The Promise Of Redundancy: Can The Impact Of A Cloud Outage Be Avoided? 

The October 20 AWS outage was a powerful reminder of conscionable really interconnected today’s applications and services person become. From banking to streaming, healthcare to logistics, organizations of each sizes and industries trust connected a analyzable web of nationalist unreality and different third-party services. As we saw, a azygous disruption tin quickly cascade, affecting not conscionable 1 institution but full industries and millions of extremity users.

Faced pinch specified disruption, it’s earthy to ask: Why aren’t much companies capable to build effective redundancies to shield themselves from disruptions for illustration these? The reply lies successful complexity.

The Hidden Complexity Behind Modern Applications

The seamless integer experiences customers and labor expect are powered by a dense cloth of infrastructure and work components, often originated from 3rd parties. Modern applications dangle connected myriad underlying services, including unreality platforms, managed databases, serverless functions and outer APIs that whitethorn themselves trust connected nan aforesaid cloud providers aliases akin outer dependencies. This intricate web makes it operationally and economically challenging to build afloat redundant systems.

Even pinch engineered failovers, specified arsenic switching to different unreality supplier region, these strategies are acold from straightforward. Each further furniture of redundancy introduces its ain group of dependencies and management challenges.

Full Redundancy Isn’t Possible 

For organizations that do person immoderate redundancy successful place, knowing erstwhile to invoke failover is simply a difficult calculus. Redundancy tin beryllium architected successful respective ways: Maintaining aggregate discrete nonaccomplishment zones, wherever instances and workloads are distributed crossed different cloud providers (multicloud), aliases employing active-active architectures wherever workloads tally successful parallel and work tin beryllium maintained if either becomes unavailable. For example, an e-commerce level mightiness replicate its captious databases and exertion servers crossed 2 chopped regions wrong nan aforesaid cloud supplier to guarantee service continuity if 1 region experiences an outage.

However, failovers and remediation actions tin themselves beryllium disruptive and require clip to execute. Data consistency, convention authorities synchronization and DNS propagation delays tin each present complications and imaginable work degradation during a transition. In immoderate cases, a failover mightiness create caller issues if nan secondary situation isn’t afloat up to day aliases if it shares hidden limitations pinch nan superior one.

Making nan correct determination depends connected knowing nan outage’s scope (localized aliases widespread), long (temporary aliases prolonged), nan behaviour of underlying limitations and nan existent effect connected users and business outcomes. Without this insight, remediation tin beryllium delayed aliases moreover worsen nan business by disrupting users aliases compounding method challenges.

The Case for Visibility and Dependency Mapping

To meet these challenges, organizations should prioritize improving visibility into nan environments they dangle on, whether they are self-managed aliases provided by 3rd parties.  Mapping exertion and work limitations is basal for uncovering hidden risks, specified arsenic chartless azygous points of failure, and for forming redundancy strategies. During an outage, real-time penetration into really each dependency is performing and really extremity users are affected becomes captious for making fast, informed decisions.

Provider position updates tin beryllium delayed aliases excessively wide to address a circumstantial company’s situation. Direct visibility into work behaviour and personification effect enables organizations to pass intelligibly and enactment decisively, minimizing business disruption.

The Role of Digital Resilience

Cloud supplier outages punctual america that resilience depends not only connected smart architecture, but besides connected intelligence and visibility crossed nan full service. As organizations proceed to clasp cloud, SaaS and now AI workloads, whose architectures often summation dependency complexity, it’s basal to admit that each introduces some tremendous opportunity and caller categories of risk.

The expertise to navigate outage events and different disruptions depends not conscionable connected redundancy, which tin ne'er beryllium perfect, but connected really efficaciously organizations tin see, understand and respond to their environments nether duress. This biology consciousness requires end-to-end visibility, making it a cornerstone of integer resilience.

Group Created pinch Sketch.