Google Cloud’s recent outage offers system and application architects a teachable moment to illustrate how to better prepare your organization for such failures.
Not sure where to begin when thinking about protecting your data across multiple clouds and services? Ned Bellavance shares his “wish list” of what his perfect solution would include. Be sure to hear even more in Ned’s recent webinar with Commvault, Managing Multi-Cloud Data Protection.
Maish Saidel-Keesing makes the case for the end of Docker, focusing on the loss of the container orchestration market, competitive runtimes, and a less engaged community. It’s an interesting read and shows how quickly the industry narrative around a company can change.
Backblaze seems to relish providing transparency into their infrastructure. We’ve been sharing their annual drive failure reports whenever we come across them, because it gives a fascinating real world look into overall drive reliability.Now the company has given an overview of how their Storage Pod rack infrastructure has evolved over time.
File system performance is often something that goes unconsidered, but can have a noticeable impact. Andre Leibovici recently compared Oracle performance on XFS versus EXT4. The results show why you might want to give your file system of choice a little closer look.
When I think of a NAS, I generally think of it in terms of a storage device. Then again, that kind of ignores the whole “network-attached” part!
That may seem like a minor point, but not when it comes to measuring data and transfer speeds. Confusing storage bytes with networking bits doesn’t just cause confusion, it might have implications for what is actually possible with your NAS. Tony Bourke put together an excellent primer to straighten everything out.
It’s become a media narrative that there is a “talent gap” in cybersecurity. The story goes that there are plenty of good paying jobs, but not enough qualified people to do them. On the face of it, there seem to be good reasons. The spate of high profile data leaks and hacks have made security a front and center concern for organizations of all sizes. As the sophistication of these attacks increase, it makes sense that the number of jobs needed would also increases. But Ben Tomhave can’t stand this narrative. He makes the case that this talent gap is actually symptomatic of organizational dysfunction.
In terms of repurposing old computer hardware, it would be hard to pick a project that would bring a bigger smile to my face than retrofitting an Intel Core2Duo-based Sun workstation to use modern AMD EPYC platform. It’s the best of both worlds, modern hardware in a chassis with perhaps the best logo in all of tech.
Kubernetes is increasingly important in IT, but that doesn’t make it any less opaque. Luckily Eric Shanks put together an excellent series of posts to get you started. This first goes over the theory behind different elements, then jumps into specific practical examples.
The Raspberry Pi Foundation released a new iteration on their module enterprise board, the Compute Module 3+. This adds an upgraded processor, configurable flash storage, and guaranteed availability through 2026. While Raspberry Pi units are never interested in bleeding edge performance, it’s a nice upgrade for organizations looking for a drop-in replacement.