David Trossell, CEO and CTO of Bridgeworks speaks to Digitalisation World, “it’s possible to enable organisations to securely move massive datasets across thousands of miles at up to 98% of their existing bandwidth capacity, proving that data proximity is no longer a barrier to global scalability.”
October 14, 2026
Moving huge SAN-block and encrypted datasets across hybrid cloud environments remains hampered by WAN latency and packet loss, but AI-driven acceleration tools aim to bypass these physical bottlenecks without costly bandwidth upgrades.
Enterprises love the flexibility of NetApp, Veeam and IBM cloud environments; however, moving massive, high-definition media or encrypted SAN-block datasets across countries routinely hits a wall. This barrier is often called latency and packet loss, and it is particularly important to note that SAN-block datasets are highly sensitive to them. Most problematic is the fact that data growth is vastly outstripping the speed of network protocol evolution.
The traditional way of moving SAN-block datasets over a wide area network (WAN) often requires much careful planning. This includes ensuring that your network architecture meets specific performance and security requirements. Low latency is required by SAN protocols like iSCSI or Fibre Channel over IP; high levels of WAN latency or packet loss can cause iSCSI timeouts, data corruption, jitter and disconnected storage volumes, such as virtual disks and drives.
Manual intervention
Data could be lost and applications could crash. When data becomes unavailable, an administrator is often needed to manually remount or re-synchronise those volumes. Latency and packet loss also impact the heartbeat signal of storage systems. When WAN latency spikes severely, the response may take too long – leading to the system assuming that the network or the storage device is down. Subsequently, latency, jitter and packet loss can compromise business continuity, damage brand credibility during critical interactions, and often indicate deeper security or configuration vulnerabilities – many of which can be addressed.
Another issue is that, when it comes to traditional throughput optimisation, standard TCP/IP can’t handle high bandwidth WAN links well. The response to this predicament is to introduce WAN Optimisation Controllers, or specialised data mobility software that uses UDP-based acceleration protocols. While the ideal situation is to be able to encrypt data and to send it in transit, WAN Optimisation tools for deduplication and compression just can’t handle it. They often fail when they deal with compressed or already encrypted files.
However, some form of data encryption can be achieved with IPsec or TLS. Yet, if double-encryption is used, the CPU overhead increases on your storage processors or network appliances. The most common response might be asynchronous replication. With this method, the local SAN takes snapshots and replicates the encrypted blocks to the remote SAN on a schedule, or continuously in the background. This method allegedly tolerates latency well.
Pitfalls of synchronicity
Synchronous replication is an alternative option, but it should be avoided if the distance between nodes is greater than 60 miles apart: this is because latency increases as the distance does. To help with this, an ultra-low latency fibre connection is needed to receive an acknowledgement from the remote site for every write, which can degrade local application performance if the WAN is slow because of high levels of latency.
However, there is also storage-assisted migration, using tools such as VMWare vMotion, which, the company claims, has long-distance capabilities. There are also storage array native migration tools that can orchestrate any move while preserving block integrity. Then, in an era where artificial intelligence (AI) is often a bolted onas an afterthought, there are WAN Acceleration tools, such as WANrockIT and PORTrockIT. They deploy AI, machine learning and data parallelisation to expedite data across WANs – handling encrypted data without having to unlock and lock it at each end.
Their non-touch data handling also enables better security as not even the Bridgeworks staff can see the data. The technology doesn’t resolve latency and packet loss. Instead, they mitigate the effects of latency and packet loss. Plus, by integrating WANrockIT and PORTrockIT into environments powered by major storage, such as Commvault and NetApp, and database providers such as Oracle, enterprises bypass the hardware limitations of latency and packet loss.
This unlocks existing bandwidth capacity without undergoing any costly network infrastructure upgrades, while buying more bandwidth leads to a 200x paradox whereby more bandwidth doesn’t necessarily lead to a more performant pipe and faster data flows. In fact is can achieve the inverse.
‘Enterprises are trapped’
David Trossell, CEO and CTO of Bridgeworks, explains: “Enterprises are trapped in a 200x paradox where throwing more bandwidth at hybrid cloud storage fails to solve the physical limitations of WAN latency and packet loss. Traditional tools force a terrible compromise between data velocity and security – especially when handling sensitive SAN-block datasets and encrypted files.”
“By using AI and data parallelisation to mitigate these network limitations rather than fight them, it’s possible to enable organisations to securely move massive datasets across thousands of miles at up to 98% of their existing bandwidth capacity, proving that data proximity is no longer a barrier to global scalability.”
So, data needn’t be stored within the same proximity in order to address the effects of latency and packet loss. The use of AI and data parallelisation means the data can be stored in different countries, different regions and sent in between storage sites thousands of miles away. In CloudFest’s ‘The Cloud Latency and Packet Loss Guide: How to Fix Them?’, Trossell comments: “Data should not be situated close together, even when using the 3-2-1-1-0 approach. This will help to minimise downtime and mitigate any potential risk of data loss.”
Cloud storage bottlenecks
The 200x paradox demonstrates that resolving hybrid cloud storage bottlenecks requires strategic adaptation, rather than simply increasing bandwidth to combat latency and packet loss. The physical limits of hybrid cloud storage are determined by hardware capacities, network bandwidth constraints, and the laws of physics governing data transmission speed over distance. By leveraging WAN Acceleration and data parallelisation, enterprises can securely transfer encrypted SAN-block datasets across vast distances without compromising performance or security.
To address latency and packet loss, and to make the most of their existing bandwidth, organisations need to look beyond traditional tools by first of all auditing what they have and by comparing it by conducting a proof of concept to realise how any potential performance benefits could be attained with WAN Acceleration. It doesn’t necessarily replace traditional technologies, but completements them – helping to overcome the 200x paradox in a way that restores greater rapidity to today’s data flows without manual intervention – freeing up IT staff to focus more on their hybrid cloud imperatives.