Your company maintains a hybrid deployment with GCP, where analytics are performed on your anonymized customer data. The data are imported to Cloud Storage from your data center through parallel…

Question

Options

• AIncrease the CPU size on your server.
• BIncrease the size of the Google Persistent Disk on your server.
• CIncrease your network bandwidth from your datacenter to GCP.
• DIncrease your network bandwidth from Compute Engine to Cloud Storage.

• A
  10% (2)
• B
  5% (1)
• C
  85% (17)

Explanation

Increasing network bandwidth from your datacenter to GCP (C) directly addresses the bottleneck: data must physically traverse that WAN link before it ever reaches GCP, so no amount of optimization inside GCP can compensate for a saturated or slow ingress connection. A is wrong because CPU is not the limiting factor in a bulk transfer workflow - network I/O is, and adding CPU cycles won't move bits faster. B is wrong because the server's persistent disk size is irrelevant; the data flows through the transfer server to Cloud Storage, and disk size doesn't affect throughput (disk speed might, but that's not what's offered). D is wrong because the Compute Engine-to-Cloud Storage link is on Google's internal network, which is already extremely high-bandwidth and is rarely the bottleneck compared to the external WAN connection from your datacenter.

Memory tip: Visualize the data's full journey - datacenter → WAN → GCP edge → Cloud Storage. The slowest segment sets the ceiling for the entire pipeline, and in hybrid scenarios the WAN link between on-prem and cloud is almost always that segment.

No community discussion yet for this question.