# autoscaling Published articles for autoscaling. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Lyft Moves Streaming Fleet to Apache Flink Kubernetes Operator DevFeed: [Lyft Moves Streaming Fleet to Apache Flink Kubernetes Operator]() Original publisher: [Read original article]() Author: Mark Silvester Published: 2026-09-16T11:00:00Z Content type: news Language: en Sources: [InfoQ]() Topics: [flink](), [apache-flink](), [Kubernetes](), [autoscaling](), [Deployment](), [legacy](), [Testing]() Tags: [apache-flink](), [architecture-design](), [autoscaling](), [deployment](), [development](), [devops](), [flink](), [kubernetes](), [kubernetes-operator](), [legacy](), [lyft-flink-k8s-operator](), [news](), [testing](), [upgrade]() ### AI overview Lyft moved hundreds of production Apache Flink jobs from its in-house Kubernetes operator to the Apache Flink Kubernetes Operator. The change enabled last-state upgrades, in-place autoscaling, and resource autotuning, while Lyft adapted legacy deployment specifications through its deploy API and contributed a fix for a configuration-renaming bug. ### Source excerpt Lyft has moved hundreds of production Flink jobs from a 2020 in-house Kubernetes operator to the Apache Flink Kubernetes Operator, unlocking last-state upgrades, in-place autoscaling and resource autotuning across the fleet. By Mark Silvester ## Enterprise-Grade Precision for Long-Context Multimodal Embedding Inference on Cloud TPU DevFeed: [Enterprise-Grade Precision for Long-Context Multimodal Embedding Inference on Cloud TPU]() Original publisher: [Read original article]() Author: Anthony Su; Injae Kwak Published: 2026-09-12T11:04:33.891311Z Content type: article Language: en Sources: [Google Developers Blog]() Topics: [Embeddings](), [vllm](), [Inference](), [long-context](), [autoscaling](), [Kubernetes](), [Cloud](), [Google](), [Artificial Intelligence](), [model-deployment](), [Open Source]() Tags: [autoscaling](), [embedding](), [github](), [gpu](), [inference](), [kubernetes](), [long-context](), [multimodal](), [open-source](), [retrieval](), [tpu](), [vllm]() ### AI overview This article explains how Google Cloud integrated native TPU support into vLLM to serve long-context, multimodal embedding models at production scale. It describes elastic TPU provisioning with Google Kubernetes Engine, autoscaling across accelerator types, and TPU-specific optimizations for long sequences and chunked prefill. The resulting setup is designed to preserve numerical parity with GPU reference baselines while supporting high-throughput semantic retrieval applications. ### Source excerpt Google Cloud has natively integrated TPU support into the vLLM serving engine, allowing developers to elastically scale high-demand embedding pipelines using Google Kubernetes Engine (GKE). To handle massive 15K+ token contexts for models like Qwen3-Embedding-8B, the engineering team implemented TPU-specific optimizations such as hardware-safe tensor alignment, JAX/XLA compilation pre-warming, and a hybrid StepPool architecture for chunked prefill management. These enhancements achieve near-perfect numerical parity with reference GPU baselines, and developers can immediately leverage the open-sourced setup recipes on the AI-Hypercomputer GitHub to build their own high-throughput semantic retrieval applications. ## Reduce inference cold starts on Amazon SageMaker HyperPod with model caching DevFeed: [Reduce inference cold starts on Amazon SageMaker HyperPod with model caching]() Original publisher: [Read original article]() Author: Kareem Syed-Mohammed Published: 2026-09-10T21:37:49Z Content type: tutorial Language: en Sources: [Artificial Intelligence]() Topics: [Caching](), [Inference](), [model-deployment](), [Large Language Model](), [autoscaling](), [Kubernetes](), [Amazon S3]() Tags: [amazon](), [amazon-s3](), [amazon-sagemaker](), [amazon-sagemaker-hyperpod](), [announcements](), [autoscaling](), [benchmark](), [caching](), [expert-400](), [inference](), [kubernetes](), [llm]() ### AI overview Amazon SageMaker HyperPod model caching preloads model weights and container images onto cluster nodes, reducing inference-pod cold starts by serving assets from local NVMe storage. ### Source excerpt Amazon SageMaker HyperPod now supports model caching for inference, which pre-loads model weights and container images onto cluster nodes so pods read from local NVMe storage instead of downloading over the network. Learn how model caching cuts cold starts from tens of minutes to seconds, how it works, and how to enable it. ## Announcing On-Demand Compute: Instant compute for your most intensive workloads DevFeed: [Announcing On-Demand Compute: Instant compute for your most intensive workloads]() Original publisher: [Read original article]() Author: Melvyn Peignon Published: 2026-09-10T14:22:58Z Content type: release Language: en Sources: [ClickHouse Blog]() Topics: [clickhouse](), [autoscaling](), [cloud-infrastructure]() Tags: [autoscaling](), [clickhouse](), [cloud](), [compute](), [cost](), [infrastructure](), [production](), [scale](), [workers]() ### AI overview ClickHouse announces a private preview of On-Demand Compute, which allocates shared workers to eligible individual queries so intensive workloads can run without competing with a service's primary compute. ### Source excerpt ClickHouse On-Demand Compute lets you scale individual queries with additional workers, run intensive workloads without disrupting production, and use compute when you need it. ## Improving Lakebase Postgres compute cache DevFeed: [Improving Lakebase Postgres compute cache]() Original publisher: [Read original article]() Author: David Wein; Sunil Kamath; Haoyu Huang Published: 2026-09-10T13:47:03Z Content type: article Language: en Sources: [Databricks]() Topics: [Caching](), [databricks](), [Databases](), [autoscaling](), [Amazon S3](), [Filesystems](), [cloud-infrastructure]() Tags: [autoscaling](), [cache](), [caching](), [data](), [data-engineering](), [databricks](), [engineering](), [filesystem](), [latency](), [low-latency](), [memory](), [s3](), [storage]() ### AI overview This Databricks article describes improvements to compute-side caching for Lakebase Postgres, a disaggregated storage system backed by object storage such as Amazon S3. It explains PostgreSQL shared buffers, the operating system page cache, and the planned use of dynamically autoscaling shared buffers consuming up to 75% of compute memory. It also introduces a local file cache as an incremental solution for fixed compute instances. ### Source excerpt The disaggregated storage model of Lakebase Postgres provides a feature rich, flexible... ## Improving Lakebase Postgres Compute Cache on Neon, Part 1 DevFeed: [Improving Lakebase Postgres Compute Cache on Neon, Part 1]() Original publisher: [Read original article]() Author: Sunil Kamath Published: 2026-09-09T12:00:00Z Content type: article Language: en Sources: [Blog -- Neon Docs]() Topics: [Caching](), [Databases](), [Operating system](), [Amazon S3](), [cpu](), [Latency](), [autoscaling]() Tags: [autoscaling](), [cache](), [compute](), [cpu](), [engineering](), [latency](), [memory](), [os](), [postgres](), [s3](), [storage]() ### AI overview Neon describes a Lakebase Postgres compute-cache change that allocates most memory to shared buffers backed by huge pages. The goal is to keep hot pages in DRAM, reducing storage reads, CPU use, and latency; the article reports up to roughly 2x throughput on specified fixed-size computes. ### Source excerpt On large fixed-size Lakebase Postgres computes on Neon, we now put most of the machine's memory into Postgres shared buffers and back that cache with huge pages. Hot pages stay in DRAM instead of falling through to a local disk cache, so the same working set is served faster and with less CPU. ## Cloud Native Computing Foundation Announces Karmada Graduation DevFeed: [Cloud Native Computing Foundation Announces Karmada Graduation]() Original publisher: [Read original article]() Author: Haley White Published: 2026-09-08T02:06:17Z Content type: news Language: en Sources: [Cloud Native Computing Foundation]() Topics: [Kubernetes](), [Orchestration](), [Cloud Native Ecosystem](), [Inference](), [autoscaling](), [GPU](), [cloud-infrastructure]() Tags: [ai](), [ai-infrastructure](), [announcements](), [autoscaling](), [china](), [cloud](), [critical-infrastructure](), [ecosystem](), [enterprise](), [global](), [gpu](), [inference](), [infrastructure](), [kubernetes](), [open-source](), [orchestration](), [production](), [pytorch](), [release](), [resilience](), [scale]() ### AI overview Karmada has graduated from the CNCF after reaching production maturity as an open-source engine for orchestrating applications across multiple Kubernetes clusters, clouds, and regions. Its v1.19 release adds multi-component scheduling for distributed AI training jobs, while the project supports hybrid-cloud capacity, multi-region resilience, and multi-cluster AI infrastructure. ### Source excerpt Multi-cluster, multi-cloud Kubernetes orchestration project reaches production maturity as global enterprises scale AI training and inference across hybrid infrastructure Key Highlights SHANGHAI, China - KubeCon + CloudNativeCon + OpenInfra Summit + PyTorch Conference China 2026 -... ## Kubernetes v1.37: Scale Workloads to Zero with HorizontalPodAutoscaler DevFeed: [Kubernetes v1.37: Scale Workloads to Zero with HorizontalPodAutoscaler]() Original publisher: [Read original article]() Author: Johannes Würbach Published: 2026-09-02T18:30:00Z Content type: release Language: en Sources: [Kubernetes Blog]() Topics: [Kubernetes](), [autoscaling]() Tags: [autoscaling](), [deployment](), [kubernetes](), [metrics](), [prometheus]() ### AI overview Kubernetes v1.37 adds beta, default-enabled HPA support for scaling eligible workloads to zero and back using object or external metrics. It explains queue-based scaling, cold-start trade-offs, and the metrics adapter setup required for an external Prometheus metric. ### Source excerpt Kubernetes v1.37 includes API support for horizontal autoscaling of workloads down to zero replicas. This feature is now Beta and enabled by default. A HorizontalPodAutoscaler (HPA) that uses a suitable object metric or external metric can now scale a workload to zero replicas, then bring it back when the metric changes. Before v1.37, you needed an add-on or external component, or you had to enable the Alpha feature gate, to scale from zero. It is now part of core Kubernetes. Scaling to zero removes the last idle Pod from workloads such as queue consumers and batch processors. The savings are largest when each Pod reserves expensive resources, including dedicated CPUs or GPUs. The trade-off is cold-start time: the HPA must observe the metric, schedule a Pod, and start the application. This works well when work can wait in a durable queue. Kubernetes Services do not buffer requests while no Pods are ready, so HTTP and other request-driven workloads need a separate buffering layer. Why scaling from zero needs a different metric The HPA commonly scales on CPU or memory usage. Both metrics come from running Pods. Once the replica count reaches zero, there are no Pods left to measure and no signal that can tell the HPA to scale back up. Object and external metrics do not have that limitation. A queue length, for example, exists independently of the workers that consume it. The HPA can continue reading the queue length while no workers are running. The following example scales a queue consumer to and from zero using an external metric. Configure an external metric The following example uses a Prometheus metric named queue_consumer_lag. It assumes that Prometheus already collects a series similar to this one: queue_consumer_lag{namespace="default",name="worker_tasks"} Kubernetes needs a metrics adapter to make that value available through the External Metrics API. One implementation is the Prometheus Adapter, which can expose the series using an externalRules entry: extern ## Announcing native MySQL databases: Bring your MySQL workloads to Appwrite DevFeed: [Announcing native MySQL databases: Bring your MySQL workloads to Appwrite]() Original publisher: [Read original article]() Author: Jake Barnby Published: 2026-09-02T00:00:00Z Content type: release Language: en Sources: [Appwrite Blog]() Topics: [Appwrite](), [MySQL](), [Databases](), [Provisioning](), [Availability](), [Replication](), [autoscaling](), [TLS (Transport Layer Security)]() Tags: [announcements](), [autoscaling](), [availability](), [database](), [databases](), [mysql](), [provisioning](), [replication](), [tls]() ### AI overview Appwrite announces native managed MySQL databases that can run inside the same project as an application's existing backend. The service supports MySQL 8.4 and 8.0, standard drivers and ORMs, TLS, backups and point-in-time recovery, replicas with automatic failover, database branches, scaling, and connection pooling. ### Source excerpt Provision a managed MySQL 8.4 instance inside your Appwrite project and keep using the drivers, ORMs, and tools your applications already depend on. ## Building patterns unlocked by scale to zero DevFeed: [Building patterns unlocked by scale to zero]() Original publisher: [Read original article]() Author: Carlota Soto Published: 2026-09-01T12:00:00Z Content type: article Language: en Sources: [Blog -- Neon Docs]() Topics: [Databases](), [autoscaling](), [ci]() Tags: [autoscaling](), [building](), [ci](), [compute](), [database](), [postgres](), [product](), [scale]() ### AI overview The article explains how autoscaling and scale to zero reduce idle Postgres compute costs. It presents ephemeral database branches for previews, tests, and individual development work as patterns enabled by suspending inactive compute. ### Source excerpt Most hosted Postgres pricing works like this: you pick an instance size, the vendor assumes the instance will run for about 730 hours a month, and you pay for those hours. Lakebase Postgres addresses that waste with autoscaling and scale to zero. ## Rerouting the Stream: How Lyft Moved to the Apache Flink Operator DevFeed: [Rerouting the Stream: How Lyft Moved to the Apache Flink Operator]() Original publisher: [Read original article]() Author: Maheep Myneni Published: 2026-08-31T19:08:16Z Content type: article Language: en Sources: [Lyft Engineering - Medium]() Topics: [Apache Iceberg](), [Streaming](), [Kubernetes](), [Open Source](), [autoscaling](), [Tooling](), [Architecture & Design]() Tags: [apache](), [apache-flink](), [autoscaling](), [data-platforms](), [infrastructure](), [kubernetes](), [lyft](), [open-source](), [production](), [streaming](), [streaming-data-processing](), [tooling]() ### AI overview Lyft describes its migration from an internally developed Flink Kubernetes operator to the open-source Apache Flink Kubernetes Operator. The move addressed maintenance burden, technical debt, feature gaps, outdated dependencies, and growing real-time streaming demands, while providing capabilities such as autoscaling, memory tuning, safer upgrades, and automatic rollbacks. ### Source excerpt Written by Maheep Myneni, Arda Kuyumcu, and Prem Santosh Udaya Shankar at Lyft. Why We Migrated: Technical Debt Meets Modern Streaming Demands Over the past several quarters, Lyft's Streaming Compute team retired our internally developed Flink Kubernetes operator and moved our entire streaming fleet onto the open-source Apache Flink Kubernetes operator. This post is about why we made the switch, how we pulled it off incrementally without disrupting users, and the follow-on work it took to actually get the benefits we were after. Back in 2020, when we first architected the Lyft Flink Kubernetes Operator, it was exactly what we needed. At that time, the open-source community hadn't yet built a dedicated control plane, so we built our own to manage all streaming applications on Kubernetes. It worked well for our initial workloads, but as our real-time data needs increased and our engineers' scope of ownership grew in both breadth and complexity, the cracks started to show. First came the maintenance burden. Our operator had become a relic of Lyft's early Kubernetes days, kept alive by a growing pile of custom code. Every Flink version upgrade meant carefully picking through layers of accumulated technical debt and hoping nothing broke on the way through. Second came the feature gap. Streaming tooling kept moving, and our engineers kept asking for capabilities that had become table stakes elsewhere, such as autoscaling to right-size jobs, automatic rollbacks on failed deploys, and an end to hand-tuning CPU and memory. Each request left us with two options, neither of which was ideal. We could explain why we couldn't support it yet, or spend weeks rebuilding something the open-source community had already shipped. Third was the dependency problem. We were pinned to outdated libraries. That doesn't break anything today, but it almost always creates new issues down the line. Security patches lagged, modern Kubernetes features stayed out of reach, and every quarter we waite ## Autoscaling Lakebase Postgres DevFeed: [Autoscaling Lakebase Postgres]() Original publisher: [Read original article]() Author: Carlota Soto Published: 2026-08-31T12:00:00Z Content type: article Language: en Sources: [Blog -- Neon Docs]() Topics: [Database](), [cpu]() Tags: [architecture](), [autoscaling](), [cache](), [compute](), [cpu](), [database](), [engineering](), [memory](), [postgres](), [storage]() ### AI overview The article explains how Lakebase Postgres autoscales database compute without stopping Postgres. Its architecture separates compute from durable storage, while the autoscaler derives a target size from CPU load, memory use, and compute-cache working-set size. ### Source excerpt Choosing a database instance size before you know the workload is an old building pattern. Lakebase Postgres omits the sizing experience altogether thanks to autoscaling. ## Kubernetes v1.37: Metrics API graduates to stable DevFeed: [Kubernetes v1.37: Metrics API graduates to stable]() Original publisher: [Read original article]() Author: ChengHao Yang Published: 2026-08-27T18:30:00Z Content type: article Language: en Sources: [Kubernetes Blog]() Topics: [Kubernetes](), [API](), [autoscaling](), [cpu](), [Monitoring]() Tags: [api](), [autoscaling](), [cpu](), [developers](), [kubernetes](), [memory](), [metrics](), [monitoring](), [production](), [release](), [resource](), [v1]() ### AI overview Kubernetes v1.37 promotes the Metrics API to the stable v1 version. The API provides CPU and memory usage for nodes and Pods, supporting commands such as kubectl top and resource-metrics-based autoscaling, without changing the collected metrics, resource types, or fields. ### Source excerpt Kubernetes v1.37 promotes the metrics.k8s.io API to stable (v1). This API provides CPU and memory usage for nodes and Pods, and is the API behind commands such as kubectl top and resource-metrics-based autoscaling. For cluster operators and application developers, this graduation means that the API now has the stability guarantees associated with a Kubernetes stable API. The v1 API has the same resource types and fields as v1beta1; this is an API-version graduation, not a change to the metrics that are collected or returned. A long-lived API reaches stable The resource Metrics API was introduced as alpha in Kubernetes v1.6 and became beta in v1.8. It has remained unchanged and has been used in production for years by clients including the HorizontalPodAutoscaler (HPA) and kubectl top. Kubernetes v1.37 formally graduates that proven API to metrics.k8s.io/v1. The API exposes two resource types: NodeMetrics, for CPU and memory usage for a node. PodMetrics, for CPU and memory usage for a Pod, with a per-container breakdown in its containers field. The API remains intentionally small. It provides the resource metrics needed for autoscaling and basic inspection; it is not a replacement for a full monitoring pipeline or the custom metrics (custom.metrics.k8s.io) API. What changed with the v1.37 release? The v1 API surface is identical to v1beta1, except for the API version. There are no renamed fields, new fields, or changes to the meaning of the returned CPU and memory values. For example, a client can retrieve node metrics from the stable endpoint: kubectl get --raw /apis/metrics.k8s.io/v1/nodes Likewise, it can retrieve metrics for the pods in a namespace: kubectl get --raw /apis/metrics.k8s.io/v1/namespaces/default/pods kubectl top supports both API versions. It prefers v1 when available and automatically falls back to v1beta1 on clusters that do not yet serve v1. The HPA controller currently supports only v1beta1. Support for discovery-based selection between v1 and v ## Durable Digest: August highlights DevFeed: [Durable Digest: August highlights]() Original publisher: [Read original article]() Author: Temporal Technologies Published: 2026-08-27T00:00:00Z Content type: release Language: en Sources: [Temporal Blog]() Topics: [Cloud](), [AI Development](), [AWS Lambda](), [Workers](), [autoscaling](), [Amazon Web Services](), [AI Agent](), [AWS IAM](), [SDKs]() Tags: [ai](), [ai-agent](), [announcements](), [autoscaling](), [aws](), [aws-lambda](), [cloud](), [distributed](), [durability](), [genai](), [google](), [iam](), [infrastructure](), [workers]() ### AI overview Temporal's August 2026 Durable Digest highlights new and preview features for running Workers on AWS Lambda, organizing Temporal Cloud resources, invoking Nexus Operations, and building durable AI applications. ### Source excerpt Highlights this month include new ways to run Workers without managing infrastructure, organize Temporal Cloud resources and build more durable AI applications. ## Why GitHub feels less reliable lately DevFeed: [Why GitHub feels less reliable lately]() Original publisher: [Read original article]() Author: Sridhar Rajarao Published: 2026-08-23T00:00:00Z Content type: opinion Language: en Sources: [Sridhar Rajarao]() Topics: [GitHub](), [incident](), [migration](), [systems](), [Azure](), [GitHub Actions](), [pull-requests]() Tags: [autoscaling](), [capacity](), [concurrency](), [data-centers](), [deployment](), [github](), [incident](), [incident-management](), [istio](), [reliability](), [request](), [resilience](), [sre](), [transformation]() ### AI overview The article argues that GitHub's recent reliability problems reflect the difficult middle of a major infrastructure transformation. It connects incidents to migration complexity, unsafe automation, configuration mistakes, capacity and concurrency weaknesses, database migration errors, and autoscaling problems. ### Source excerpt GitHub is not having one outage problem. Its recent incident reports show the difficult middle of a platform transformation. ## A Tale of Two Flink Autoscalers DevFeed: [A Tale of Two Flink Autoscalers]() Original publisher: [Read original article]() Author: Netflix Technology Blog Published: 2026-08-21T16:01:01Z Content type: article Language: en Sources: [Netflix](), [Netflix TechBlog - Medium]() Topics: [apache-flink](), [autoscaling](), [Netflix](), [stream-processing](), [Kafka](), [Open Source](), [telemetry](), [Network]() Tags: [apache-flink](), [autoscaling](), [aws](), [infrastructure](), [kafka](), [metrics](), [netflix](), [open-source](), [operational-excellence](), [scale](), [stream-processing](), [telemetry]() ### AI overview Netflix describes running two Apache Flink autoscalers: an older in-house system and a newer one from the Apache Flink community. The article explains why autoscaling is necessary for more than 30,000 Flink jobs across AWS regions, discusses workload variability and the cost of resizing stateful jobs, and introduces the metrics-driven design of the first autoscaler. ### Source excerpt Samuel Yeboah, Francesco Di Chiara and Mingliang Liu Today, Netflix runs two Flink autoscalers. That is exactly one more than we want. We built the first one in-house years ago, when there was no mature option suited to our platform. The second came from the Apache Flink community, and it can scale workloads our homegrown system was never designed for. We now run both in production and are steadily converging on the open-source one. Along the way we learned some hard lessons about metrics, cost, and the real price of maintaining infrastructure you could instead adopt, and we hope they are useful whether you run a handful of Flink jobs or tens of thousands. Why autoscaling is not optional at our scale Netflix has run stream processing on Apache Flink since 2017. As of 2026 we operate more than 30,000 Flink jobs across multiple AWS regions. Most are not deployed by hand; they are generated by our managed platform Data Mesh, so the majority of users never touch a Flink job directly. A smaller but growing set are custom jobs, built and operated by teams across the company for use cases like personalization, Ads, and Live events. They range from single-operator jobs that shuttle records between Kafka topics to stateful pipelines with branches, joins, and terabytes of state, and their load swings with daily cycles, launches, and regional failovers. Provisioning every one of those jobs for its peak is wasteful; provisioning for the average causes lag during surges. And in our platform a scaling action is not free: by default it means taking a savepoint, stopping the job gracefully, and restarting it at the new size, which for a large stateful job can take minutes. That leaves a genuinely hard question: how do you give each job the resources it needs, when it needs them, without a human in the loop and without breaking anything? The first autoscaler: watching from outside Our first answer, built around 2019, was an autoscaler shaped like a stream-processing job. It ran on M ## Neon for Herdr: A Neon pane in your Herdr workspace DevFeed: [Neon for Herdr: A Neon pane in your Herdr workspace]() Original publisher: [Read original article]() Author: Dominik Koch Published: 2026-08-13T12:00:00Z Content type: release Language: en Sources: [Blog -- Neon Docs]() Topics: [SDKs](), [TypeScript](), [Terminal](), [dashboards](), [OAuth](), [Databases](), [autoscaling](), [Command-line interface]() Tags: [autoscaling](), [browser](), [community](), [databases](), [linux](), [macos](), [oauth](), [typescript]() ### AI overview Neon for Herdr is a terminal workspace plugin that adds a Neon project-management pane built with Ink, Effect, and the Neon TypeScript SDK. It supports project and branch operations, compute controls, connection-string retrieval, fuzzy search, and OAuth authentication with PKCE. ### Source excerpt I just released Neon for Herdr, a Herdr plugin that puts your Neon projects in a pane next to whatever you are already doing, so creating a branch, grabbing its connection string, or suspending a compute is only a keypress away. ## Fair by design: orchestrating background jobs in Ruby DevFeed: [Fair by design: orchestrating background jobs in Ruby]() Original publisher: [Read original article]() Author: Travis Turner (richardturner@evilmartians.com) Published: 2026-08-11T00:00:00Z Content type: tutorial Language: en Sources: [Evil Martians]() Topics: [Ruby](), [jobs](), [Latency](), [autoscaling]() Tags: [autoscaling](), [how-to](), [jobs](), [latency](), [performance](), [rails](), [redis](), [ruby](), [sidekiq]() ### AI overview This tutorial examines fairness in Ruby background-job processing. It explains how queue latency affects quality of service, why adding workers or autoscaling may be limited by shared resources and operational cost, and introduces background-job prioritization as a way to address bottlenecks. ### Source excerpt Are you treating your users fairly? They could be stuck in the queue while a greedy user monopolizes resources. And you might not even know it! In this post, you'll see if it's time for you to take background job prioritization seriously, and how to make it fair for all users. ## Pulling multi-gigabyte container images in seconds on Amazon EKS DevFeed: [Pulling multi-gigabyte container images in seconds on Amazon EKS]() Original publisher: [Read original article]() Author: Sri Saran Balaji Vellore Rajakumar Published: 2026-08-10T16:00:00Z Content type: article Language: en Sources: [Kubernetes Overview, News and Trends | The New Stack]() Topics: [Amazon Elastic Kubernetes Service](), [container images](), [Machine learning](), [autoscaling](), [CUDA]() Tags: [ai-infrastructure](), [amazon-eks](), [autoscaling](), [aws-marketplace](), [container](), [container-images](), [containers](), [cuda](), [kubernetes](), [machine-learning](), [post-contributed](), [sponsor-aws-marketplace](), [sponsored-post-contributed]() ### AI overview This article explains how a production ML platform on Amazon EKS reduced multi-minute pulls of roughly 30 GB container images to seconds. Profiling found that software use of available hardware, rather than network bandwidth or the registry, was the main constraint; the resulting changes are available by default in EKS Auto Mode and were contributed upstream to containerd and the SOCI snapshotter. ### Source excerpt When the image is the bottleneck: Machine learning changed what a container image looks like. A typical application ships in The post Pulling multi-gigabyte container images in seconds on Amazon EKS appeared first on The New Stack. ## Managed Queues: Autoscaling Queue Workers on Laravel Cloud DevFeed: [Managed Queues: Autoscaling Queue Workers on Laravel Cloud]() Original publisher: [Read original article]() Author: Laravel Team Published: 2026-08-06T15:15:00Z Content type: article Language: en Sources: [Laravel Blog]() Topics: [autoscaling](), [Laravel](), [Cloud](), [dashboards]() Tags: [autoscaling](), [cloud](), [compute](), [cost](), [feature](), [infrastructure](), [laravel](), [operations](), [scale]() ### AI overview Laravel Cloud's managed queues run isolated queue workers that autoscale according to queue pressure, scale to zero when idle, and restart quickly when jobs arrive. The feature also improves visibility into failed jobs and charges for compute and queue operations used. ### Source excerpt Laravel Cloud's managed queues autoscale workers on queue pressure, surface failed jobs in a built-in dashboard, and scale to zero when idle. ## Fixed cadence to seconds: making ClickHouse Cloud autoscaling more reactive DevFeed: [Fixed cadence to seconds: making ClickHouse Cloud autoscaling more reactive]() Original publisher: [Read original article]() Author: Marvin Beckers Published: 2026-08-04T13:19:15Z Content type: article Language: en Sources: [ClickHouse Blog]() Topics: [Orchestration](), [Kubernetes]() Tags: [autoscaling](), [clickhouse](), [cloud](), [concurrency](), [cpu](), [event](), [kubernetes](), [latency](), [memory](), [orchestration](), [reactive](), [real-time](), [resource](), [scale]() ### AI overview ClickHouse describes rebuilding Cloud autoscaling orchestration to combine periodic recommendation passes with a reactive fast path for urgent scale-up signals. ### Source excerpt How we rebuilt ClickHouse Cloud's autoscaling orchestration on Kubernetes' controller-runtime and a ClickHouse-powered signals table, adding a reactive fast path that scales services up in seconds instead of waiting for the next scheduled pass. ## What's new in clickhousectl v0.4.0 DevFeed: [What's new in clickhousectl v0.4.0]() Original publisher: [Read original article]() Author: Al Brown Published: 2026-07-31T11:44:09Z Content type: release Language: en Sources: [ClickHouse Blog]() Topics: [clickhouse](), [Command-line interface](), [autoscaling](), [Cloud](), [Kafka](), [MySQL](), [API keys]() Tags: [api-keys](), [autoscaling](), [cli](), [clickhouse](), [cloud](), [cloud-services](), [kafka](), [mysql](), [telemetry]() ### AI overview clickhousectl v0.4.0 introduces horizontal autoscaling for ClickHouse Cloud services, beta ClickPipe schema discovery for Kafka and Kinesis sources, more controls for object storage and MySQL/Postgres ingestion, global OAuth token storage, safer local server lifecycle behavior, and opt-out anonymous usage telemetry. ### Source excerpt clickhousectl v0.4.0 adds horizontal autoscaling for Cloud services, ClickPipe schema discovery, finer control over ClickPipe ingestion, and opt-out anonymous usage telemetry. ## Announcing Serverless Workers for Google Cloud Run DevFeed: [Announcing Serverless Workers for Google Cloud Run]() Original publisher: [Read original article]() Author: Brandon Chavis Published: 2026-07-30T00:00:00Z Content type: release Language: en Sources: [Temporal Blog]() Topics: [Cloud Run](), [autoscaling](), [Serverless](), [Google Cloud Platform (GCP)](), [Deployment](), [container](), [IAM]() Tags: [autoscaling](), [cloud-run](), [deployment](), [docker](), [google-cloud-run](), [iam](), [product-news](), [serverless]() ### AI overview Temporal announces Serverless Workers support for Google Cloud Run. The service integrates with Cloud Run Worker Pools and dynamically adjusts capacity, including scaling to zero when appropriate, reducing the infrastructure planning and autoscaling work required from users. ### Source excerpt Temporal now supports Serverless Workers on Google Cloud Run, autoscaling Worker Pools for you so you never manage infrastructure or pay for idle compute. ## ModelExpress: Distributing Model Artifacts at the Speed of Light DevFeed: [ModelExpress: Distributing Model Artifacts at the Speed of Light]() Original publisher: [Read original article]() Author: Elizabeth Goodman Published: 2026-07-24T16:45:00Z Content type: article Language: en Sources: [NVIDIA Developer](), [NVIDIA Technical Blog]() Topics: [model-deployment](), [GPU](), [P2P](), [Amazon S3](), [autoscaling](), [Caching](), [cloud-infrastructure]() Tags: [agentic-ai-generative-ai](), [ai-foundation-models](), [ai-inference](), [autoscaling](), [data-center-cloud](), [developer-tools-techniques](), [dynamo-triton](), [featured](), [gpu](), [inference](), [inference-performance](), [jit](), [llms](), [local](), [model](), [s3](), [scale](), [speed](), [streaming]() ### AI overview NVIDIA ModelExpress (MX) reduces model-serving startup time by locating compatible model weights at the fastest available source. It prioritizes direct GPU-to-GPU transfer over P2P RDMA from an existing serving replica, while using object or local storage for initial bootstrapping and fallback. The article also describes reusing JIT kernel caches and distributing reinforcement-learning weight updates. ### Source excerpt Every byte moved has a cost. As model checkpoints grow to hundreds of gigabytes or even a terabyte, that cost adds up quickly. To make things even worse, moving... [Next page]()