Prashant Lakhera

From On-Premises to Multi-Cloud: Yahoo's Kubernetes Transformation Scaling Kubernetes for platforms like Yahoo Sports, Finance, and News is no small feat. Yahoo’s journey from on-premises to multi-cloud architecture is a masterclass in innovation and problem-solving at scale. In a recent talk by Nandhakumar Venkatachalam and Payal Patel, they shared valuable insights into how Yahoo tackled this transformation, balancing complexity, performance, and security. Here are some standout lessons from their experience: 1️⃣ 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱𝗶𝘇𝗶𝗻𝗴 𝘁𝗵𝗲 𝗦𝗲𝘁𝘂𝗽: In-house tools like KDNC and FLEKS simplified managing Kubernetes clusters across environments. These tools leveraged Terraform and CloudFormation to create a consistent, efficient way of provisioning clusters. 2️⃣ 𝗦𝗲𝗰𝘂𝗿𝗶𝗻𝗴 𝗜𝗱𝗲𝗻𝘁𝗶𝘁𝘆 𝗮𝗻𝗱 𝗔𝗰𝗰𝗲𝘀𝘀: Yahoo adopted Athenz and Istio to automate RBAC and establish app/service identities. Security and scalability were seamlessly integrated when using OIDC for EKS or Workload Identity Pools for GKE. 3️⃣ 𝗦𝗰𝗮𝗹𝗶𝗻𝗴 𝗢𝗯𝘀𝗲𝗿𝘃𝗮𝗯𝗶𝗹𝗶𝘁𝘆: With Prometheus, Grafana, and Splunk, Yahoo monitored their infrastructure across 37+ clusters, supporting 135,000 pods, 385,000 containers, and 1.8 million requests per second—a truly massive operation. 4️⃣ 𝗢𝗽𝘁𝗶𝗺𝗶𝘇𝗶𝗻𝗴 𝗡𝗲𝘁𝘄𝗼𝗿𝗸𝗶𝗻𝗴: Yahoo’s approach to cloud-native networking stood out. Custom CIDRs, VPC Peering, and Istio ingress configurations helped reduce costs and improve performance across AWS and GCP regions. 5️⃣ 𝗗𝗿𝗶𝘃𝗶𝗻𝗴 𝗖𝗼𝘀𝘁 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆: Bin-packing strategies and resource optimization allowed Yahoo to achieve significant savings. A notable example is their GKE-based Header Bidding project, which reduced costs by 33% while maintaining high throughput. What’s remarkable about Yahoo’s approach is its focus on abstraction—allowing developers to concentrate on building applications while the platform handled the heavy lifting around infrastructure, security, and observability. It’s a model of balancing complexity and usability in a large-scale, multi-cloud environment. You can dive into the full presentation here: https://lnkd.in/eSh77edu

7

Long trusted for expertise in application security and analysis, Veracode adopted Kubernetes approximately two years ago. As they migrated to Amazon EKS, they recognized the need to ensure compliance with reliability and security policies across their Kubernetes deployments. Our latest case study dives into how they built reliability and security into their K8s infra right from the start: https://bit.ly/48DVU8G #kubernetes #infrastructure #veracode #fairwindsInsights #K8s #reliability #security #casestudy

3

Don't miss this fireside chat on August 21 at 2 PM ET/11 AM PT designed for technology leaders responsible for Kubernetes management, efficiency, and cost optimization. Dieter Matzion, Sr. Cloud Governance Engineer at Roku, will join Andy Suderman, CTO at Fairwinds, and Emilia A’Bell, Chief Revenue Officer, Platform9 who will provide actionable insights to elevate your Kubernetes strategy. Register here: https://bit.ly/3AgmUOj #webinar #kubernetesmanagement #kubernetesefficiency #costefficiency #roku #kubernetesstrategy #kubernetes #managedkubernetesasaservice #devops #finops

1

🚀 Amazon OpenSearch Service just got a major upgrade: 📊 Scale to 1000 data nodes on a single cluster 💾 Manage up to 25 petabytes of data 🔥 Support for 1000 hot nodes and 750 warm nodes This update eliminates the need for multiple clusters when handling large workloads. Say goodbye to complex refactoring and hello to simplified management! Key benefits: • Reduced operational overhead • Streamlined configuration and monitoring • Available for OpenSearch 2.17 and above • Accessible in all AWS regions with OpenSearch Service Ready to supercharge your data operations? Check out the AWS documentation for more details on scaling your cluster beyond 200 nodes. #AWS #OpenSearch #BigData #CloudComputing #DataManagement Sources [1] https://lnkd.in/dt3GaZdy [2] Amazon OpenSearch Service now scales to 1000 data nodes on a single cluster - AWS https://lnkd.in/dt3GaZdy

23

1 Comment

Announced at #AWSreinvent: Cribl now supports AWS Private Image Build! 🔐 How do you ensure your telemetry pipelines stay secure and compliant? With Cribl Stream and Cribl Edge, combined with AWS Private Image Build, you can create Amazon Machine Images (AMIs) tailored to your specific requirements. This lets you build “golden” or “hardened” AMIs that meet your organization’s security and compliance standards. 💡 Learn more about the benefits: https://gag.gl/1i9d7h

1

Yahoo’s Kubernetes Journey from on-Prem to Multi-Cloud   🎯 Key Innovations:     - Migrated from on-prem infrastructure to a multi-cloud ecosystem with GCP and AWS.     - Developed ‘Omega’ interface simplifying app deployment to Kubernetes clusters.     - Introduced ‘Cadence’ tool for seamless cluster creation across multiple clouds.    💡 Notable Features:     - **Omega YAML**: A simplified deployment template ensuring quick app setups.     - **Cadence Tool**: A Golang-based tool for managing EKS and GKE workloads without deep cloud provider knowledge.     - **Node Manager Operator**: Automates Kubernetes version upgrades while maintaining workloads.    🛠️ Perfect for:     - Cloud Engineers     - Kubernetes Administrators     - DevOps Teams     - Platform Engineers     - Organizations transitioning to multi-cloud environments    ⚡️ Impact:     - Enhanced deployment efficiency by streamlining application setups.     - Achieved significant cost reduction for operations in cloud environments.     - Facilitated rapid responses to infrastructure changes and reduced upgrade times from months to weeks.    🔍 Preview of the Talk:     Nandhakumar Venkatachalam and Payal Patel from Yahoo discuss the company's transition from a traditional on-prem environment to a cutting-edge multi-cloud strategy using Kubernetes. They delve into the challenges faced, such as networking and security, while showcasing breakthrough tools like Omega and Cadence. The talk emphasizes how these innovations enhance user experiences and streamline operations. Participants will gain valuable insights into effective cloud strategies and operational efficiencies in Kubernetes environments. Watch the full session: https://lnkd.in/gcZS--rY

1

With AWS Machine Learning capabilities now easily accessible, individuals can build cutting-edge models that deliver real value with minimal effort. However, once you’ve proven a small use case, how do you scale and operationalize your models for enterprise use? To take a project from concept to production-ready, organizations need to incorporate enterprise-grade functionality—like tracking code changes, enforcing security standards, and ensuring policy and governance compliance. Using Terraform to build and manage the lifecycle of an ML/AI platform on AWS provides a scalable approach to address these requirements. Terraform can help orchestrate your infrastructure, ensuring your ML models are production-ready and enterprise-compliant. #MachineLearning #AWS #Terraform #AIOps #DevOps #CloudInnovation

4

🚀 Better Together: Cockroach Database + AWS ☁️ Date: 24th September 2024 Location: Grand Hyatt Hotel, Bandra Kurla Complex, Mumbai Cockroach Labs will be @ AWS Financial Services Symposium 2024 in Mumbai! #️⃣ I'll be there and invite you all to join me—let’s have a quick catch-up and kickstart the conversation of "Why Distributed SQL with CockroachDB + AWS?" . We Welcome FSI leaders to explore innovative cloud-native solutions that tackle RDBMS Horizontal Scale and hybrid cloud challenges. 🔍 Key Discussions: ✅ Why are so many FSI and fintech's talking about distributed databases? How can CockroachDB + AWS be the perfect solution. ✅ How can you scale your RDBMS workloads with CockroachDB’s multi-region capabilities/ Muti-master / Active-Active capabilities. ✅ We invite all FSI leaders to join us for insightful conversations on how Better Together: Cockroach Database + AWS can accelerate your business transformation. Showcasing the power of "Better Together Cockroach Database + AWS" to solve the most complex data challenges. By bringing the best of distributed SQL and the scalability of AWS. Let’s connect at the event! Stop by our booth to learn more or schedule a chat with our team. #FinancialServices #DistributedSQL #AWSCloud #BetterTogether #CockroachDB #HybridCloud #FSI #Fintech #InnovationInFSI

42

2 Comments

New Relic Announces One-Step Observability for Kubernetes to Deliver Automatic Instrumentation and Insights 𝐊𝐧𝐨𝐰 𝐌𝐨𝐫𝐞👉 https://lnkd.in/gct97Abd #NewRelic, the Intelligent #Observability Platform, announced one-step observability for Kubernetes, addressing #keymonitoring challenges #developers face in managing #dynamic Kubernetes #environments. New Relic #NewRelic Manav Khurana Ashan Willy Peter Marelas SWAPAN ROY NCN Magazine #ncnmagazine #ncnonline #NCN

1

Effortless Cluster Management Let's be honest, managing Kubernetes clusters can be a real headache. It's complex, time-consuming, and requires specialized expertise that not every team has readily available. But with AWS EKS, you can say goodbye to those infrastructure worries and focus on what truly matters: building and deploying your amazing applications. #Containers #devops #insights #Kubernetes

1

Graviton 4 with this instance refresh.. EBS perf is parity with previous Gen.

6

Interesting feature of AWS S3's 𝗖𝗼𝗻𝗱𝗶𝘁𝗶𝗼𝗻𝗮𝗹 𝗪𝗿𝗶𝘁𝗲 !! This looks like a simple feature, but it is a powerful one. AWS announced this recently on August 20th. https://lnkd.in/gbkabTx4 This feature helps to check for the existence of an object before uploading it and thus prevents overwriting of the object. To use this, we can pass an an additional header --𝗶𝗳-𝗻𝗼𝗻𝗲-𝗺𝗮𝘁𝗰𝗵 in the aws command to upload the file. The command is mentioned in this AWS doc https://lnkd.in/gS4vnXtP To put it in simple, it's like doing Skip All when copying photos into a backup directory. If the files already exist, we click on "Skip All" so it doesn't copy the existing files and continues to copy the rest new files. 𝗕𝗲𝗻𝗲𝗳𝗶𝘁𝘀 𝗳𝗿𝗼𝗺 𝘁𝗵𝗶𝘀 𝗳𝗲𝗮𝘁𝘂𝗿𝗲: ✔️ Saves time spent doing re-copy operation. ✔️ Saves upload cost by avoiding copying of same files which already exist in the bucket. ✔️ Prevents accidental loss of data due to overwrite when versioning is not enabled. ✔️ Using of this feature does not incur additional charge. ✔️ Using of Conditional write can avoid the duplications thus keeping the versions count lower. There are different use-cases in which this feature can play crucial role. I can think of below 2 use-cases: 𝗨𝘀𝗲 𝗖𝗮𝘀𝗲 𝟭 Let's assume you have a solution in which multiple Analysts upload the stock market related csv files into a common share bucket. Now let's say you have configured an S3 trigger action (for ObjectCreated) in a Lambda function. The Lambda function will read the file in the S3 bucket, does some computation to extract and stores the data into the DB. Now, if the file already exist in the S3 bucket then without the conditional write feature, the same file uploaded by multiple analyst will make Lambda function to trigger multiple times resulting in Lambda invocation costs, S3 upload costs and maybe duplication of data. 𝗨𝘀𝗲 𝗖𝗮𝘀𝗲 𝟮 Let's assume you have a solution in which the Analysts upload the datasets into S3 bucket and using which an automation process kicks off to train the ML model. If the same dataset file is overwritten into S3 bucket, then it will be used to re-train the model which results in wasted compute power, increased costs, and inefficiency. You don't have to build the additional programming logic, automation or additional API requests to check for the file existence and avoid rewrites. Instead, use the conditional write option while writing the objects to S3 bucket. This feature helps to maintain 𝗱𝗮𝘁𝗮 𝗶𝗻𝘁𝗲𝗴𝗿𝗶𝘁𝘆, 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗮𝗻𝗱 𝗰𝗼𝘀𝘁-𝗲𝗳𝗳𝗲𝗰𝘁𝗶𝘃𝗲𝗻𝗲𝘀𝘀. Follow me for simplified learning and best practices on DevSecOps Save for future reference. 🔁 Consider a Repost if this is useful. #aws #devops #DataManagement #DataIntegrity #S3 #CloudStorage #CostOptimization #CloudArchitecture

13

A standout session from Day 2 at re:Invent was “Optimizing Workloads for Speed and Cost Using Amazon ECS and AWS Fargate” by Rachel Leekin and Praveen Edem. Some key tips included: ☑️ Save 30-70% by right-sizing tasks with AWS Compute Optimizer and using Compute Savings Plans. ☑️ Boost performance and reduce costs with Graviton instances and AWS Fargate Spot. ☑️ Optimize auto-scaling with the new predictive scaling feature. ☑️ Minimize container image sizes for better efficiency. ☑️ Use SOCI (Seekable OCI) to speed up image pulls with lazy loading. If you couldn't attend this session, check out our article, by Dennis Ruzeski, for more insights: https://lnkd.in/giKBS9Cd #ippontech #reinvent2024 #amazonecs #costsavings #performanceoptimization

11

3 Comments

Check out the updated #aws #observability guide . https://lnkd.in/ggZPiMKq

12

After over three years of maintaining #JCenter as a read-only repository for legacy #Java packages, JFrog is moving forward with an official sunset of JCenter and redirection of all requests to Maven Central. Learn more about it here: https://lnkd.in/gNyvsZxP

36