TL;DR
Get business pricing on monitors, keyboards and dev 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, including guidance on disaster recovery and software that relies on AI models. The material sets out recovery targets and discusses provider, infrastructure and model-change risks; it does not provide evidence that AI has caused a measured rise in outages.
The Pragmatic Engineer has published excerpts from Sam Newman’s new book, Building Resilient Distributed Systems, outlining how software teams can prepare for major infrastructure failures and risks in AI-powered systems. The guidance focuses on setting recovery targets, choosing disaster-recovery arrangements and accounting for changing models and provider dependencies.
In the excerpt from Chapter 10, Newman describes disaster recovery (DR) as the procedures used when a severe event disrupts a large part of a system. Examples include fires, floods, power failures and problems in underlying cloud infrastructure. He says teams face trade-offs: reducing acceptable downtime or potential data loss generally raises cost and complexity.
Newman recommends defining two measurable objectives. The recovery point objective (RPO) sets the maximum acceptable period of data loss, based on how recently data was backed up or synchronized. The recovery time objective (RTO) sets how long a service can remain unavailable. Restoring data, making computing resources available and redirecting traffic all contribute to recovery time.
The report’s overview says the book also compares backup-and-restore, pilot light, warm failover and active-active recovery patterns. Its sections on AI-powered software address reliance on model providers, changes to models and the need for observability. The overview also points to using multiple models or vendors, making costs transparent and reviewing vendors regularly as ways to manage uncertainty.
Recovery Targets Shape Outage Plans
The excerpt turns resilience planning into decisions teams can measure: how much data they can afford to lose and how long users can tolerate a service being unavailable. Those targets help determine whether a less costly backup-and-restore plan is sufficient or whether a faster, more complex recovery setup is justified.
AI services add dependencies that can change outside a software team’s direct control. A model provider might be unavailable, or a model may change; the report says Newman’s book addresses these issues through resilience planning and observability. Multiple providers or models may reduce reliance on one option, but they can also add operational and cost complexity. The excerpt does not quantify those trade-offs.
Products Worth Considering
Logitech MX Keys Combo for Business Keyboard and Mouse – Graphite
Bose SoundLink Flex Bluetooth Speaker (2nd Gen) – Portable Outdoor Speaker with Hi‑Fi Audio, Waterproof and Dustproof, USB‑C, Up to 12 Hours
LG UltraGear 32GX850A-B 32-inch 4K UHD Glossy OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From Distributed Systems to AI
The Pragmatic Engineer says Newman spent three years writing the book, rather than the two he had initially planned, and revised its material to account for AI. The publication describes the book as covering 16 chapters and presents excerpts from two chapters, including one on disaster recovery and another on uncertainty and AI-related resilience.
The source introduction says software teams are seeing more outages and suggests that increased code deployment, including code written by AI agents, and more production bugs may be contributing. That is the publication’s framing, not a quantified finding in the excerpt: it provides no outage data, comparison period or evidence establishing the scale or cause of any increase. Newman’s book is published by O’Reilly Media; the publication says the excerpt is used with permission.
“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, in the excerpt from Building Resilient Distributed Systems
Outage Trends Remain Unquantified
The source does not give data to establish whether outages are increasing, how many incidents involve AI-written code, or whether AI code is a cause of production bugs. Those points appear in the publication’s introduction as a broad observation rather than a measured result.
The supplied excerpt ends during its explanation of RTO and does not include the full chapters or detailed implementation guidance. It also gives no specific recovery targets, vendor assessments, model examples or cost comparisons. The appropriate plan will depend on each system’s requirements and constraints, according to the framework described.
Teams Must Set Their Own Targets
The immediate next step for a team applying the excerpt’s advice is to agree on acceptable data loss and downtime, then compare recovery options against those targets, costs and infrastructure constraints. The excerpt’s framework calls for considering the full return-to-service process, rather than treating restored backups alone as proof that a system is ready.
For AI-dependent services, the report points readers toward checking provider dependencies, model changes, observability, costs and vendor arrangements. The publication says its article includes further excerpts from Newman’s book, but it gives no dates for follow-up material or specific implementation milestones. The book is available from its publisher, according to the report.
Key Questions
What is the news?
The Pragmatic Engineer published excerpts from Sam Newman’s 2026 book on resilient distributed systems, covering disaster recovery and challenges in AI-powered software.
What do RPO and RTO mean?
RPO is the maximum acceptable period of data loss, measured by how long it has been since data was backed up or synchronized. RTO is the maximum acceptable time a service can be unavailable after an incident.
Which disaster-recovery approaches does the report mention?
The report overview lists backup-and-restore, pilot light, warm failover and multisite active-active approaches. The supplied excerpt does not provide the detailed comparison of each pattern.
Does the source prove AI is causing more outages?
No. The publication’s introduction suggests that more deployments and bugs, including those associated with AI-generated code, are contributing to outages, but it provides no incident data or analysis establishing that claim.
What AI-related risks does the book address?
The report says the book discusses model-provider resilience, model changes and observability, as well as approaches such as using multiple models or vendors, tracking costs and reviewing vendors.
Source: rss
