|
|
You're in the
dotnet distributed apps
group:Beginner's Questions
dotnet distributed apps:
I have written an app with a Windows Forms UI that is deployed to clients across the internet via Terminal Services. It uses SQL Server as a data store, and ADO.NET for CRUD. The data access layer is currently completely ensnarled within the code for the Windows Forms app, and calls stored procedures on the data source using Windows Integrated Authentication. The user's identity is used by several triggers to identify who last updated data. I am considering moving to a distributed architecture that would involve a web services data access layer on a web server, and a deployed Windows Forms Presentation layer client on user machines accross the internet. I know this is quite a radical change, but hardware and bandwidth costs will eventually make it very justifiable. My questions.... 1. What is the single best book I can buy in this regard? 2. Currently, users sign-on to my domain when they access my app via Terminal Services. This allows me to use suser_sname in my AfterUpdate triggers. It also allows me to establish very specific permissions for these users based on AD GPO's and SQL server roles. Furthermore, the user's Window's username is the primary key in a Client Access Table and is required to return, via sproc, the name of the database he/she is allowed to connect to, completing an SQL connection string that initially reads "Initial Catalog=". How could I achieve this in a distributed app environment? I believe a standard approach is to have the client authenticate at the DALC, and then have the DALC run as a single domain user account in order to take advantage of connection pooling. Obviously, this blows my afterupdate triggers and initial user-to-database redirection strategy. 3. I am considering sticking with Windows Forms as my app uses a fair amount of drag-drop and graphical image manipulation (also reasons to get away from Terminal services). Are these valid reasons for avoiding a web forms app for the presentation layer? 4. There is obviously no real data going over the internet with the Terminal Services configuration. The data involves protected healthcare info, and WILL go over the internet if I adopt the distributed structure. Would SSL be the recommended means of securing the data? If so, would this offer any alternative authentication strategies? I know these are BIG questions, and I don't expect to see any books written in reply. But any pointers to good references, walkthroughs, etc, would be great. (Already read MS Distributed App Architecture Guide and Microsoft Guide for Designing Data Tier Components and Passing Data Through Tiers). -- Thanks.
Hello.
As for books, I recommened the following two: XML Web Services in the Organization http://www.microsoft.com/mspress/books/6534.asp Understanding Web Services Specifications and the WSE http://www.microsoft.com/MSPress/books/6708.asp The reason I recommend the last one is for security. SSL is a proper way to protect your data. Another is WS-Security and WSE. Here is also a good article for WSE: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnwse/html/ wse2wspolicy.asp Hope this help, Luke
Hi JT. Yes this is quite a change. We are going under a similar process and
I think your question is of interest. We validated this solution with Microsoft and this architecture is very secure. We do use Windows form on the presentation layer which is on terminal server and call web services on the business logic side. IIS in the entry point for our SOA architecture and we are using only Windows integrated autentication in order to autenticate who is gaining access to the server side. This way, the credential of the client side are transported to the server side and the thread still knows the identity of the autenticated user. So the information is still availlable for you to connect to your database. Knowing that and knowing you want to use connection pooling, there are ways to still do what you want. On our side, we are using an Oracle Database. So we decided to use "Proxy" authentication on the DB Side. Which means that we are going to have only one real user interfacing the database but the identity of the authenticated user will be passed as a parameter. Then the database user will work "on behalf" of the real user. Thus all transactions will be stamped with the identity of the real authenticated user. I know this information is very roughly said but there are many books about SOA architecture that can give you good information regarding identity passing and how you can achieve that. Also, there may be an equivalent way of using "proxy" authentication on SQL Server. If your architecture reside on the same domain and you are not going on the internet, to my point of view there is no need to go for SSL. But if so, I will strongly recommend it. The infromation between the two tier will by passed using XML which to my point of view is still a good way to bind to data controls on the presentation layer. Hope this help. [quoted text, click to view] "JT" wrote:
> Hi all, > I have written an app with a Windows Forms UI that is deployed to clients > across the internet via Terminal Services. It uses SQL Server as a data > store, and ADO.NET for CRUD. The data access layer is currently completely > ensnarled within the code for the Windows Forms app, and calls stored > procedures on the data source using Windows Integrated Authentication. The > user's identity is used by several triggers to identify who last updated data. > > I am considering moving to a distributed architecture that would involve a > web services data access layer on a web server, and a deployed Windows Forms > Presentation layer client on user machines accross the internet. I know this > is quite a radical change, but hardware and bandwidth costs will eventually > make it very justifiable. > > My questions.... > > 1. What is the single best book I can buy in this regard? > > 2. Currently, users sign-on to my domain when they access my app via > Terminal Services. This allows me to use suser_sname in my AfterUpdate > triggers. It also allows me to establish very specific permissions for these > users based on AD GPO's and SQL server roles. Furthermore, the user's > Window's username is the primary key in a Client Access Table and is required > to return, via sproc, the name of the database he/she is allowed to connect > to, completing an SQL connection string that initially reads "Initial > Catalog=". How could I achieve this in a distributed app environment? I > believe a standard approach is to have the client authenticate at the DALC, > and then have the DALC run as a single domain user account in order to take > advantage of connection pooling. Obviously, this blows my afterupdate > triggers and initial user-to-database redirection strategy. > > 3. I am considering sticking with Windows Forms as my app uses a fair > amount of drag-drop and graphical image manipulation (also reasons to get > away from Terminal services). Are these valid reasons for avoiding a web > forms app for the presentation layer? > > 4. There is obviously no real data going over the internet with the > Terminal Services configuration. The data involves protected healthcare > info, and WILL go over the internet if I adopt the distributed structure. > Would SSL be the recommended means of securing the data? If so, would this > offer any alternative authentication strategies? > > I know these are BIG questions, and I don't expect to see any books written > in reply. But any pointers to good references, walkthroughs, etc, would be > great. (Already read MS Distributed App Architecture Guide and Microsoft > Guide for Designing Data Tier Components and Passing Data Through Tiers). > -- > Thanks.
Thank you both for your answers. I have three follow-on questions...
1. Authenticating the user. My users will all be communicating with my web service(s) accross the Internet using the downloaded Windows Forms client I supply them. What is my best bet for authenticating the users to my web service? To my database (SQL Server)? I am assuming that Windows Integrated Auth will not work for the web service. Or perhaps it would, if I gave the users domain accounts. Can this be done outside of the RDC/Terminal Services environment? If not, would it be reasonable to store usernames and passwords on SQL Server, and use basic auth with https? 2. SQL Server 2005 Web Services. If I upgrade to SQL Server 2005, I can expose its native web services, and perhaps avoid using a middle data access tier and IIS altogether? If I require all of my Internet users to use a Windows 2K or greater OS, can I authenticate them to this type of web service using Windows Integrated Auth? I would like to avoid mixed auth to the SQL server. 3. My current terminal services app has about 25 typed datasets, each with from one to 40 datatables (average of three per dataset). Should I create one huge web service with methods on all of these business entities, or many smaller services? Thanks for any further advice. -- John [quoted text, click to view] "[MSFT]" wrote:
> Hello. > > As for books, I recommened the following two: > > XML Web Services in the Organization > http://www.microsoft.com/mspress/books/6534.asp > > Understanding Web Services Specifications and the WSE > http://www.microsoft.com/MSPress/books/6708.asp > > The reason I recommend the last one is for security. SSL is a proper way to > protect your data. Another is WS-Security and WSE. Here is also a good > article for WSE: > > http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnwse/html/ > wse2wspolicy.asp > > Hope this help, > > Luke >
Hello,
1. Windows Integrated Auth can be used for web service, but not for internet user. You may consider VPN conncetion and your user can log in with domain account on a VPN connection. 2. SQL Server 2005 can expose data via web service, but Internet users cannot log in with Windows Integrated Auth since they don't have domain account. 3. In my opnion, smaller services are more flexible than a huge one. Hope this help, Luke
Don't see what you're looking for? Try a search.
|
June 2003
July 2003
August 2003
September 2003
October 2003
November 2003
December 2003
January 2004
February 2004
March 2004
April 2004
May 2004
June 2004
July 2004
August 2004
September 2004
October 2004
November 2004
December 2004
January 2005
February 2005
March 2005
April 2005
May 2005
June 2005
July 2005
August 2005
September 2005
October 2005
November 2005
December 2005
January 2006
February 2006
March 2006
April 2006
May 2006
June 2006
July 2006
August 2006
September 2006
October 2006
November 2006
December 2006
January 2007
February 2007
March 2007
April 2007
May 2007
June 2007
July 2007
August 2007
September 2007
October 2007
November 2007
December 2007
January 2008
February 2008
March 2008
April 2008
May 2008
June 2008
| home · blog · groups | Questions? Comments? Contact the dn |