File Transfer
High-Speed File Transfer
Introducing INNORIX's high-speed file transfer, which connects high-speed networks to actual file transfer performance.
REAL THROUGHPUT
Files move as fast as your line speed.
Building a high-speed network doesn't automatically make file transfer just as fast.
Even with bandwidth like 10Gbps or 40Gbps, actual transfer throughput varies depending on latency, RTT, packet loss, file profile, storage I/O, concurrent traffic, and the transfer method.
Rather than relying on the characteristics of a single protocol, INNORIX high-speed file transfer uses parallel transfer, dynamic concurrency, and long-distance optimization to put the available capacity of your existing network to work in actual file transfer.
TRANSFER CONDITIONS
Connects line speed to actual file throughput.
In high-speed file transfer, the important number isn't just the maximum bandwidth shown on the network interface.
Even with the same 10Gbps line, actual file transfer performance can differ between a short distance inside a data center and a WAN spanning continents.
INNORIX manages transfer performance based on the entire data path the file actually travels from source to target, not just a single link speed.
TRANSFER METHODS
Uses network capacity according to the transfer method.
TCP and UDP are often compared when explaining high-speed transfer.
UDP-based acceleration methods implement reliability, congestion control, and rate control at the application layer, which can deliver high performance over long-distance networks.
Approaches in the Parallel TCP and High-Speed TCP family can increase throughput through multiple connections and transfer optimization while still using the existing TCP network.
Network
- UDP-Based Acceleration
- Parallel TCP
- High-Speed TCP
- WAN Optimization
Rather than assuming any single method is always superior in every environment, it's important to choose a transfer structure suited to the actual network architecture, security policy, operating environment, and required throughput.
INNORIX scales performance within a transfer architecture that can operate alongside your existing network environment.
EXISTING NETWORK
Configures high-speed transfer on top of your existing network.
INNORIX does not presuppose UDP-based data transfer.
It applies parallel transfer and transfer optimization while using your existing TCP-based network and security environment.
Existing Enterprise Network
Rather than changing the network architecture itself into a separate transfer-dedicated structure to adopt high-speed file transfer, it's configured to raise transfer performance between the source and target that are already connected.
PARALLEL SESSIONS
Expands a single connection into multiple transfer paths.
In high-speed, long-distance networks, a single connection can sometimes struggle to fully utilize the available bandwidth.
INNORIX uses parallel sessions to process data across multiple transfer paths.
Source
Target
Through parallel sessions and dynamic concurrency, transfers are handled according to source, target, and network conditions.
The goal is not simply to increase the number of connections, but to expand the capacity of the current network that can actually be used as file throughput.
CONCURRENCY
Adjusts concurrency according to transfer conditions.
Creating more sessions doesn't always lead to higher performance.
This is because source storage, target storage, the network, and other workloads are all affected at the same time.
Rather than using a single fixed concurrency value, INNORIX operates multiple transfers according to the capacity available in the actual transfer environment.
LONG-DISTANCE
Makes use of available bandwidth even over long-distance networks.
RTT increases as distance grows.
Even on networks with the same bandwidth, a transfer within a data center in Seoul and a transfer from Seoul to New York are not under the same conditions.
LOCAL
Server A
Server B
LONG DISTANCE
Seoul
New York
In long-distance transfer, RTT, packet loss, and connection behavior determine actual throughput along with bandwidth.
INNORIX uses parallel transfer and transfer optimization designed for long-distance environments to put available WAN capacity to work in file transfer.
PACKET LOSS
Continues transfer even in packet-loss environments.
In real WAN and internet environments, packet loss is never completely constant.
When loss occurs, the transport and transfer layers must handle it while continuing data movement.
Network Condition
In high-speed transfer, what matters isn't just measuring peak speed under ideal network conditions, but how stably the actual transfer maintains throughput and completes while network conditions change.
INNORIX manages transfer state and recovery together, keeping data movement going even in environments where network conditions change.
END-TO-END PATH
Views the network and storage as a single performance path.
Having a 10Gbps network doesn't mean the source can read data at a speed equivalent to 10Gbps or that the target can write at the same speed.
Actual file transfer throughput is limited by the slowest segment.
Source Read → Transfer Processing → Network → Target Write
Therefore, rather than explaining the performance of high-speed file transfer with a single network number, it should be verified by the actual throughput of the end-to-end data path, including source and target storage.
FILE PROFILE
Applies the transfer method according to the file profile.
Even on the same network, performance characteristics differ between transferring a single file and transferring millions of files.
For large files, network and storage throughput matter most, while for high-volume transfers, file discovery, open/close operations, metadata, and queue processing also have an impact.
INNORIX considers both file size and file count as part of actual transfer performance.
MULTI-NETWORK
Uses multiple networks as a single transfer capacity.
In environments where the source and target have multiple network interfaces or available network paths, multi-network transfer can be configured.
Source
Target
Rather than using just one network path, the capacity of multiple networks can be used for transfer to expand overall file throughput.
The actual implementation is configured according to the source, target, and network configuration.
SHARED NETWORK
Operates transfer capacity even on a shared network.
An enterprise network is not dedicated infrastructure for file transfer alone.
Applications, databases, user traffic, and other services all share the same network.
Network
Queue / Concurrency / Capacity
INNORIX manages multiple transfers using parallel transfer, dynamic concurrency, and queuing, and operates available transfer capacity within the existing network environment.
Rather than isolating high-speed transfer as a separate network island, it is connected as part of the enterprise infrastructure.
RECOVERY
Manages speed and recovery within a single transfer.
If a high-speed transfer fails midway and the entire file has to be sent again from the start, the actual time to complete the task increases significantly.
Especially when moving files of hundreds of gigabytes or terabytes over long distances, recovery becomes as important a performance factor as throughput.
Large Transfer
INNORIX applies resume, retry, and recovery together with high-speed transfer, allowing transfer to continue based on the data already delivered.
Rather than treating speed and stability as separate features, they are managed together as part of the actual completion time.
SHARED CAPACITY
Manages the speed of multiple transfers within overall capacity.
In enterprise environments, it's more common for multiple transfers to run at the same time than for a single file to use the network alone.
Shared Capacity
Rather than just maximizing the peak speed of a single transfer, what matters is how the overall capacity is used when multiple transfers run at the same time.
PLATFORM INTEGRATION
Connects high-speed transfer to your existing infrastructure.
If high-speed file transfer is isolated as a separate, dedicated transfer environment, it has to be reconnected every time a new data movement arises.
With INNORIX, high-speed transfer can be used between servers, cloud, object storage, and other endpoints.
Source
INNORIX
Rather than using high-speed transfer only as a specific point-to-point accelerator, it is applied as a common transfer capability available across the entire INNORIX platform.
DATA PATHS
Uses the same performance model across various data paths.
Even when the types of source and target differ, the way actual end-to-end transfer performance is verified based on bandwidth, RTT, loss, storage I/O, and file profile remains the same.
TEST CONDITIONS
Verifies transfer performance under real conditions.
It's difficult to judge real-world transfer performance from a single number like 'up to 10Gbps' alone.
This is because results for the same product can vary significantly depending on test conditions.
Therefore, it's important to verify high-speed transfer performance by presenting bandwidth together with RTT, packet loss, file profile, storage, and concurrency conditions.
LINE UTILIZATION
Checks actual line utilization alongside peak speed.
The purpose of high-speed transfer is not to display the biggest number, but to put the network capacity customers have already built to work in actual data movement.
Transfer Efficiency
Therefore, when checking performance, it's important to look not only at maximum throughput but also at how much of the capacity the actual transfer used in the given network environment.
This perspective applies equally across networks of different scales.
AUTOMATION
Connects high-speed transfer to automation and flow.
Fast transfer doesn't have to be a separate task that an operator must run manually every time.
High-speed transfer can be started by a schedule, file event, API, or external request, and the completion result can be connected to the next task.
High-speed transfer and automated file transfer are operated within the same flow model.
PERFORMANCE LAYER
Applies the same performance capability to large files and high-volume transfers.
High-speed transfer is not an independent file type, but a performance capability needed across many kinds of data movement.
Therefore, high-speed file transfer is not an isolated island of speed separated from other transfer products, but expands into a common performance layer applied to large files, high-volume transfers, servers, storage, and AI data delivery.
MONITORING
Checks transfer performance and results on the same platform.
Even with high-speed transfer, operators need to check not just speed but also whether the transfer completed successfully.
RUN-1842
Progress is checked in Runs and Monitoring, and the final result is confirmed through file status and receipts.
Performance, recovery, and completion results are connected within a single transfer operation.
TEST ENVIRONMENT
10Gbps isn't the definition of high speed — it's one environment to verify.
High-speed file transfer isn't a product that begins at a specific network speed.
Performance improvement matters even in environments where a 1Gbps line's actual transfer stays at only tens of Mbps, and on networks of 10Gbps or above, what matters is how much of the already-built capacity can be put to use in actual file transfer.
10Gbps is not the definition of high-speed transfer, but a test environment for verifying how much of the actual network capacity can be converted into transfer throughput.
GET STARTED
Connect your network speed to actual file throughput.
It considers bandwidth, RTT, packet loss, storage I/O, and file profile together, and applies parallel transfer, dynamic concurrency, and long-distance optimization to put your existing network's available capacity to work in actual data movement.
The same high-speed transfer capability is applied across large files, high-volume transfers, servers, object storage, and AI data delivery, with recovery and final results managed within a single platform.
Considering high-speed file transfer?
We'll help you review a high-speed file transfer setup that fits your current network and transfer environment.
- Review of long-distance and global transfer environments
- Performance configuration suited to your network
- Guidance on applying it to your existing infrastructure