A global gaming enterprise is preparing to migrate its core infrastructure to Google Cloud… — Google Cloud Professional Cloud Architect

A global gaming enterprise is preparing to migrate its core infrastructure to Google Cloud within a 3-week target window. The legacy environment consists of two primary datasets located in a single on-premises data center:

1. A $2.4\text{ PB}$ static archive of telemetry and analytics logs stored on an S3-compatible object storage cluster.
2. A $12\text{ TB}$ operational transactional MySQL database supporting live player accounts, which allows a maximum maintenance downtime of 4 hours during final cutover.

The enterprise has a Dedicated Interconnect connection to Google Cloud with an unallocated available bandwidth of $300\text{ Mbps}$ reserved for migration operations. Security governance strictly mandates that data transfer operations must comply with perimeter-based exfiltration controls using VPC Service Controls.

Which migration and data transfer strategy should you recommend to meet the business and technical requirements?

Order and deploy Google Cloud Transfer Appliance devices to physically ingest and transfer the $2.4\text{ PB}$ static archive into Cloud Storage, while utilizing Database Migration Service (DMS) with continuous Change Data Capture (CDC) over the Cloud Interconnect connection for the $12\text{ TB}$ MySQL database.Answer
B
Configure Storage Transfer Service on-premises agents to stream both the $2.4\text{ PB}$ static archive and the $12\text{ TB}$ MySQL database simultaneously over the $300\text{ Mbps}$ Dedicated Interconnect connection using maximum parallel transfer threads.
C
Migrate the $12\text{ TB}$ MySQL database directly to Cloud Spanner by backing up to Transfer Appliance devices, while establishing high-bandwidth HA VPN tunnels to stream the $2.4\text{ PB}$ archive directly into BigQuery.
D
Provision an array of HA VPN tunnels over the public internet to transfer the $2.4\text{ PB}$ archive using Storage Transfer Service, and perform a manual `mysqldump` export and restore of the $12\text{ TB}$ MySQL database during the 4-hour cutover window.

Answer

Deploy Google Cloud Transfer Appliance devices for the

2.4\text{ PB}

static storage archive and use Database Migration Service (DMS) with continuous Change Data Capture (CDC) over Cloud Interconnect for the

12\text{ TB}

MySQL database.

Transferring

2.4\text{ PB}

of data over a

300\text{ Mbps}

network link would take over two years, making offline transfer via Transfer Appliance mandatory to meet the 3-week timeframe. For the

12\text{ TB}

MySQL database, continuous replication using Database Migration Service (DMS) with Change Data Capture (CDC) over Cloud Interconnect ensures the database remains synchronized online, enabling final promotion and cutover well within the 4-hour downtime limit.

Step-by-Step Solution

Calculate network throughput limitations for the static archive dataset

At an available bandwidth of

300\text{ Mbps}

(

37.5\text{ MB/s}

), transferring

2.4\text{ PB}

(

2,400,000\text{ GB}

) online would require

\frac{2,400,000\times 10^9\text{ bytes}}{37.5\times 10^6\text{ bytes/s}} \approx 64,000,000\text{ seconds} \approx 740\text{ days}

An online network transfer cannot satisfy the 3-week migration timeline requirement.

Select physical offline transfer mechanism for petabyte-scale storage

Utilizing Google Cloud Transfer Appliance enables offline physical transport of

2.4\text{ PB}

to Cloud Storage within the 3-week window.

Transfer Appliance bypasses bandwidth bottlenecks for multi-petabyte datasets.

Select replication and cutover mechanism for the operational database

Configuring Database Migration Service (DMS) with continuous CDC continuously syncs transactions over the

300\text{ Mbps}

Interconnect link, allowing final cutover to take minutes, well under the 4-hour downtime threshold.

Offline dumps of

12\text{ TB}

would exceed the 4-hour downtime window during transfer and restore.

Key Concept

Selecting hybrid data transfer strategies based on bandwidth constraints, dataset scale, and operational cutover downtime limits.

Estimated Time:3m 0s

Question

Answer

Step-by-Step Solution

Key Concept

Common Mistakes

Topics