Showing posts with label Cloud. Show all posts
Showing posts with label Cloud. Show all posts

Wednesday, April 24, 2013

Azure


Steve Sanderson gave fabulous presentation on Single Page Applications at TechDays.  Bart Jolling shows how to publish this application to Azure.
I was able to get the Delivery Tracker demo working on Azure after a few tweaks.  I've outlined the details below.  I believe the main problem was that the database gets created on the fly in code.  The Azure version never creates a database and this may be due to the "cluster index" on primary key issue.  Azure SQL requires a "cluster index" and the Delivery Tracker doesn't create one.
Issues to overcome:
  • System.Net.Http version problem on my development system
  • Get code working locally on the development system
  • Preparation on Azure - create empty web site with database
  • Change the Firewall settings on Azure to allow MMSM access
  • Database is created in code, which did not work on Azure
  • Connect to the Azure database in MMSM using server name and user account
  • Export database from development to Azure will fail with Cluster Index error
  • Primary Key needs a Cluster Index for Azure using MMSM script
  • Download the proper publishing profile after all this is done
  • Use "Publish to Azure"  in Visual Studio and load Azure profile

Monday, December 10, 2012

YQL for Web Data

The web is your own personal database.  All the information is out there, you simply need an efficient way to access it.  YQL is the Yahoo Query Language and goes a long way toward bring your data directly to you in a well structured format.

So how do you get started in YQL and do you need to be a programmer?  Let's start here: try grabbing some stock quotes.  Sure, it looks a little imposing with angle brackets and keyword tags, but play around with it -- test it, give it a go.

Yes, the data is out there and someone might already have your data in a YQL table -- check it out.

The old way of downloading web information used what is called a screen scrapper or some other form of parsing the HTML from a web page.  The problem here, if the web page changes, the data access is broken.  Or you could grab a spreadsheet like format, like the Yahoo stock quotes in CSV format and use a programming environment like the Windows .NET framework to parse the data in a proprietary fashion.

Now you can get the stock data using a standard YQL approach.

Other sources of information:






Tuesday, August 28, 2012