What are the Best Practices for Ensuring Cloud Resilience?

Cloud resilience has become a strategic priority for market operators. As exchanges, central counterparty clearing houses (CCPs), and central securities depositories (CSDs) modernize critical systems, cloud infrastructure can improve scalability, flexibility, data management, and innovation.

These themes are explored in a new white paper from Nasdaq. For financial market infrastructures, resilience comes first, and cloud transformation is as much an operating-model and governance issue as a technology one. The same source argues that resilience cannot be assumed from the platform alone. It must be built into architecture, workflows, controls, readiness, and organizational decision-making.

This article summarizes those lessons on resiliency best practices (informed by Nasdaq experiences) and looks at the practical framework for market operators designing, assessing and improving cloud resilience.
 

Cloud resilience is the ability to keep critical services predictable, recoverable and governable under stress, disruption or change. In the white paper, resilience is defined as broader than availability alone. It includes control, recoverability, observability, governance, evidence, security and compliance at scale. For market operators, that matters because core workloads underpin market continuity and broader financial stability, which raises the bar for resilient design and operating discipline. 

In practice, that means a resilient cloud environment is not just one that stays up. It is one where teams understand dependencies, know how systems fail, can act on credible signals and can recover in a controlled way without creating additional market disruption. The paper describes this as a “resilience-by-design” philosophy: predictable behavior under stress comes from deliberate decisions such as multi-region architectures, automated recovery procedures, pre-provisioned capacity and continuous improvement based on operational learning. 

For exchanges, CCPs, and CSDs, that framing shifts the question from “Is the cloud resilient?” to “How do we maintain resilience across cloud services, internal teams, third parties and operating decisions?”
 

The hardest part of maintaining cloud resilience is not any single technology choice; it is managing responsibility, dependencies, failover and evidence together.

• One common challenge is unclear responsibility. In traditional environments, infrastructure and operational accountability often sit more directly within the institution. In cloud-forward models, responsibility is distributed across internal teams, providers and service partners. The paper is explicit that the FMI remains accountable for resilience, which is why shared-responsibility mapping, clear ownership and defined escalation paths matter.
• A second challenge is limited dependency visibility. Cloud services may be resilient by design, but each workload still depends on regions, availability zones, managed services, identity systems, network paths, control planes and administrative access. The paper warns that failures can cascade across those layers. To overcome that, market operators need dependency mapping, failure-domain awareness and resilience planning that covers infrastructure, applications, data flows and operational processes together.
• A third challenge is single-region exposure and uncontrolled failover. Single-region deployments may not always be enough for critical systems. Also, failover is not just a technical switch. It requires decisions about replication, state, routing, latency, degraded operations and authority. The way to overcome that is through multi-region design, tested recovery paths and decision trees that define when automation can act and when operators must intervene.
• A fourth challenge is weak evidence and governance. Architecture diagrams alone do not prove resilience. Operators need records of testing, incident response performance, recovery outcomes, control effectiveness, dependency management, change governance and lessons learned. In other words, resilience has to be demonstrable, not implied.

It can be argued that one of the more effective way to improve cloud resilience is to match architecture, operations and recovery planning to workload criticality.

Not every workload has the same latency, control, recovery or regulatory requirements. Some functions may need infrastructure closer to market participants, while others can take advantage of more elastic cloud resources. Hybrid architecture exists for this reason. Best practices the paper explores include: 

Another important point is that elasticity still has a role to play, but it should be governed by known limits and tested operating controls.
 

Multi-region architecture may be suitable when a workload has continuity requirements that cannot tolerate the concentration risk of a single-region deployment. That does not mean multi-region is a checkbox. It requires decisions about data replication, state management, routing, latency, failover authority and the circumstances under which degraded operation is acceptable. Multi-region design can help improve resilience only when redundancy is understandable, testable and governable.
 

Multi-AZ deployments can reduce localized failure risk by distributing critical systems across availability zones so that disruption in one location does not automatically affect the whole service. The white paper groups multi-region and multi-availability-zone deployment together as part of the resilience foundation for critical systems.
 

FMIs should consider defining RTO and RPO targets by classifying systems according to criticality and aligning recovery objectives to business impact. The white paper gives the example that trading engines may operate with near-zero RTO and RPO, while internal tools or batch systems may tolerate longer recovery windows.

Download the White Paper
 

The white paper’s guidance can be summarized into five recommended best practices for market operators:

1. Acknowledge scope of resilience beyond uptime. Treat resilience as a combination of recoverability, control, observability, governance, evidence, security and compliance.
2. Match architecture to workload criticality. Use workload placement, static stability and pre-provisioned capacity where business and market impact require predictable behavior.
3. Design for regional disruption. Avoid single-region dependence for critical systems and make failover intentional, testable and governed.
4. Test under realistic conditions. Use disaster recovery exercises, surprise drills and chaos testing to validate not just technology but also operator readiness and escalation practices.
5. Build assurance around evidence and learning. Maintain recovery test results, audit trails, incident records, change data, dependency inventories, communications logs, and remediation tracking, then turn those findings into system improvements.

These practices matter because the paper treats resilience as a continuous operating cycle: detect, assess, respond, communicate, and improve. That cycle becomes more important as trading windows extend, cloud services evolve, and market operators face new combinations of workload complexity and third-party dependency. 
 

For FMIs, cloud resilience is no longer just an architecture question. It is an operating-model question: how to maintain control, continuity and confidence as critical market systems modernize.

Managed Services for Nasdaq Eqlipse platforms brings that operating layer to exchanges, CCPs and CSDs. It leverages Nasdaq’s cloud resilience practices, operational discipline, monitoring, incident response, and change management to environments built for mission-critical market infrastructure.

For market operators, the value is practical: reduce execution risk, accelerate resilience maturity, and keep internal teams focused on market innovation while Nasdaq applies the operational expertise gained from decades of running global markets and cloud workloads to help manage resilient, cloud-forward infrastructure.
 

Discover How Nasdaq Financial Technology Empowers Over 3,800 Leading Finance Organizations

We are committed to helping market operators and participants overcome infrastructure, operational and regulatory challenges. Our solutions empower you to focus on your core competencies, driving market growth and progress.

Learn More
 

_{Information set forth in this post contains forward-looking statements. Nasdaq cautions readers that any forward-looking information is not a guarantee of future performance and that actual results could differ from those contained in the forward-looking information. Forward-looking statements can be identified by words such as “will,” “believe” and other words and terms of similar meaning. Forward-looking statements involve a number of risks, uncertainties or other factors beyond Nasdaq’s control. Nasdaq Eqlipse is a product of Nasdaq’s Financial Technology business and is operationally independent and distinct from The Nasdaq Stock Market, LLC.}

_{© 2026 Nasdaq, Inc. The Nasdaq logo and the Nasdaq ‘ribbon’ logo are the registered and unregistered trademarks, or service marks, of Nasdaq, Inc. in the U.S. and other countries. All rights reserved. This communication and the content found by following any link herein are being provided to you by Nasdaq Financial Technology, a business of Nasdaq, Inc. and certain of its subsidiaries (collectively, “Nasdaq”), for informational purposes only. Nothing herein shall constitute a recommendation, solicitation, invitation, inducement, promotion, or offer for the purchase or sale of any investment product, nor shall this material be construed in any way as investment, legal, or tax advice, or as a recommendation, reference, or endorsement by Nasdaq. Nasdaq makes no representation or warranty with respect to this communication or such content and expressly disclaims any implied warranty under law. At the time of publication, the information herein was believed to be accurate, however, such information is subject to change without notice. This information is not directed or intended for distribution to, or use by, any citizen or resident of, or otherwise located in, any jurisdiction where such distribution or use would be contrary to any law or regulation or which would subject Nasdaq to any registration or licensing requirements or any other liability within such jurisdiction. By reviewing this material, you acknowledge that neither Nasdaq nor any of its third-party providers shall under any circumstance be liable for any lost profits or lost opportunity, direct, indirect, special, consequential, incidental, or punitive damages whatsoever, even if Nasdaq or its third-party providers have been advised of the possibility of such damages.}

_{© 2026 Nasdaq, Inc. All rights reserved.}

https://www.nasdaq.com/legal