Meta Open Source’s Post

When you're running your tests in an environment like Docker, you might have seen or even passed the following parameters in your Chrome options - -- disable-gpu -- no-sandbox -- disable-dev-shm-usage But what are these and why do we use it? -- disable-gpu : GPU hardware acceleration offloads certain graphical operations to the GPU (Graphics Processing Unit), making these operations faster and freeing up CPU resources for other tasks. But GPU isn’t usually available inside a Docker container, unless you’ve specially configured the host. Setting this flag explicitly instructs Chrome not to try and use GPU-based rendering. -- no-sandbox : These disable Chrome’s sandboxing. Sandbox is a security measure that separates running processes. It's used to prevent malicious or malfunctioning software from damaging or snooping on the system. Each tab in Chrome, for example, runs in its own sandbox, isolating its process from other tabs and the rest of the system. -- disable-dev-shm-usage : Shared memory, or /dev/shm, is used as a temporary storage area by programs running on your computer. Some applications, like web browsers, use it intensively. If an application runs out of shared memory, you might see errors or crashes. When you use Docker, the amount of shared memory is much smaller than on a regular machine. This asks Chrome will write into /tmp instead. #knowledgesharing #testautomation #testing #concepts #docker #selenium #seleniumdocker

I encountered an error while running a Docker container: FATAL ERROR: Ineffective mark-compacts near heap limit Allocation failed - JavaScript heap #out_of_memory After researching, I found this article very helpful #DOCKER

So this announcement went out last week. I believe the momentum around RISC-V is approaching critical mass and we are likely to see a long tail of hardware announcements. This announcement is interesting because it's around the SW ecosystem, a project that could dramatically change the trajectory of RISC-V in the market. Check it out.. https://riseproject.dev/ #riscv #software #ecosystem #rise #hardware

DHH continues upending the prevailing style of software development by re-shoring just about everything outta the cloud... "To me, the most satisfying part of the improved performance of modern developer CPUs is the possibility to simplify our stacks. Installing, operating, and caring for a remote CI setup is a substantial complication. Either you do it on your own hardware, and deal with that complexity directly, or you pay through the nose for a cloud-based setup. Getting to flush all of it down the simplification drain is an amazing step forward. In fact, it's what I like most about paying attention to the progress of our platforms. Oh, browsers now have really good JavaScript and CSS engines? Awesome. Let's go #nobuild. Oh, developer CPUs now have dozens of cores? Sweet. Let's pull CI home. Oh, single-core performance is way up? Wonderful. Let's drop gotcha-hinged accelerators like Spring." https://lnkd.in/gVVA4xAc

New Post: Kubernetes horizontal POD Autoscaling – Hello World - https://lnkd.in/drhBqAZb For a quick test of horizontal autoscaling of an NginX deployment, we need the following ingredients: a metrics server installation to be able to measure the POD’s CPU an nginx deployment with a CPU reservation an auto scaler configuration a process that causes high CPU This can be tested on: https://lnkd.in/e9y53frq Install Metrics (quick&insecure way) curl -s -L https://lnkd.in/dE9QNdvZ

During your busy development day, there are times when Docker Desktop isn't in active use. Whether you're code reviewing, in meetings, or taking a breather, Resource Saver Mode steps in. It cleverly saves resources during these idle moments, cutting down on memory and CPU usage. This keeps your system nimble and responsive, even with Docker Desktop running quietly in the background. With performance intact, you can seamlessly hop between tasks, boosting your inner-loop development process. https://lnkd.in/gQpueA2t

Live / hot checkpoint and migration of #GPU for #hpc workloads was thought to be impossible, at least it was to me. It’s extremely difficult to stub out the gpu memory and registers w/o deep knowledge of the architecture and having access to open frameworks. Still obviously no #cuda / #nvidia support (that I know of, also it would have to be developed deep inside the hallowed halls in Santa Clara in any case), so this to me is #amd taking a legitimate #opensource leadership position right here. Trust me folks, this is a really big dealio. Watch this space. It will impact purchasing decisions, hot migrating running workloads is a giant carbon and cost reduction strategy. GPUs are now in play. Good docs in the GitHub also. https://lnkd.in/eFRziGaa

Virtualization, in the context of frontend development, is a technique used to improve the performance of rendering long lists of data, such as a large set of items in a scrollable container. The idea behind virtualization is to render only the visible items on the screen, rather than rendering all items at once. This can significantly reduce the memory usage and improve rendering performance, especially when dealing with a large number of items. #frontenddevelopment

#FreeRTOS FreeRTOS (pronounced "free-arr-toss") is an open source real-time operating system (RTOS) for embedded systems. FreeRTOS supports many different architectures and compiler toolchains, and is designed to be "small, simple, and easy to use". https://lnkd.in/diPe-t-A