There is a lot of interesting buzz around specialized container hosts, rump kernels, and unikernels because they hold the potential to revolutionize certain workloads (embedded, cloud, etc.). Keep your eye on this exciting, fast moving space, but cautiously.
Currently, unikernels seem quite similar to building printed circuits. They require a lot of upfront investment to utilize and are very specialized, providing benefits for certain workloads. In the meantime containers are quite interesting even for conventional workloads and don't require as much investment. Typically an operations team should be able to port an application to containers, whereas it takes real re-engineering to port an application to unikernels and the industry is still not quite sure what workloads can be ported to unikernels.
Here's to an exciting future of containers, rump kernels, and unikernels!
A few days ago when delivering benchmarks of the new CPUFreq "Schedutil" governor in Linux 4.7 the P-State comparison results on this Git kernel looked particularly terrible. I've since done some P-State tests on the same system using the Linux 4.5 and 4.6 kernels that further point towards a regression having taken place.
Meet Linus Torvalds, the man who created Linux software, which runs in millions of computers all around the world. Hear the story of how he invented the software in this awesome TED Talk.
Bryce Harrington announced the release today of the release candidates for Wayland 1.11 and the reference Weston 1.11 compositor.
The official 1.11 release candidate announcements can be found via the Wayland-devel mailing list.
Convergence is not a word on everybody's lips. But if Canonical Software, the company that controls Ubuntu, has any say, it soon will be.
Others may be more skeptical.
Canonical describes convergence as "a single software platform that runs across smartphones, tablets, PCs, and TVs. It is designed to help make converged computing a reality: one system, one experience, multiple form factors."
Whom can I recommend CentOS to? Probably to people I mentioned in the very beginning of this article: students who want to dedicate their life to system administration. You need to learn how to search for the answers, how to do things manually. That is your profession. That is your bread. CentOS gives you a brilliant opportunity to learn all of that along with learning the system itself.
But CentOS is not for home users who want things done quickly and easily, I'm afraid.
I’m not going to argue that the Raspberry Pi should always be the device of choice for every situation. Sometimes it just doesn’t cut it and using it in a given situation will cause more work than necessary. Whenever I am asked the above question, I usually get the details of what the person is intending to do, and then talk about the pros and cons of the Raspberry Pi for that use. One of the things I always remind the person is that no matter how good Device X might be, you need to consider the community behind the device. In my opinion, a constantly growing supportive community is what the Pi offers over all other devices.
RapidDisk is an open-source and enhanced Linux RAM drive solution led by BDFL Petros Koutoupis (who also writes for Linux Journal) that allows users to create, resize and remove RAM drives dynamically or map those same RAM drives as a cache to slower data volumes. The latest version 4.0 release adds a series of complementary improvements, such as kernel module optimizations, code cleanup/redesign and bug fixes. RapidDisk consists of a collection of kernel modules, an administration utility, high-availability scripts and a RESTful API for third-party integration. By design, RapidDisk volumes are thinly provisioned and will allocate memory only upon usage.