Archives

I was on PTO this week. I was working on a number of side projects and issues I’ve been trying to resolve for quite some time. This includes soft launching a podcast! Deep Cuts is a podcast that looks at the news behind the news. Things that were significant but didn’t make the newsletter for one reason or another. Prepare to embrace People, Process, and Tools in your ear holes. I say soft launch because it’s important for me to take things iteratively. Anchor is the platform for the moment. They’re working the difficult task of getting the podcast listed in all the major directories. Iterative improvements will be the name of the game. Feedback on anything I do is always appreciated. How companies adopt and apply cloud native infrastructure–from O’Reilly Survey results reveal the path organizations face as they integrate cloud native infrastructure and harness the full power of the cloud. SPONSORED DevOps’ish Last Week’s Top Five How I use Slack—alone—to get more done After 4 years with nginx, we switched to Caddy - Here is why I spend too much time in Zoom… A deep dive into Linux namespaces How SRE teams are organized, and how to get started People DevOps for doubters: How to deal with 9 kinds of people who push back — This article is stuffed with quotes about dealing with all sorts of personalities practitioners have experienced. I was one of many quoted in the article. ...

I spent a lot of time this week struggling with an odd problem. How do you make a multicloud platform without having to do a ton of work? I was trying to figure out why it’s so damn hard to get a multicloud platform going. People want this for various reasons but, struggle with it. Why? What was I trying to do? Host the DevOps’ish web site where it was less likely to be blocked based on a visitor’s geography. The idea was if someone were in Hong Kong they’d hit a bucket in Alibaba Cloud. The US and Europe would likely draw from Google Cloud Storage. Everywhere else was going to fall to the algorithms. Turns out Route53 doesn’t let you route traffic based on geography to anything outside of AWS. Same for Google’s Cloud DNS. Digital Ocean spaces aren’t quite ready for prime time yet either. Folks have to pick a cloud or build a platform across them. This is why Kubernetes is such a big deal. This is why Google’s dominance in the Kubernetes space matters so much. No one is going to be able to work a cloud providers primitives into the ANYCAST hybrid multicloud of their dreams. But if something is built on top of the primitives magic might be possible. For now, though, I’m using Google Cloud CDN. At least it’s not blocked in most places. Turns out there are some handy sites for testing a domain’s accessibility in other nations: ...

2019 State of Multicloud A Report on the Underlying Dynamics Fueling Multicloud Strategies. Download Today! SPONSORED [Webinar] Every commit should have an issue ticket number Referencing an issue ticket in each commit is a development best practice. It improves code reviews, creates audit trails, and keeps you compliant. Learn how to implement this scalably with CircleCI + Datree.io. SPONSORED DevOps’ish Last Week’s Top Five Kubernetes for the impatient 10 YAML tips for people who hate YAML How to get started with site reliability engineering (SRE) GitHub shocks top developer: Access to 5 years’ work inexplicably blocked Automate Kubernetes with GitOps People The worst morale boosting gesture I’ve experienced — This might be the single dumbest morale booster I’ve ever heard (and I’ve seen some really bad ones). When Your Job Is Your Identity, Professional Failure Hurts More — “…when you take professional kicks personally you compromise your own ability to recover and see the bigger picture…” Perhaps you are the one doing the kicking and your organization is reacting accordingly. ...

Dear reader, lend me your eyeballs. Let’s discuss the DevOps’ish Summer 2019 Survey for a brief moment. It’s vital feedback that I will use to pivot, fork, or modify the newsletter if needed. Therefore it’s important that you take the survey today! I’m in San Francisco this week for JFrog swampUP. If you’re around the Red Hat office on Monday, Embarcadero Center Tuesday and Wednesday, or going to the SF SRE Meetup Wednesday night and want to chat, let me know! 2019 State of Multicloud A Report on the Underlying Dynamics Fueling Multicloud Strategies. Download Today! SPONSORED [Datree.io] Trace commits to tickets Tracing commits to JIRA tickets is not only a development best practice, but also a mandatory policy for compliance standards like SOC 2. Learn how to integrate policy checks into your CircleCI pipeline using Datree in this webinar. SPONSORED Have an upcoming webinar, event, conference, or eBook you want to publicize? Ever thought about putting it in front of 3,000+ insanely sharp admins, engineers, technical marketers, and managers? This ad slot is open. Sponsor DevOps’ish today! ...

This week’s image was something I whipped up when talking to a friend. When I explained that running Kubernetes on a Raspberry Pi cluster or Digital Ocean didn’t mean they had to be an expert at running Kubernetes, it tilted their world off axis a little. I blew away a cluster and rebuilt it from a handful of files in less than thirty minutes a couple of weeks ago. This should be a rare occurrence (and there should be backups). But, if there isn’t time to troubleshoot (and you maintain proper configuration management), burn it down and start over if you make a mistake. It points to the practice Rob Hirschfeld discovered of keeping Kubernetes clusters small to reduce blast radius and snowflakes. 2019 State of Multicloud A Report on the Underlying Dynamics Fueling Multicloud Strategies. Download Today! SPONSORED DevOps’ish Summer 2019 Survey Please take this year’s survey. It is a handful of questions, will provide actionable feedback, and be greatly appreciated. The intent is to improve the overall experience and increase the value this newsletter provides to its readers. Take the survey today! ...