COA COA High Availability & Fault Tolerance 2

Question 1

How many controller nodes are typically recommended as the minimum for quorum-based OpenStack HA?

Accepted Answer

3

Answer

Three controller nodes are the minimum recommended for quorum, preventing split-brain scenarios in Galera and Pacemaker clusters.

Question 2

What happens to a running instance during a successful OpenStack live migration?

Accepted Answer

The instance continues running with minimal downtime

Answer

Live migration transfers a running instance's memory and CPU state to another host while keeping the instance operational throughout the process.

Question 3

Which OpenStack alarming service evaluates metric thresholds and triggers actions for automated fault response?

Accepted Answer

Aodh

Answer

Aodh is OpenStack's alarming service that evaluates metric thresholds from Gnocchi and triggers notifications or Heat scaling actions.

Question 4

What is the purpose of a floating IP in an OpenStack high availability architecture?

Accepted Answer

Enabling a public IP that can be reassigned to a standby instance during failover

Answer

Floating IPs can be disassociated from a failed instance and quickly reassigned to a standby instance, maintaining external service availability.

Question 5

Which HAProxy directive configures HTTP health checks against backend OpenStack API nodes?

Accepted Answer

option httpchk

Answer

The `option httpchk` directive tells HAProxy to send HTTP requests to backend OpenStack API nodes to detect unhealthy backends.

(COA) Certified OpenStack Administrator Exam Practice Test