I don’t have much experience with Ansible but I thought I would give it a try on a lab environment I am setting up to test load balancing. This is the planned configuration of the lab, which is really a proof of concept to see if we will get the same or better performance out of a more professional network setup.
(1) Backup server – Medium system with 8Gb, 12TB RAID5, GigE
(2) Database server – Medium system with 8Gb, 4TB RAID1, GigE
(2) Web server – Medium system with 8Gb, 4TB, GigE
(1) Master – Medium system with 8Gb, 4TB RAID1, GigE
- Setup the backup server as a management system, Nagios/Icinga, and SSH server for file backups
- Setup the database server with MySql
- Setup the Web servers with a dummy WordPress environment, just for testing load balance capabilities
- Setup the Master server as HAProxy
This is only the test environment, there isn’t enough memory in these systems for production. In production we will probably have 4-6 web servers, and everything will have at a minimum double that memory. The storage requirements for the Backup server shouldn’t be too high so that might stay the same. This backup server would perform a weekly dump to another backup server offsite at one of the branch offices. It is possible that I will have virtual machines preconfigured and ready to fire up offsite in the event of disaster as well, if so I will have to figure out how they would automatically update from the latest backup before being started – but that is for a different project.
I’m going to start building the site.yaml tonight or tomorrow and document everything here afterwards. If I can get everything up and running the way I want I will post a cleaned up version of how I did afterwards. If not I will at least try to figure out where I went wrong so you can use it as a guide of what not to do.