Linux kernel developer Jiri Slaby announced the release of the fifty-eighth maintenance build of the long-term supported Linux 3.12 kernel series, as well as modification to its EOL status.
Read more at Softpedia
Arch Linux Operating System Is Now Powered by Linux Kernel 4.5
As of this morning, April 14, 2016, the famous Arch Linux operating system has received a major kernel upgrade, Linux 4.5, which users can now install via the official channels.
Read more at Softpedia.
RDO Delivers OpenStack Mitaka for CentOS Linux
This week, the RDO community announced the general availability of its freely-available, community-supported distribution of OpenStack, the popular open source project for building private, public, and hybrid clouds.
Read on to learn about some of what’s new in the Mitaka release of RDO, or head over to the RDO Project site to download the software and meet the community.
Management & Deployment
Simplified cloud management and deployment is one of the headline themes for the Mitaka OpenStack release. For instance, the teams behind the Nova compute and Keystone identity services have managed to trim some of the required configuration elements, replacing them with additional default settings, as can be seen by comparing the current and previous versions of the Nova installation guide.
Elsewhere in Nova, the Mitaka release brings improvements to the functionality and reliability of virtual instance live migration, with support for pausing and aborting live migrations, and for providing progress reports on these migration operations.
On the deployment front, RDO continues to offer the packstack tool for producing quick, proof of concept OpenStack installations on limited hardware. To get up and running quickly with a more production-appropriate deployment of RDO, there’s a quickstart install option based on the OpenStack TripleO project (previously known as RDO Manager) that builds on the Nova, Ironic, Neutron, and Heat components to automate cloud management.
For still another take on OpenStack deployment, the Kolla project provides images and deployment tools for operating OpenStack clouds with components running in Docker containers. The Kolla project defaults to RDO as a base for its container images, and uses Ansible scripts for deployment. New for Mitaka, Kolla has added Ansible scripts for conducting service upgrades and centralized log aggregation using the ELK stack of ElasticSearch, LogStash, and Kibana.
User Experience
OpenStack Mitaka also promises to improve general user experience, with enhancements both to the project’s Horizon web dashboard application and to the command line tools available to end users and operators alike. Users of the Horizon web interface can expect improved performance thanks to wider use of Angular.JS and of template caching across the project.
Where individual stack components have previously hewed to their own separate command line client applications, the latest release includes a major push toward a common cli application, called OpenStack Client, that provides a consistent set of calls for creating resources so that end users can apply their knowledge from familiar components to additional OpenStack services, without having to learn the quirks of individual service APIs.
Scalability
As OpenStack racks up more time running the cloud environments of large users, the project has accrued various enhancements related to operating at scale, and this is another theme of the Mitaka release.
The Nova compute service now offers a disk space scheduling filter, which enables Nova to take free disk space into consideration, alongside available RAM, when allocating new instances on compute hosts. The project has also firmed up Nova’s Cells v2, a feature first introduced in OpenStack Liberty that allows admins to scale a compute cloud in a distributed fashion without relying on complicated database and message queue clustering.
OpenStack’s Cinder block storage backup component has been decoupled from the volume storage component, allowing both for greater scalability and for improved data protection, as well as opening the way for new volume backup options such as the OpenStack Cinder backup driver for Google Cloud Storage driver announced this week.
Getting Involved
The RDO project will be holding a pair of test days on April 13-14, 2016. The testing will be coordinated through the #rdo channel on Freenode, through the RDO project website and the rdo-list mailing list. Drop into the IRC room or introduce yourself on the mailing list to connect with other RDO community members as they explore this new release.
Beyond the test days, if you’re interested in getting more involved with packaging, building and testing OpenStack (and other cloud platforms) on CentOS Linux, check out the CentOS Cloud SIG, which provides build and continuous integration resources for the RDO effort.
Working Late May Be Destroying Your Organization: Colin McNamara [Video]
Many of us who work in the IT field are aware of the grim reality of working late at night. People often end up working long hours as they take on additional work and projects. But, is that good for you? Is it good for your organization? Is it good for your teams and clients? By doing so, are you helping your company or hurting it?
That was the topic of the keynote that Colin McNamara, Practice Director, DevOps at Dimension Data, delivered at DevOps Networking Forum.
Looking at the raised hands, McNamara said, “you might be destroying your organizations. You might be destroying your teams.”
He cited an example of when you are working on a project and suddenly your team is asked to add new features at the last minute. Every time you say yes, you are basically agreeing to take on additional work and deliver in the same amount of time.
That situation can set things off the rails. Now you have more code to deliver in the same amount of time. Things can get so bad that the project you previously could have delivered ahead of time won’t even be delivered on time anymore. Now your teams are not happy; they are grumpy. They are working over time, and the coordination between teams has gone downhill because some people are working late at night, during odd hours.
In a nutshell, this habit is counterproductive: It’s damaging to you, your teams, your organization, and your clients.
Is there a solution? There are many, actually, and McNamara talked about some of them in his keynote, suggesting several ways to maintain a balance and keep energy levels high.
To avoid late-night work, he suggested implementing the 11/7 rule, which essentially means no coding between 11 p.m. and 7 a.m. That takes care of a lot of things. You don’t work late at night, so you get more sleep and you spend more time with family and friends. All of that recharges you. At the same time, because you are not working while the rest of your team is sleeping, there is no communication disruption. The result? Increased productivity.
McNamara also suggested a tip that many of us likely received before taking a test in school: Get a good night’s sleep. Eat well.
If you are continuously writing code, you will burn out. To avoid that, McNamara uses a strategy that comes from the Marine Corps: It’s a 72-hour stand. McNamara said that after 72 hours of coding, it’s time for rest, “Do not open your laptops. Go spend time with your family, as we don’t want you being divorced.”
McNamara’s teams noticed significant improvements when adopting his advice, “Four days later, our guys had started to think outside the box again.” Don’t you want your teams to be thinking outside the box?
Watch all 13 keynote videos from DevOps Networking Forum.
