Highly transactional databases in the cloud face significant challenges, particularly latency issues caused by network-based storage solutions like AWS FSx. These databases require low-latency, high-performance storage to efficiently process large volumes of reads and writes, but network-based options introduce bottlenecks that can degrade performance. When data must travel across availability zones or network hops, applications can experience slowdowns, timeouts, and even downtime, impacting business continuity. To address these issues, locally attached storage, such as Elastic Block Store (EBS) volumes, can provide a more reliable and cost-effective alternative.
Dave Bermingham, Senior Technical Evangelist at SIOS Technology, explains that network-based storage solutions are often unsuitable for workloads where even millisecond delays can affect performance. Unpredictable storage latency can lead to inefficient database transactions, negatively impacting user experience. “If the backend database can’t process transactions quickly, it can lead to application timeouts, bad user experiences, and ultimately downtime,” Bermingham explains. While applications may appear to be running, if storage cannot keep up, businesses face hidden downtime that can disrupt operations.
📹 Going on record for 2026? We're recording the TFiR Prediction Series through mid-February. If you have a bold take on where AI Infrastructure, Cloud Native, or Enterprise IT is heading—we want to hear it. [Reserve your slot
Bermingham highlights how SIOS DataKeeper addresses these challenges by eliminating network-based storage dependencies. The solution enables organizations to use locally attached EBS volumes within AWS, ensuring that storage remains co-located with the active server. This setup removes the added latency of network-based storage while maintaining high availability (HA). By guaranteeing that storage and compute resources are always in the same availability zone or placement group, SIOS Technology prevents the inconsistencies that often arise with cloud-based databases.
Another key factor in database performance is placement consistency. When workloads are spread across multiple zones without proper placement controls, unpredictable delays occur as data moves between locations. SIOS DataKeeper ensures that compute and storage resources remain together, providing a stable and high-performance environment for transactional workloads.
In addition to performance improvements, SIOS DataKeeper offers cost savings by eliminating the need for expensive network-attached storage. Bermingham pointed out that high-performance tiers in FSX can be costly, whereas locally attached EBS volumes provide a more affordable alternative. This allows businesses to optimize their cloud spending while maintaining the high availability (HA) required for mission-critical applications.
SIOS Technology also enhances disaster recovery (DR) by synchronizing data in real time across availability zones and asynchronously between regions. Bermingham emphasizes that this approach ensures minimal downtime and near-instant recovery in the event of a regional failure. By keeping storage and compute resources consistently aligned, SIOS DataKeeper provides high availability (HA) and disaster recovery (DR) in cloud environments.
Guest: Dave Bermingham (LinkedIn)
Company: SIOS Technology
Show: Data Driven
This summary was written by Emily Nicholls.





