TL;DR
Get business pricing on networking and server gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
The Pragmatic Engineer published excerpts from Sam Newman’s 2026 book, Building Resilient Distributed Systems, on disaster recovery and resilience for AI-powered software. The material explains how recovery targets, infrastructure constraints, model changes and vendor uncertainty should shape planning. It does not document a specific outage or provide data showing that AI coding tools have increased incidents.
The Pragmatic Engineer has published excerpts from Sam Newman’s 2026 book, Building Resilient Distributed Systems, outlining disaster recovery practices and challenges for software that relies on AI. The material matters to engineering teams because it connects service recovery targets with newer dependencies, including model providers whose availability and systems can change.
Newman’s excerpt on disaster recovery focuses on failures that affect an entire cloud region, data center or other infrastructure site. It defines a recovery point objective (RPO) as the maximum acceptable period of potential data loss, and a recovery time objective (RTO) as the acceptable duration of service unavailability. For example, hourly backups could leave up to an hour of data at risk after an incident. RTO includes the work of restoring data, making computing resources available and redirecting traffic.
The report describes several recovery patterns, including backup and restore, pilot light, warm failover and multisite active-active. They offer different balances of cost, complexity and recovery speed. Newman’s broader discussion also addresses disaster recovery objectives and service-level objectives, and the need to plan around constraints in infrastructure availability.
For AI-powered services, the excerpt identifies model provider resilience, model changes and observability as concerns. The report also points to strategies for handling uncertainty, including using multiple models or vendors, making costs visible and regularly reviewing vendors. It says continuing GPU and CPU shortages mean teams should not assume infrastructure will always be available and should prioritize critical services.
Recovery Targets Shape System Readiness
The excerpt frames resilience as a set of decisions teams can make before an incident. Setting RPO and RTO targets makes the acceptable loss of data and downtime explicit, helping teams choose recovery patterns that fit the service. Faster recovery and lower data loss can require more infrastructure, expense and operational complexity.
AI services add dependencies that can shift: a model provider may face availability limits, a model may change, and computing resources may be constrained. Observability and vendor review can help teams detect and respond to those changes. The report offers engineering guidance, but it does not establish how often these risks occur or quantify their effect on outages.
cloud backup software for disaster recovery
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Site Failures to AI Dependencies
Newman’s book addresses resilience across 16 chapters, according to The Pragmatic Engineer. The published excerpts draw from two chapters, covering disaster recovery and uncertainty in AI-powered systems. Newman said he spent three years writing the book, one year longer than planned, and revisited its material to account for AI.
The newsletter introduces the excerpt by saying outages are occurring more often and linking that trend partly to more deployments and production bugs as AI agents write more code. That is the newsletter’s characterization; the supplied material provides no outage dataset or comparison baseline to verify the trend or measure AI’s contribution. The excerpt itself focuses on planning and recovery principles rather than reporting a particular incident.
“The recovery point objective (RPO) defines the maximum amount of time that can have passed since the data was last backed up or synchronized.”
— Sam Newman, author of Building Resilient Distributed Systems, in the excerpt
As an affiliate, we earn on qualifying purchases.
The Scale of AI-Linked Outages
The supplied report does not identify specific outages caused by AI-generated code or provide figures comparing incident rates before and after AI agents became more common. Its statement that outages are becoming more frequent is not supported here with a dataset, so the scale and cause of any increase remain unclear.
The excerpt also gives no universal recovery targets or cost estimates. Appropriate RPO, RTO and disaster recovery patterns depend on each system’s needs and infrastructure. The report does not specify how often model providers change their systems, or how much redundancy a particular AI service should maintain.
multi-region cloud failover solutions
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Teams Must Set Their Own Targets
The next step for engineering teams is to define acceptable data loss and downtime for each service, then select and practice a recovery approach that meets those targets. Teams building AI-powered software can also assess provider dependencies, model-change monitoring, infrastructure limits and whether multiple vendors or models are appropriate.
The Pragmatic Engineer presents the material as excerpts from Newman’s book, which is published by O’Reilly Media. No specific follow-up release, implementation milestone or industry response is given in the source material.
enterprise data backup and restore software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is the news development?
The Pragmatic Engineer published excerpts from Sam Newman’s 2026 book on disaster recovery and resilience for AI-powered software.
What do RPO and RTO mean?
RPO sets the maximum period of potential data loss after an incident. RTO sets how long a service can remain unavailable while it is restored.
Which disaster recovery approaches does the excerpt mention?
It names backup and restore, pilot light failover, warm failover and multisite active-active. These approaches differ in cost, complexity and recovery speed.
What AI-related risks does the report identify?
It points to model provider resilience, model changes and observability, along with potential limits on GPU and CPU availability. It also describes multiple models or vendors, cost transparency and regular vendor reviews as ways to manage uncertainty.
Does the source prove that AI agents are increasing outages?
No. The newsletter says more deployments and production bugs from AI-written code are contributing to more outages, but the supplied material includes no supporting data or comparison baseline. The scale of that claimed trend remains unclear.
Source: rss
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
