The goal is to deploy a new version without increasing the number of instances (minimizing cost) and ensuring each new instance is healthy before proceeding. Managed Instance Groups (MIGs) use maxSurge and maxUnavailable to control rolling updates.

• maxSurge: Determines the maximum number of new instances created above the desired count. Setting it to 0 ensures no extra instances are added, keeping costs low.
• maxUnavailable: Specifies the maximum number of instances that can be temporarily unavailable during the update. Setting it to 1 allows one instance to be replaced at a time.

With maxSurge=0 and maxUnavailable=1, the MIG replaces instances one-by-one: it first stops an old instance (making it unavailable), then creates a new one in its place. The deployment waits for the new instance to be healthy before proceeding to the next replacement. This ensures no instance count increase and health checks are enforced.

Options C and D are invalid as maxHealthy/maxUnhealthy are not valid parameters. Option A uses maxSurge=1, which increases instance count temporarily, violating the cost constraint.