Cloud foundry is an open source, platform as a service governed by the Cloud Foundry Foundation.It allows anyone to deploy application and make it available to the world in few minutes. Previous post we had discussed about PCF architecture for application staging. Now let’s focus on the architecture, that ensures high availability of application. PCF – there are four levels of availability.
- Availability Zone:
PCF distributed the application evenly across multiple availability zone. So even one zone is down, still application will be available from other zone.
- Bosh managed process – Cloud Controller Process
“Bosh” is an important component of PCF architecture for high availability. It ensures high availability in two ways
1. If “Cloud controller process” is down:
The “Monit” component restarts the “Cloud Controller Process”. Bosh agent sends the restart event to the “Bosh/ ops Manager” for investigation and notification.
2. If Cloud Controller VM is down:
The Bosh agent running in the VM , sends health report of the VM to “Bosh” constantly. If the “health Monitor” component of “Bosh” does not receive health report from Bosh agent, it asks “Bosh Director” to restart the VM.
- Self healing
Apart from that PCF Diego has self healing capability. “Rep” constantly reports the application status of a cell to BBS. BBS passes it to “Brain”. “Converger” component of the “Brain” checks the actual state vs desired state. If it does not match, Brain throws the cell and goes for a new auction to select cell and stage application.
With these four levels of availability , PCF ensures high availability of the application deployed in PCF.