Main Site Documentation

Connection to MSSQL database


#1

Hello to all!

I am new to this topic and hope somebody can answer my question. My aim is to build a traceability system that would work in the production. The system should collect some process data and send them to a central MSSQL server.(SELECT, INSERT, UPDATE)
A friend of my has recommended a possible solution with the FEZ spider. I have installed the Visual C# 2010 Express development tool, the .NET Micro Framework 4.1 SDK and the Gadgeeter SDK to get familiar with this tool. I saw, the micro framework does not provide the namespace System.Data like I know it from the framework for windows applications, therefore my Question: is there an other possibility to realise a connection between the FEZ spider and MSSQL database or is there no possibility?

Thanks ahead and sorry if my english is not perfect…


#2

In order to make SQL calls from .NET MF to SQL Server the platform would have to have TDS protocol implementation. Since this is a limited resource solution a TDS driver is not available. Instead you should think about having a thin client of some sort that would accept calls from your board and transfer them to your database. First thing that i can think of is a simple REST API frontend in a web app. .NET Micro Framework supports also DPWS so you can try using WebServices as well.


#3

Oh, and welcome to the community :slight_smile:


#4

It will be best to abstract the database layer - but in an efficient way.

Have a look at how the people at https://pachube.com do it. You can use their service - or just copy the idea for your own internal service :slight_smile:

How much process data do you plan to collect and at what frequency? Issues like queueing, guaranteed delivery, idempotence (handling duplicate submissions) and firehosing the data in become interesting.


#5

thank you for replies. @ realiser: There are only few process data, I’ve created a table with 9 Fields, and each machine would send two times data per process step. Insert at the beginning, and update at the end of a step. There are all together abouth 50 machines and I will get estimated 1 protocol per second.