Is “observability” just “instrumentation”? Or, monitoring sucks? No, you suck.
Software Defined Talk LLC·2017-09-22·via Software Defined Talk
Episode 106
·
September 22nd, 2017
·
59 mins 11 secs
The DevOps kids have decided to come up with a new term “observability.” We get to the bottom of the WTF barrel on what that is - it sounds like a good word-project. Also, there’s a spate of kubernetes news, as always, and some interesting acquisitions. Plus, a micro-iOS 11 review.
Meta, follow-up, etc.
Patreon - like anyone who starts these things, I have no idea WTF it is, if it’s a good idea, or if I should be ashamed. Need some product/market fit.
Check out the Software Defined Talk Members Only White-Paper Exiguous podcast over there.
This guy: “Thinking directionally, Monitoring is the passive collection of Metrics, logs, etc. about a system, while Observability is the active dissemination of information from the system. Looking at it another way, from the external ‘supervisor’ perspective, I monitor you, but you make yourself Observable.”
So, yes: if developers actually make their code monitorable and manageable…easy street! It’s a good detailing of that important part of DevOps.
Cloud Native Java has a good example with the default “observability” attributes for apps, and then an overview of Zipkin tracing.
Mesosphere adding K8s support - “Guagenti also noted that he believes that Mesosphere is currently a leader in the container space, both in terms of the number of containers its users run in production and in terms of revenue (though the company sadly didn’t share any numbers).”
"I think it’s fair to call Kubernetes the de facto standard for how enterprises will do container orchestration,” Derrick Harris.
Meanwhile, an abstract of a containers penetration study, from RedMonk: "Docker, is running at 71% across Fortune 100 companies. Kubernetes usage is running in some form at 54%, and Cloud Foundry usage is at 50%”
This update from the Cloud Foundry Foundation is a little more, er, “responsible” in pointing out flaws. Instead it just says there’s lots of growth and tire-kicking: 2016/2017 y/y shows those evaluating containers went up from 31% to 42%, while “using” ticked up a tad from 22% to 25%, n=540.
Oracle’s in the CNCF club! K8s on Oracle Linux, K8s for Oracle Public Cloud. “At this point, there really can’t be any doubt that Kubernetes is winning the container orchestration wars, given that virtually every major player is now backing the project, both financially and with code contributions.”
Rackspace acquires Datapipe “The reason we’re buying them is that we want to extend our leadership in multi-cloud services,” Rackspace chief strategy officer Matt Bradley told me. “It’s a sign and signal that we’re going for it.” Bradley expects that the combined company will make Rackspace the largest private cloud player and the largest managed hosting service.
“In May, the company launched its Kubernetes dashboard K8S. It allows users to connect repositories, build images from source, then deploy them to that Kubernetes cluster. You can also set up automated pipelines to push images from one cluster to another, promote software from test/dev to prod, quickly roll back and do all this in the context of one or more Kubernetes clusters… The Kubernetes service is offered as a hosted service or in an on-prem version. It provides notifications through Slack.”
Is there anything to do here? Setup layers of credit cards? Require Touch ID (etc.) approval of all financial decisions and transactions in your “account”? Food & Safety like inspectors for security?
This guy has written a big Solaris-brain to Linux-brain manifesto/guide, plus: “[n]owadays, Sun is a cobweb-covered sign at the Facebook Menlo Park campus, kept as a warning to the next generation.” SICK BURN!
Layoffs and more: “In particular, that employees who had given their careers to the company were told of their termination via a pre-recorded call — “robo-RIF’d” in the words of one employee — is both despicable and cowardly.”
