The primary reason to migrate was that I don’t like the ClearDB MySQL database option as the free version is limited to 4 connections and the paid version is not cheap and/or scalable as running Azure SQL. I really want to have a site that can run as light weight as possible to recommend to people looking for low cost approaches for running sites on Azure Web Sites.
Unfortunately, it seems that the Brandoo project has run into a lot of challenges with the ongoing changes in WordPress. I heard from the Brandoo project owner himself:
I’m the maintainer of Brandoo WordPress. Changes in WordPress upstream is making debugging of Brandoo WordPress almost impossible today. There is many “strange” changes in WordPress, making translation to AzureSQL/MS SQL impossible at all. We do not want to change WordPress core and build fork. Without community we will be not able to work on it. There is no support form developers community so we are fighting alone. Second thing is that today it is possible to connect Azure Web Site with Your own MySQL cluster or single VM via VPN. So You can create HA WordPress on Azure without AzureSQL – I think that this will be a new direction for Brandoo WordPress. I’m father of this project and it is not so simple to tell You, that we are slowly going to the end of Brandoo WordPress in form You know and use.
What this means is that if running Brandoo is a use at your own risk option. While my experience has been positive so far (except for search not working), it means that it’s not a well supported option and one that I couldn’t recommend as a solution to those wanting to run a low cost web site or blog on Azure Web Sites.
My next step for this blog is to look at alternatives to WordPress that will work on a standard Azure Web Site with an Azure SQL database (or even better, no database requirement at all). I’ll keep everyone posted as I look at various options – currently I’m looking at C1 Composite, Ghost, and a few others (I’ve ruled out both Joomla an Drupal as they don’t seem to deploy properly to Azure SQL).