Deploying IRI in a carrier grade fashion
A Proof-of-Conecpt, deploying IRI in a carrier grade fashion. Having a dynamically scaling infrastructure requires some planning and automation. How do we ensure our deployment is stable, scalable and robost?
This blog touches the subject of deploying IRI behind highly available load balancers using service discovery and health checks. In addition, it includes some examples a user can test to better comprehend the concept.
In addition, I included a link to the repository where I am currently working on a complete solution for private clusters wanting to use a highly available load balancer solution.
This tutorial will help you getting started with a clustered load-balancers:
- Dockerized IRI version + monitoring
- HAProxy with Consul as dynamic backend
- High availability of the load balancers
- lbreg: tool to help manage load balancer backends
- Multiple nodes deployment
- Apply for and configure SSL certificate for multiple nodes at ones
Update: glad to announce that the the load balancer feature has matured from a PoC to a working solution. Consul+Haproxy has been added to the "dockerized" iri-playbook. See links section on the left for the load balancer feature in the IRI Playbook.