Docker provides a powerful Logging feature for containerized applications that allows you to define how the logs should be handled.
In this short article, we will demonstrate how we can take advantage of this, collecting logs from a Docker Container in real time to aggregate them back into an Elasticsearch database. For these purposes, we will introduce Fluent Bit, an open source and lightweight data collector for Linux.
Read more at Fluent Bit.
This post into delves into the non-technical aspects of adopting microservices within a company. With the world now being driven by technology, companies must learn to adapt, stay agile and continue to increase velocity in their core business.
The transition towards a microservice architecture are usually thought of as a process driven by technical limitations of an existing system. While that’s true in most cases, many of the other reasons for moving in that direction are led by higher level requirements related to the business and team dynamics.
The post will cover motivations, the migration path, what success may actually look like and the tradeoffs which are made in such a transition.
Read more at Asim Aslam’s Blog
Kroah-Hartman presented this talk to Google’s Kubernetes development team. Kubernetes is also undergoing rapid growth, and Kroah-Hartman draws on his extensive experience to provide tips on how to manage such a high-velocity project.
Linus Torvalds’ travel plans mean version 4.7 of the Linux kernel will be delayed by a week.
“We’ve had a nicely calm week, which is what I expected – the last rc really was bigger just due to random timing issues, and not some worrying pattern about this release cycle,” Torvalds wrote on Sunday.”
Read more at Softpedia
The Crate Mesos Framework, developed as an open source project, integrates the data storage layer with Mesos. It is a good fit for stateful services that run on top of Mesos. It avoids operations teams to manually install Crate on their cluster with resources, such as disk space, statically partitioned.
Dhilip Kumar S of Huawei Technologies shares how he built a thin, high-performing Redis framework on Mesos, which delivers Redis’s good performance and simplifies running it in a cluster.
A scalable and adaptable machine learning platform is essential for an organization to harness the full potential of their data. This talk outlines how Docker, Spark, Hadoop and several other building blocks can be integrated into a machine learning framework on Mesos.
The openSuSE project offers two distributions: Tumbleweed, which is a rolling distribution that gets continuous updates, and Leap, which is a point distribution that gets periodic updates.
Looking at it a different way, I think of Tumbleweed as being a development distribution, so I expect it to get the latest version of all its major packages very quickly, but I am not surprised when there is some minor instability. I consider Leap to be a stable distribution, so some of the major/critical packages only get updates when a new point release is made, and I expect it to be very dependable.
Read more at ZDNet
A new open source PaaS, Tsuru, is out to ease the application deployment process by reducing it to little more than a Git push command.
The workflow for Tsuru, according to its documentation, consists of writing an app, backing it with resources like databases or caching, and deploying it to production with Git. Tsuru handles the rest, including crating up the apps in Docker containers and managing their workloads. Its creators claim it can be deployed both locally and on services like AWS, DigitalOcean, or Apache CloudStack.
Read more at InfoWorld
The speed at which data is generated, consumed, processed, and analyzed is increasing at an unbelievably rapid pace. Social media, the Internet of Things, ad tech, and gaming verticals are struggling to deal with the disproportionate size of data sets. These industries demand data processing and analysis in near real-time. Traditional big data-styled frameworks such as Apache Hadoop is not well-suited for these use cases.
As a result, multiple open source projects have been started in the last few years todeal with the streaming data. All were designed to process a never-ending sequence of records originating from more than one source. From Kafka to Beam, there are over a dozen Apache projects in various stages of completion.
With a high overlap, the current Apache streaming projects address similar scenarios. Users often find it confusing to choose the right open source stack for implementing a real-time stream processing solution. This article attempts to help customers navigate the complex maze of Apache streaming projects by calling out the key differentiators for each. We will discuss the use cases and key scenarios addressed by Apache Kafka, Apache Storm, Apache Spark, Apache Samza, Apache Beam and related projects.
Read more at The New Stack
