PROFESSIONAL-CLOUD-ARCHITECT · Question #197
You need to deploy an application to Google Cloud. The application receives traffic via TCP and reads and writes data to the filesystem. The application does not support horizontal scaling. The…
The correct answer is D. Use an unmanaged instance group with an active and standby instance in different zones, use a. Option D is correct because the application's constraints - no horizontal scaling and exclusive filesystem access - demand exactly one active instance at a time. An unmanaged instance group with an active/standby pair in different zones achieves zone-level redundancy while…
Question
Options
- AUse a managed instance group with instances in multiple zones, use Cloud Filestore, and use an
- BUse a managed instance group with instances in multiple zones, use Cloud Filestore, and use a
- CUse an unmanaged instance group with an active and standby instance in different zones, use a
- DUse an unmanaged instance group with an active and standby instance in different zones, use a
How the community answered
(66 responses)- A6% (4)
- B12% (8)
- C21% (14)
- D61% (40)
Explanation
Option D is correct because the application's constraints - no horizontal scaling and exclusive filesystem access - demand exactly one active instance at a time. An unmanaged instance group with an active/standby pair in different zones achieves zone-level redundancy while ensuring only one instance owns the filesystem at any moment. The storage type in D (a regional persistent disk) can be detached from the failed instance and reattached to the standby, maintaining exclusive read/write access and preventing corruption. This architecture also aligns with the business accepting brief downtime during failover.
Options A and B are wrong because managed instance groups can automatically spin up additional instances (e.g., during auto-healing), which would result in concurrent filesystem access and data corruption - exactly the failure mode the requirements forbid.
Option C is wrong because it uses Cloud Filestore, which is an NFS-based shared filesystem accessible by multiple instances simultaneously; even with an active/standby setup, the risk of concurrent access during a failover transition remains and violates the exclusivity requirement.
Memory tip: Map the constraints to the architecture - "no horizontal scaling + exclusive filesystem = active/standby with a single attachable disk, not a shared NFS." When you see "concurrent access causes corruption," immediately eliminate any shared-filesystem (Filestore/NFS) options.
Topics
Community Discussion
No community discussion yet for this question.