I've alluded to this project before, in this question, but the scope of redesign has been slightly tightened, i.e. I can't redesign the whole thing, so I 'd like some general advice on how to structure the existing artefacts in the application as an incremental step in improving the design.
The site has two areas of functionality, v.i.z. reporting and maintenance. This is the major function of the site, not data manipulation, but just presentation. The site includes a small number of maintenance pages that use standard GridViews and FormViews for maintaining a small area of data. This is why I have decided against a rich, complex DAL in favour of the Enterprise Library DAAB and plain vanilla datasets.
Each report is an isolated page that uses results of a dynamic SQL query to explicitly render HTML table rows for the report. In order to not maintain one set of queries for both mySQL and MSSQL for each report, I will move all data access to stored procedures, removing coupling to either DB engine through the DAAB.
I am looking at reporting tools to decouple the report structure definition from the report presentation. I would prefer to not define report structure in classes, as telerik reporting does, but have yet to look at other reporting tools.
All reports share a common filter page that is presented on choosing a report from a menu, which redirects to the chosen report once the user is happy with their filter selections. Is there any guidance available for this very common scenario that I seem to have to keep reinventing?
I am looking for general advice on how to move this toward a better structured product, without actually restructuring the whole project. Simple stuff like separating maintenance pages from report pages in sub-directories, etc.
I'm not looking for others to do my work, and will in due course have implemented many improvements of my own making, but I would appreciate general opinions on how other people would handle a project like this.
For presentation of reports based upon a database of facts I would investigate one of the tools that has got all the core functionality already implemented, such as BIRT (and I am sure that there are some .NET alternatives). You might think a table of data is good enough for an intelligent person to parse, but down the line someone will ask for graphs and PDFs.
The maintenance / admin pages can be a standard website - forms and backend, using whatever framework you are comfortable with.
Architecturally wise, are you going to be running the reports against a live database, or an archive database (or the failover/replication DB)?
Are you generating aggregate data tables (fact tables) to generate reports from, or running against the more complex main schema?
I'm going to go with a management system for metadata on reports, with a simple directory structure for a page-per-report system, using the DevExpress ASPxGridView. This grid is more than capable of meeting the reporting needs of this application.
I'll be implementing maintenance and admin as standard web forms, with a shared filter form for most reports, dynamically configured based on report metadata. There will be no runtime maintenance of reports, as adding a report requires adding a page. This is acceptable as maintenance will be infrequent.
Related
I just wrapped up an arch review and next-gen recommendation for a client of ours that needs about the deepest level of customization I’ve ever seen for an application. Their desire is to customize their enterprise web application from the UI to the back-end by customer (40+ customers needing control-level customization). The customization will even include special business rules engines and very complex logic involving the transportation industry. As much as is possible, they want developer nirvana by automating everything so customizations can be driven by their customers and have minimal to no involvement by their devs.
Based on my research, though there will need to be some additional plumbing built in as well as security, the DDF will get them closer to their goals more than anything else out there. However, they're requesting more detailed information than what I provided for them.
I really need a case-study or some other such testimony of an enterprise-level company that has successfully implemented the DDF and gives details as to the enterprise problems it solved for them. Any direction or help would very much be appreciated. Thanks!
Since it is now July, your question is probably OBE by now. However, I have designed and fielded a transportation scheduling web app (ASP.Net 4.0) currently in use by 15 facilities within the Army and Air Force using Dynamic Data. This is a single instance, scalable web appplication adapting to customer requirements through database resident configuration settings. I extended the field templates to use Telerik ASP.Net controls and be configurable by user role and facility.
I have found little in the metadata that was much of a hindrance in providing a flexible configurable UI.
Well at least one word of caution. One important aspect (and selling point) of DDF is the assignment of metadata attributes to help scaffold columns and tables and the use of new dynamic data controls to gain advantage of that metadata (like QueryableFilterUserControl or DynamicDataManager or PageAction). One aspect of metadata however is that it is assigned at run time, and cannot be manipulated once the application has started. Therefore different users would all be logging into basically the same metadata set, and customization based on user would be a nightmare. You can certainly set security and permissions based on group roles, but control level customization would be difficult. I hope this helps.
my next assignments is to build 2 information portals for customers. These portals will be login protected sites and contain a set of pages displaying information like orders, invoices, pdf-files ... for the authenticated user (all presented as lists with links to detail pages). The users and the data are stored in an Oracle database. The portals differ in some of the features and in the layout.
My standard approach is to build an individual ASP.net Web Application for every portal.
But this is not the best way to get something reusable. So for these two projects my idea is to create a set of WCF services to get the Data from the Oracle database and to build user controls to display the different elements in Umbraco. This way I hope to get a set of independent, reusable “modules” which can be used to build these portals.
Now my question: is Umbraco a good platform for this type of projects? And is my “concept” a valid approach?
Kind regards
Volkmar
Umbracois very flexible. ON the one hand there is the question about security: With Umbraco you can use any Membership Provider you want for all visitors ( also with member roles).
On the other hand you have the question of the integration: With Umbraco you can create usercontrols, xslts or razor files as macros (which can be seen as the reusable modules).
For Xslt you can implement your own XsltExtension which pulls the external content as XPathNodeIterator you can use in every Xslt macro. For ascx files or razor you can use LinQ2Umbraco, your own objects etc to connect to the oracle database.
You also can use some sort of caching functionality to reduce the db-calls. On the other hand is one of the biggest advantages that Umbraco stores all the content as xml and object tree in memmory. So it is very fast in content rendering. With every database call you are loosing a little bit of this advantage.
hth, Thomas
Ruben Verbourgh began the Oracle4Umbraco project to create an abstracted fork for the Datalayer to support running on an Oracle DB. You can find it at http://oracle4umbraco.codeplex.com/, although it has no active releases, so build from source and YMMV.
Volkmar, your concept is perfectly sound - although you might want to consider using the Umbraco data store as the persistence layer for your data rather than in the Oracle DB itself. You get XML content versioning, caching, and all the benefits of the content-management side of things, in a robust and flexible framework which you can expose to other apps later should you so need to, through the Umbraco APIs and web services.
HTH,
Benjamin
content management of website becomes simplified with Umbraco.
But if you are planning to use Oracle as backend, Umbraco does not have support for it.
So decide carefully as to what parameters can be compromised.
Good luck.
I was involved in a SharePoint(WSS) project that was very data centric. The project consisted of more than 500 lists that has very complex relations between them. The client also asked for more than 350 Reports. Don't tell me why did you use SharePoint from the beginning. It was a managerial decision and we already delivered the project after 14 months of pain (this is 6 months overdue to the deadline)
When we first started the project, we didn't know anything about SharePoint development (believe it or not). The management said that they will take the risk. They were very convinced that SharePoint is the optimal solution for anything!!!(well, that proved wrong at the end of the project).
Anyway, we were learning SharePoint while we were developing. Our development were mainly based on SharePoint designer to customize all the AllItems/NewForm/EditForm/DispForm for every list to provide the needed logic/validation that the client asked for (Using JavaScript). We also implemented around 15 Custom Fields (e.g. master-details fields). We also made an event receiver to handle all the adding/updating/deleting... events for all the lists in the site. Plus around 40 ASP.Net user controls.
The main problems that faced us (we worked-around it but sadly in an inefficient way)
1- The Client asked for a search web part in each AllItems.aspx. The search web part should have multiple keys for the client to search with. We did that using Form Web parts using SPD no problem. But the real problem was how to search for a related field that was not in the current list. (So, in such cases we had to save these fields values in our list to be able to search for (crap, I know!!)). You might ask, why didn't you implement ASP.Net user controls for such task? Well, that would require us to forsake the default AllItems web part, and were already customized hundereds of AllItems.aspx pages with alot of customization that would take us a lot of time to reimplement them from the beginnging. Also, even if we used user controls, CAML is very inefficient in retrieving data from multiple related lists!
2- I think you can guess this one, if we've already faced a big time doing search web parts, how on earth will be able to do the 350 reports!!:D But we figured out a work-around (as usual :S) we made an Access DB file with links to all the 500 sharePoint lists, then we implemented a user control that has a report viewer control. This user controls takes an ordinary T-SQL Query to query on the Access DB, the Access DB retrieves the data from the SharePoint DB and pass it back to the user control which views the DataSet on the report viewer.
There are other administration related problems, but I would like to focus on the development here.
So, after I showed you the picture (sorry for the long post). What do you think was the best SharePoint development technique that we should have taken in such a data centric Project, if any?
I heard that some companies doesn't use lists at all in such projects, and builds there own SQL database tables instead of the SharePoint Database. But I can't keep my self from wondering, If I'm making my own DB, and hence implementing my CRUD web parts from scratch (We will also lose the security module benefits provided by SP Lists), what would be the benefits of SharePoint?
Once again I apologize for the long post.
I think you found out exactly what I did. Sharepoint just isn't good at handling large enterprise type applications. We ended up creating a custom database to house our data. We used Webparts for the user interface, but otherwise, the entire application was independent of Sharepoint.
In my opinion, Microsoft is overselling Sharepoint. It's actually good at team collaboration sites and simple Excel services applications, but anything beyond that it just isn't capable of handling.
I have to disagree with both Geoff and Abu in regards to SharePoint being a bad choice for large enterprise applications.
As you state yourself Abu your team were learning on the job as you had no SharePoint development experience, the issues you faced was more a Management error that a platform problem, management should have brought in SharePoint contractors to work alongside your team to help build what sounds to be a fairly complex system.
As a developer who has worked with SharePoint for a number of years many of the projects I have worked on some that I myself would not have believed suitable for SharePoint with in my first few years of developing on this platform, however now with more experience I know how to leverage the power of the platform far better and I realise the advantages gained using SharePoint for projects of that nature. That said I have a number of issues with parts of the platform but this is no different to any other platform I have worked on including parts of the ASP.Net platform.
If I was asked to develop a solution using a bespoke Java based system (or perhaps the new MVC platform) I am sure I would experience many problems similar to what you experienced where I simply don’t know what the right approach would be. That would not in any way be an issue with the platform but more with my inexperience.
I am sorry to hear that both of you experienced pains working within the bounds of the SharePoint platform that was forced on you by your management. Though I am disappointed that you are so fast to point the blame away from yourselves and your management.
Was SharePoint the best platform for your projects I can’t say, but that doesn’t make it a bad platform for enterprise applications.
I disagree with Geoff that SharePoint is no good for large enterprise type apps. You have to remember from the start that SharePoint is a development PLATFORM. This means it gives tyou a lot of functionality out of the box, but is extremely customisable.
Being a platform does not mean every bit of customisation needs to be done based on SharePoint lists. Seeing as it is built on ASP.NET you can do anything in SharePoint you could do in ASP.NET as well.
I have built a great deal of ASP.NET apps that were hosted in SharePoint, letting SharePoint do the authentication etc.
I have to say though, determining where SharePoint should stop being your base and you should switch to regular ASP.NET is sometimes hard..
If you are looking for a data-centric solution in SharePoint, the best solution is to use the Business Data Catalog (BDC). This keeps your rich data relationships and all of the SQL (or other DBMS) goodness you want where it should be - in a repository designed to be optimal at storing data.
For an overview of what the BDC functionality can do, see this post on the SharePoint Team Blog. For much more details, read the series on SharePoint Magazine. Note that these features requires an Enterprise license of SharePoint 2007.
This may sound a bit general, but I have a startup that is working on an ASP.NET (greenfield) suite of software applications. We are aiming to spend a substantial amount of time in the architecture phase to develop a strong foundation for our software. I was wondering if anyone has any advice, anything we should focus on or any suggestions for areas we should focus on to build a better suite.
Some things we are focusing on right now:
1. Session state requirements - should sessions be sticky or should we take server clustering session migration into account.
2. User login authentication - what are the major concerns in this space - LDAP, AD, custom SQL authentication systems etc.
3. The DAL - ORM vs Stored Proc
4. Integrating multiple ASP.NET applications in a single software suite. How it should look/feel. How it should be architected, etc.
I would appreciate any advice from any architects out there that have built similar systems from the ground up.
I know there are lots of solutions to session, but if you can create your framework to be session-free, you will avoid a lot of potential headaches. (There are lots of session-free options, but an obvious one is a hidden form field, somewhat like ViewState.)
Just some quick notes. I can't get too detailed since we went through this exercise where I worked last year - and I don't work there any more!
Start from the beginning using Enterprise Library, especially the Logging and Exception Handling application blocks. I've also found their Unity dependency injection library to be very useful.
Consider using Visual Studio Team Foundation Server. It's not just for source control, but can create you a complete continuous integration solution, complete with integrated bug tracking, code quality tracking, etc. If you've got the time and people, it's well worth a man-month to learn how to do an initial deployment.
You may want to buy one or more licenses of one of the Visual Studio Team System editions. You don't need these versions in order to use TFS, but they work well with it.
Consider globalization right from the start. Same with customization, if your suite will run on customer premises and be customizable by them.
You haven't said how large your team is, or is expected to be. If it's large enough, you'll want to spend at least a man-week learning a bit about what's available to you in terms of Visual Studio Extensibility. Your developers (and maybe also your QA folks) will all but live in Visual Studio, so the ability to customize it to meet your needs can be a big win. Whether it's just some macros and maybe some customized project or item templates, or whether you want to do add-ins or more, Visual Studio is very extensible.
Be certain to use WCF for any web services work. The older ASMX web service technology is now considered by Microsoft to be "legacy software".
Finally, be sure to find out whether you qualify for BizSpark, "A program that provides Software, Support and Visibility for Software Startups." And does so almost for free.
I saw a demo of Silverlight 3 at the PhillyDotNet User Group last night - WOW. Wow for business applications, not graphic applications. There is a learning curve, but you get a lot for it. For example, the demo showed a grid being bound to a table without needing to write any code.
Right out of the box you had sorting, editing, paging, etc. But it wasn't the lame stuff you normally get and then have to rework. For example the paging was smart enough to write the sql that would only bring back the 20 rows you needed for the page.
The demo continued with him putting a detail form on the page for editing. Again no code, but it was smart enough to know that it had the same datasource as the grid on the page. So as you were moving row to row on the grid - the detail form was showing the current row (and it was very fast).
Both the grid and the detail form were editable and as you changed a field in one the other would reflect the new value. The editing was smart enough to validate the field on its own. So you couldn't put a letter in a field that was an integer type, etc. It also limited the number of characters that could be entered based on the column size found in the database. All the date fields on the detail form automatically had a calendar next to them. You get the idea - no coding for any of this.
If this weren't enough, it can be used to build occasionally connected applications. So he showed how he updated a few records on a few different pages, had the option to revert back a field later (ctrl-Z), and then at the end submitted all the changed records to be saved.
Also, they said it works with Linq2SQL and the entity fraimwork.
So if I were building a new product now, I would really look into this as a way of differentiating my product. And I suspect that if you don't do it with Silverlight now, you will be rewriting it in a few years anyway.
Here is a link to a demo (not the one I saw.)
Some general thoughts. If you'd like me to expound on any of these, let me know.
Inheriting from a custom subclass of
Page instead of Page itself is a
great way to share functionality
across your site.
Nested MasterPages are good.
Charting: I've tried DevExpress,
Syncfusion, and MSChart control and
all have their own issues.
The built-in forms authentication is
pretty good. Building a site that
allows both integrated authentication
and forms authentication is tricky
but can be done.
I've tried using cross page postbacks
and I'm still not sure if I like
them.
Localization takes a lot of time
Come up with a good structure for your App_Themes and css.
Use Elmah to track unhandled exceptions
I am in the early stages of planning a conversion of a large classic ASP database application to ASP.Net and I'm having trouble picking out which data access method to use. I have played around with Linq To SQL, Dynamic Data, strongly typed datasets, Enterprise Library (Data Access Application Blocks), and a tiny bit with Entity Framework, but none of them have jumped out to me as "the one". There are just too many choices - my head is swimming, help me choose!
Perhaps it would help to give some background on the application that I am converting along with the priorities...
The back end is Microsoft SQL Server (2005 or later) and we are committed to that, so I don't need to worry about ever supporting a different database platform.
The database is very mature and contains a great deal of the business logic. It is highly normalized and makes extensive use of stored procedures, triggers, and views. I would rather not reinvent two wheels at the same time, so I'd like to make as few changes to the database as possible. So, I need to choose a data access method that is flexible enough to let me work around any quirks in the database.
The application has many data entry forms and extensive searching and reporting capabilities (reports are another beast which I will tackle later).
The application needs to be flexible enough to deal with minor changes to the database structure. The application (and database) may be installed at different sites where minor custom modifications are made to the database. Ideally the application could identify the database extensions and react appropriately. In other words, if I need to store an O/R mapping in the application, I need to be able to swap that out (or refresh it easily) when installing the application and database at a new site.
Rapid application development is critical. Since the database is already done and the user interface is going to closely match the existing application, I'm hoping to find something where we can crank this out fairly quickly. I am willing to sacrifice not using the absolute latest and greatest technology if it will save time in development. In other words, if there is a steep learning curve to using something like Entity Framework, I'm fine with going something like strongly typed Datasets and a custom DAL if it will speed up the process.
I am a total newbie to ASP.Net but am intimately familiar with Classic ASP, T-SQL and the old ADO (e.g. disconnected recordsets). If any of the data access methods is better suited for someone coming from my background, I might lean in that direction.
Thanks for any advice that you can offer!
Look at all three articles in this series:
High Performance Data Access Layer Architecture Part 1
Great advice.
You may want to look at decoupling the database layer from the asp layer so that you can not only give more flexbility in making the decision, but when you have to make changes to a customer's database you can just swap in a new dll without changing anything else.
By using dependency injection you can use xml to tell the framework which concrete class to use for an interface.
The advantage to doing this is that you can then go with one database approach, and if you later decide to change to another, then you can just change the dll and go on without making any changes to other layers.
Since you are more familiar with it why not just go directly to the database at the moment by making your own connections? Then you can move the rest of your code and along the way you can decide which of the myriad of technologies to use.
For a new application I am working on I am starting with LINQ to SQL for it, mainly because development will be quicker, but, later, if I decide that won't meet my needs I will just swap it out.
nHibernate might be a good fit. You can store the mapping in external configuration files which would solve your needs. Another option might be using ActiveRecord, which is based upon nHibernate.
nHibernate has a neat feature which you might find helpful. It's called a Dynamic property which is basically a name value pair collection populated by pulling the column names from the mapping file. So when you add a column at your client site, you update the mapping file and you'd be able to access the data through a collection on the object.