Cloud Therapist Reveals Surprising Trends in 2025! скачать в хорошем качестве

Cloud Therapist Reveals Surprising Trends in 2025!

Observability, Generative AI/ML, Disaster Recovery, AIOps, Multicloud, Public Cloud, Uptime/Downtime, Data Centre, Energy/Power, Automation, LLM. We have it all and more in this June episode of Cloud State Of Play! #observability #genai #cloudnative #podcast #cloudtherapist #cloudcostmanagement #finops #disasterrecovery #businesscontinuity Welcome to the Cloud State Of Play - June 2025 episode. Paul Bevan (Director of Infrastructure Research at Bloor Research) is back with me for one final month, before he disappears for a long-overdue and well-deserved summer break. === The topics for this month are: 1 - What Level Of Uptime Do Your Business Services Require? 2 - Dynatrace - How Modern Cloud Observability Tames Complexity 3 - McKinsey - AI, Data Centres, and the Energy Equation === 1 - What Level Of Uptime Do Your Business Services Require? Based on an AquaSec LiN post about how GenAI Models exponentially increase your attack surface: https://www.linkedin.com/posts/aquase... This initiated an e-mail exchange between me and Paul, which led us to discussing uptime for business services. My main point was that for GenAI workloads, specifically, there's no guarantee that everyone will want to build their own GenAI platform There's a likelihood that most SMEs will go to an AI-SaaS vendor solution instead and just dump their data into such, industry-optimised models Paul's initial point was "it all relies on the professionalism and reliability of your AI-SaaS provider!" This is where I differed from Paul, because as I stated above, not everyone will build their own GenAI platforms, so most (in my opinion) will rely on SLAs instead, from the AI-SaaS vendors This is an easier risk calculation to make, than 'guessing' your own service uptime requirements, through RTO/RPO/SLO, etc. === 2 - Dynatrace - How Modern Cloud Observability Tames Complexity The document explores how modern cloud observability confronts the increasing complexity IT environments face as enterprises shift to cloud-native architectures. https://www.dynatrace.com/info/ebooks... Highlights 🌐 Modern cloud observability tackles the complexity introduced by Multicloud, microservices, and containerized architectures. 🤖 AI and Automation transform observability from reactive monitoring to proactive problem solving. 📊 Observability includes three pillars—metrics, [events], logs, traces (MELT)—enhanced with context and user data for actionable insights. 🏗️ Digital transformation amplifies cloud complexity, making observability a critical success factor. 🔄 Teams evolve from legacy monitoring tools to unified AI-powered Observability platforms for instant visibility. 🛠️ Dynatrace’s platform automates data collection, Root Cause Analysis, and cross-team collaboration. 🚀 Real-world case study highlights how Observability supports essential government services during crisis. MELT = Metric, Events, Logs and Traces (what Observability tools collect and what I had nicknamed as "MELT" but according to Paul, it turns out that the Observability community already use this acronym for the same thing! Why stop at Observability? Where's the actions?! Why can't Observability tools also do full-blown Remedy and Prevention? Extend the MELT acronym to "MELTReP" (Re = Remediation, P = Prevention) === 3 - McKinsey - AI, Data Centres, and the Energy Equation: What Leaders Should Know - Webinar   / theater     / about   Lucia Rahilly (Global Editorial Director) Pankaj Sachdeva (Snr Partner) Jesse Noffsinger (Partner) 4 things to consider: 1 - AI value chain - customer-driven view for infra utilisation 2 - How can you max value from existing Capex 3 - What are durable business models and investment opportunities 4 - Identify potential risks === More info on Bloor Research: https://www.bloorresearch.com/ More info on all topics in the pinned comment...coming soon! === Thanks to Paul Bevan, for his insights. All opinions expressed in this video are solely of the person who gave them and do not reflect opinions/stance/policy of any other person(s)/employer(s)/organisations/entities/vendors, etc. All copyrights are of their respective owners and are used here (in various formats) purely for reference purposes and under fair use policy. E&OE. === Intro + Outro Track: "Know Myself" - Patrick Patrikios (YouTube Music Library).