Prepare and test your skills
Prepare and test your skills
Assessing transfer tools for specific use cases means matching data size, available network, and transfer frequency to the right Google Cloud path. Storage Transfer Service (opens in a new tab) is an online, fully managed service that moves data over the network, while Transfer Appliance (opens in a new tab) is physical hardware that is filled locally and shipped to Google. Both tools deliver data to Cloud Storage (opens in a new tab), but online transfers depend on bandwidth and time while offline transfers avoid loading the business network.
Storage Transfer Service is a fully managed online service that moves data between object and file storage systems without custom code or managed infrastructure. Data flows from the source system over the network to a Cloud Storage bucket, and integrity validation confirms the data was not corrupted in transit. The service uses a highly parallelized architecture with automatic retries and load balancing, and it is optimized for transfers larger than 1 TiB up to petabytes.
Storage Transfer Service supports one-time migrations and repeated transfers such as backup copies and data processing pipelines. Teams can choose Google-managed transfers for simplicity or self-hosted agents for granular control over network routing and bandwidth consumption. Transfers are managed through the Google Cloud console, REST APIs, the gcloud command-line tool, and Java and Python client libraries.
The service supports movement between these source and destination combinations:
Common uses include migrating from another cloud or data center, replicating a bucket to another region for backup, feeding analytics in BigQuery or Managed Service for Apache Spark, and moving cold data to Cloud Storage for archival.
Transfer Appliance is a hardware appliance for offline migration of very large volumes of data to Google Cloud. Data flows from on-premises storage to the appliance during local copy, then the appliance is shipped to Google where data is ingested to Cloud Storage. Data does not travel over the business internet connection during capture, so production work is not disrupted.
Transfer Appliance is recommended when data exceeds 20 TB and scales up to 1 petabyte. It fits one-time, large-scale migrations where an online transfer would take too long or overload the network during peak work hours. For reference, 1 GB ideally takes eight seconds over a 1 Gbps network, but 100 TB takes about 12 days before accounting for management time and bandwidth efficiency. Planning must include product cost, personnel and logistics, plus Cloud Storage storage and operation costs after arrival.
The transfer decision depends on cost, transfer time, and whether an online or offline path fits the constraints. Ingress to Cloud Storage is free, but transfers from another public cloud can incur egress and read-operation charges, and a private data center may need added bandwidth to Google Cloud. Time planning must account for dataset size, usable bandwidth, and whether transfers must continue on an ongoing basis.
Key planning factors for time and cost include:
The two tools differ across transfer method, data size, frequency, and network dependence:
| Tool | Transfer method | Ideal data size | Transfer frequency |
|---|---|---|---|
| Storage Transfer Service | Online over the network, fully managed with TLS 1.3 encryption in transit | Optimized for more than 1 TiB, up to petabytes | One-time and repeated or automated transfers, including backup and pipelines |
| Transfer Appliance | Offline with shipped hardware, copied locally then ingested to Cloud Storage | Recommended for more than 20 TB up to 1 petabyte | Mainly one-time bulk migrations with limited network or tight time limits |
Choose Storage Transfer Service when network capacity is sufficient and transfers need automation, repeatability, or movement between clouds. Choose Transfer Appliance when the dataset is extremely large, bandwidth is limited, or an online move would disrupt mission-critical work.
Exam tip: Storage Transfer Service is optimized for transfers over 1 TiB, while Transfer Appliance is recommended for offline moves over 20 TB up to 1 petabyte.
Evaluating transfer feasibility using bandwidth and time constraints means checking whether total data volume can move over usable network bandwidth before the project deadline. Storage Transfer Service (opens in a new tab) moves on-premises Network File System data to Cloud Storage over the network, while Transfer Appliance (opens in a new tab) copies data to shipped hardware without using outbound bandwidth. The comparison of volume, speed, and time decides whether an online transfer to Cloud Storage (opens in a new tab) is efficient or an offline move is needed.
Estimated online transfer time is determined by the relationship between dataset size and usable network bandwidth. Ideally, 1 GB transfers in eight seconds over a 1 Gbps network, but the same math scales quickly so a very large dataset can take days or months. A realistic estimate must include efficiency and overhead because raw division overstates usable speed.
Estimate transfer duration using these four factors:
Large transfers can also overload a company network during peak work hours and block mission-critical work. Feasibility therefore includes when the transfer can run and whether production traffic shares the same public internet connection.
The online versus offline transfer choice compares network transfer time against the project deadline and the cost of adding bandwidth. Online transfer moves data from a private data center to Cloud Storage over a public internet connection, Direct Peering, or Cloud Interconnect. Offline transfer copies data locally to a Transfer Appliance and ships the appliance to Google for upload, without consuming outbound network bandwidth during capture.
A decision tree that branches on source location, data size, and whether bandwidth meets the deadline to choose gcloud storage, online Storage Transfer Service, or offline Transfer Appliance into Cloud Storage.
The same data volume changes dramatically with bandwidth, as shown below:
| Data volume | Network bandwidth | Estimated online time |
|---|---|---|
| 100 TB | 1 Gbps | Over 10 days, about 12 days with overhead |
| 100 TB | 100 Mbps | Over 100 days |
| 300 TB | 100 Mbps | About 9 months |
With Transfer Appliance, 300 TB can be captured in under 25 days plus about 10 more business days before the data is accessible in Cloud Storage. Transfer Appliance is generally a good fit when uploading over the network would take more than one week.
Exam tip: Choose online transfer when 100 TB at 1 Gbps in about 10 to 12 days meets the deadline, but choose Transfer Appliance when 100 Mbps would stretch the same transfer past 100 days.
Transfer Appliance is the offline hardware option for large-scale transfers when fast network connectivity is unavailable or too costly to acquire in time. Storage Transfer Service for on-premises data is designed for up to petabytes of data, but it requires sufficient available bandwidth to meet the deadline. When bandwidth cannot meet the deadline, the decision shifts from Storage Transfer Service to Transfer Appliance even though appliance latency is typically higher.
Transfer Appliance is especially useful in these situations:
Using Transfer Appliance requires requesting one or more appliances in the Google Cloud console, copying data during the allowed data capture period, and shipping the hardware back for upload. Transfer Appliance is available only in certain countries, so location availability and shipping ability are also part of feasibility.
Analyzing the features and benefits of GCP transfer tools means comparing how online and offline paths move data into Cloud Storage without transforming it. Storage Transfer Service (opens in a new tab) is an online, fully managed service that moves data over a network, while Transfer Appliance (opens in a new tab) is Google-owned hardware that is filled locally and shipped back for upload. The choice depends on data volume, available bandwidth, and how quickly data must arrive in Cloud Storage (opens in a new tab).
Storage Transfer Service is a fully managed, highly scalable service that automates transfers into Cloud Storage from on-premises file systems, other public clouds, and between Cloud Storage buckets. For on-premises data, the service moves files from Network File System (NFS) storage to a destination bucket and supports up to petabytes of data and billions of files with full copies or incremental copies. For cloud sources, the service transfers data from Amazon S3 and Microsoft Azure Blob Storage to Cloud Storage, and it can also copy data from one Cloud Storage bucket to another or from a list of public URLs.
For on-premises transfers, data flows from the NFS file system through transfer agents to the destination Cloud Storage bucket. Agent software runs in a Docker container on computers in the data center that can see the NFS data, and the agents collectively move data in parallel to Cloud Storage. A transfer starts in the Google Cloud console with a source directory, a destination bucket, and a time or schedule, and the service recursively crawls subdirectories and preserves names so /dir/foo/file.txt becomes an object named /dir/foo/file.txt. The transfer depends on identical agent setup, because all agents must see the same NFS mounts with the same relative paths for the product to function.
Storage Transfer Service is built for speed, control, and reliability on large online transfers. More agents result in more speed because work is automatically parallelized across all agents to use available bandwidth, while a user-specified bandwidth cap limits transfer traffic so other data center workloads can meet their service objectives. The service automatically retries transient errors, lets teams monitor file count and transfer speed in the console, and generates a tab-delimited file (TSV) with a full record of files touched and error messages when the transfer completes. Agents are fault tolerant and self-updating and self-healing, so the transfer continues with remaining agents if one agent goes down and manual patching or restarts are not needed.
Storage Transfer Service is especially useful when bandwidth is sufficient and managed operation is needed:
Long-running transfers that take days or weeks depend on Cloud Monitoring for agent health and throughput. The service exposes agent health and throughput to Monitoring so alerts can notify teams when agents are down or need attention, which prevents slowdowns that delay the project timeline.
Transfer Appliance is a high-capacity, Google-owned storage device for offline transfer when a fast network connection is unavailable and acquiring more bandwidth is too costly or too slow. The appliance sits on the customer side of the network boundary: it is connected to the local network, data is copied directly to it, and then the hardware is shipped back to a Google upload facility where Google copies the data to the Cloud Storage bucket and wipes the device for its next use. Data flows from local storage to the appliance to Cloud Storage without consuming outbound network bandwidth during capture.
Transfer Appliance provides optimal efficiency when online transfer cannot meet time or connectivity constraints. It is especially useful when the data center is in a remote location with limited or no bandwidth, when bandwidth exists but cannot be acquired in time to meet the deadline, and when logistical resources exist to receive, connect, and return the hardware. The tradeoff is higher overall latency than online transfer, a requirement to handle shipping, and availability in only certain countries, with EU appliances shipped from and returned to Belgium so data destined for an EU region can remain within the European Union.
Performance math drives the decision between online and offline transfer. With reasonable connectivity such as 1 Gbps, transferring 100 TB online takes over 10 days, which may be acceptable, while the same 100 TB over a 100 Mbps connection takes over 100 days and favors Transfer Appliance. With a typical 100 Mbps connection, 300 TB takes about 9 months to upload online, while Transfer Appliance can capture 300 TB in under 25 days plus about 10 business days for Google to make the data accessible in Cloud Storage.
Acquiring and using a Transfer Appliance follows an ordered offline workflow:
Transfer Appliance can also run in online mode, where data copied to the appliance streams directly to the Cloud Storage bucket instead of waiting for return shipping. In online mode, data flows from the local network to the appliance to Cloud Storage with low latency, multiple appliances can connect in parallel, the connection is encrypted end-to-end, and data is removed from the appliance after the transfer completes.
Choosing between the transfer tools depends on where data starts, how much data must move, and whether network bandwidth can meet the deadline. Transfers of less than 1 TB from on-premises or between Cloud Storage regions typically use the gcloud storage command, while transfers of more than 1 TB use Storage Transfer Service or Transfer Appliance. Cloud-to-cloud moves and large bucket-to-bucket moves use Storage Transfer Service, because Google provisions the connection between cloud data centers with no network setup required.
| Tool | Best for | Network requirement |
|---|---|---|
gcloud storage command | Small to medium transfers under 1 TB, as-needed command-line copies, streaming program output, and syncing a moderate directory with low latency | Requires enough enterprise network bandwidth to meet the deadline |
| Storage Transfer Service | Large online transfers over 1 TB up to petabytes and billions of files, cloud-to-cloud moves, bucket-to-bucket moves, scheduled and recurring copies, and teams needing a console, bandwidth caps, and detailed error records | Requires sufficient available bandwidth; traffic flows over public internet, Direct Peering, or Cloud Interconnect |
| Transfer Appliance | Very large offline transfers where bandwidth is missing, too slow, or too costly to acquire in time, including remote sites | Works with poor or no internet connectivity; physical logistics are needed to receive, connect, and return hardware |
Cost, time, and security tradeoffs apply to all options. Ingress to Cloud Storage is free, but transfers from another public cloud can incur egress and read-operation charges, private data centers may need to pay for added bandwidth, and Transfer Appliance has product and personnel costs for handling hardware. Large transfers should avoid peak work hours if they share production bandwidth, and transfers over the public internet should be checked with a security administrator for company policy.
Exam tip: Choose Transfer Appliance when on-premises data exceeds 1 TB and bandwidth cannot meet the deadline, especially when network upload would take more than one week; choose Storage Transfer Service when bandwidth is sufficient for large, scheduled, or cloud-to-cloud transfers.
gcloud storage command is typically used for transfers under 1 TB, while larger online or cloud-to-cloud transfers use Storage Transfer Service.Choose Storage Transfer Service when network capacity is sufficient and transfers need automation, repeatability, or movement between clouds, and choose Transfer Appliance when the dataset is extremely large, bandwidth is limited, or an online move would disrupt mission-critical work. Storage Transfer Service is optimized for more than 1 TiB up to petabytes and supports one-time and repeated or automated transfers, while Transfer Appliance is recommended for more than 20 TB up to 1 petabyte and is mainly for one-time bulk migrations with limited network or tight time limits.
Estimated online transfer time is determined by dataset size, available bandwidth, bandwidth efficiency, and management time for setup, monitoring, and retries. Transferring 100 TB takes about 12 days at 1 Gbps but over 100 days at 100 Mbps, while 300 TB at 100 Mbps takes about 9 months online versus under 25 days of capture plus about 10 business days with Transfer Appliance. Transfer Appliance is generally a good fit when uploading over the network would take more than one week.
Storage Transfer Service moves on-premises Network File System (NFS) data to Cloud Storage through transfer agents that run in a Docker container on computers that can see the NFS data and collectively move data in parallel to the destination bucket. All agents must see the same NFS mounts with the same relative paths for the product to function, more agents result in more speed because work is automatically parallelized, a user-specified bandwidth cap can limit transfer traffic, and the service automatically retries transient errors and generates a tab-delimited file with a full record of files touched and error messages.
An organization has 300 TB of historical archive data stored on local Network Attached Storage (NAS) devices in their on-premises data center. The data center has a limited outbound internet connection of 100 Mbps, and uploading the data over the network would take several months to complete.
Which Google Cloud transfer solution should the organization use to perform this one-time migration to Cloud Storage within a few days?
Cloud Data Fusion
BigQuery Data Transfer Service
Storage Transfer Service
Transfer Appliance
Last updated