Use existent VM Instace (bitnami) for Autoscale Group of Instances

[Francisco Insúa]

Hi guys! I am using a Bitnami Wordpress for Google Cloud. Now, I need to setup a Instance Template -> Group of instances -> Load balancer and with this, my system will be autoscaling :)

But, I have the VM instance autocreated by Bitnami, and I need to put in a group of instance.

Can you helpme with this, please?

Thanks a lot
Francisco

[Francisco Insúa]

Hi guys! I am using a Bitnami Wordpress for Google Cloud. Now, I need to setup a Instance Template -> Group of instances -> Load balancer and with this, my system will be autoscaling :)

But, I have the VM instance autocreated by Bitnami, and I need to put in a group of instance.

Can you helpme with this, please? Sorry for my english..

Thanks a lot

[Faizan (Google Cloud Support)]

Hello Francisco,

Before you autosclae your Wordpress instance you need to separate your frontend and backend(MySQL database) first. For your database you can choose CloudSQL or setup MySQL on separate GCE instance. Once done, you can create an image of your Wordpress instance boot disk and than use the image to create the template for instance group. You can refer to this discussion thread which has useful information on configuring database on CloudSQL with Wordpress.

I hope that helps.

Faizan

[Matthew Lenz]

Unfortunately you can't load balance wordpress without jumping through hoops.  As Faizan stated you'll want a common database backend such as CloudSQL but wordpress also stores assets directly on the instance.   For that you would require some kind of syncing to occur between all your live wordpress instances whenever content is changed.  Is it doable? yes.  Is it going to be a big pain in the but? oh yeah.

So basically autoscaling isn't really an option in my opinion.

You are much better off using one of the static cache plugins, varnish and cdn.  Assuming the site is almost all reads with writes being limited to a few people even a very small instance could handle a huge number of views per day.