Environmments. The biggest problem we face today is with incremental deployments where we want to retain the data. Clearly builds of this type require a manual intervention, but unfortunately, at this point, with VSTS:DB, we can’t then automate the deployment but must deploy manually.

Anyway, I look forward to your future posts and please feel free to bounce things off of me. I can’t guarantee I can help, but I’ll sure try.

Permalink 3 Comments

VSTSDBE GDR RC0: No Joy in Mudville

November 14, 2008 at 10:32 am (PASS, Tools, Visual Studio) (, , )

A few posts ago I outlined the problems I was having getting the RC0 of the GDR to deploy appropriately. I was successful in a deployment, so I thought I was done. Yesterday I went to do a run through of the presentation for PASS and my deploy failed. Despite the fact that I’m running on the same machine with the same project that worked before. I’ve now gone back through the whole process again, but I’m getting nothing but failures. It’s totally hosed. At this point, barring a miracle, I’m going to present the process and then explain that I can’t demo it because of failures.

I don’t know if it’s just a configuration issue with my virtual machine or something flawed in the release. Either way, I’m at a standstill and completely bummed. I’m going to try recreating things, but I suspect that I’m going to have to assume a failure at the conference and not run the demo.

I posted the issue and the error messages I’m getting over at the MSDN Forum for VSTSDBE

Permalink 3 Comments


November 4, 2008 at 3:10 pm (Tools, Visual Studio) (, , , , )

Second round of testing. Instead of associating with a project, I tried creating a reference to a .dbschema file. Same error. This time, I’m going to clear out everything. I tried creating it initially on top of the code from CTP 17.

So, another chance to try out the reverse engineer process of “Import Database Schema.” Works great. New AdventureWorks database inside Visual Studio with a tested deployment faster than you can spit.

Created a new server project. Added a login, just to give it something to do. It deployed fine.

On to the compound project. Let’s see what happens. Just to see, I did a build and deploy before I added references or objects or anything. It worked great. Whatever that means. I’m taking it a step at a time, to try to see where it breaks, if it breaks. So, I’ve added just the Adventureworks database project (not the .dbschema file). Trying a build & deploy now. Build went through. No issues. Deploying… Damn. I’m hitting a collation error. However, that’s good because I never got that far before… It worked! On to adding the server as a second reference and then deploying that… Whoops, there it is. Failure. Remove, reset and redeploy without the Server project.

Time for some more experimentation…

It looks like I was getting an error in the server project. I rearranged a few things and now I have the whole thing working again.

There is joy in Mudville!

Permalink Leave a Comment


November 3, 2008 at 4:01 pm (nHibernate, Tools) (, , )

There is no joy in Mudville.

The install of RC0 went smoothly. I was able to create a new server project and deploy it. I was able to reverse engineer AdventureWorks into a new project and deploy that. When I went to create a new compound project, combining the output from the two… deployment failed. I got an arcane error about something in the Microsoft.Data.Schema.SchemaModel.ModelSerializationException erroring out. Useful. So, at this point, the functionality I had in my demo for the PASS Summit isn’t working. Yikes. I’ve got two weeks… no pressure…

I posted a note about it over at the MSDN discussion site. If you have a solution, swing by and drop it off.

Permalink Leave a Comment


October 31, 2008 at 11:08 am (Tools, Visual Studio) (, , )

I’ve downloaded and started the install. The first screen I’m seeing is not filling me with confidence. The executable that is currently running is called “DBProRepair.exe” However, the installed completed without any issues.

Just for giggles, I tried opening my PASS presentation (created on CTP 17). If it worked, I wouldn’t have to recode. Ah, well, no joy. Not that I expected any. Oh, but the Conversion Report that comes up afterwards is very informative.

Reverse engineer against the AdventureWorks2008 database went off without a hitch. All the objects imported into the database with no issues too. So far, so good. Created all the projects. No issues. I’m configuring for a deployment now. We’ll see how that goes. Builds worked fine. I forgot to change the deploy action the first time I did a deployment. Second time… Frell!

The compound deployment is failing. I had something configured wrong… I’ll post when I get it figured out.

Permalink Leave a Comment

Next page »