Darragh Grealish

Erlang RED https://lnkd.in/e8ANSp4H Gerrit Riessen created #Erlang RED to experiment with visual Flow Based Programming (FBP) in conjunction with Erlang. FBPs message-passed concurrency is well suited to Erlang, conversely Erlangs Supervisor design pattern is well suited to FBP. The goal is to bring the advantages of low-code visual flow-based programming to a programming language that is designed for message passing and concurrency from the ground up, Erlang. This ended up a deep conversation on design templates / philosophies, and a planned follow-up discussion focused on telecom applications using Erlang Red. Sam Machin, Viacheslav Katsuba, Scalicon Inc., Flowhub, Jonathan Eisenzopf, Ericsson, Vance Shipley, Nexmo (now Vonage), Vonage

25

6 Kommentare

🚀 Why Expanding an EBS Volume on a Running EC2 Requires Manual Steps 🤔 Ever wondered why expanding your AWS EBS volume while your EC2 instance is running still requires you to run xfs_growfs or resize2fs — but not when it’s stopped? Here’s the key: 🔹 When you stop the EC2 and expand the volume, the Linux OS detects the new disk size at boot. Tools like cloud-init or systemd may even automatically resize the partition and filesystem. 🔹 But when the EC2 is running, the Linux kernel doesn’t automatically know the block device size has changed — it needs a manual nudge. ✅ Steps on a running EC2: ##bash sudo growpart /dev/nvme0n1 1 sudo xfs_growfs / 💡 Why this matters: If you're running Docker workloads, databases, or any I/O-heavy applications — understanding disk growth behavior is critical to avoid outages or misconfigured storage. 📌 Bookmark this tip for your next on-the-fly volume expansion! #AWS #EC2 #EBS #DevOps #Linux #CloudTips #SRE #xfs_growfs #resize2fs #DevOpsCommunity

33

Why can’t Big Tech just build their own chips? A common question: “If Microsoft, Google, Amazon, or Apple spend billions on R&D, why can’t they end NVIDIA, TSMC, ASML, or AMD’s dominance?” Here’s the reality 👇 -> #Microsoft (Maia chip) Maia is designed to work seamlessly with Azure. It may not yet rival NVIDIA’s top GPUs, but it helps Microsoft cut costs and optimize performance inside their own cloud. -> #Google (TPUs – Tensor Processing Units) Google’s TPUs power many of its own AI services and are tightly integrated into Google Cloud. They are highly efficient for specific workloads, showing how custom silicon can drive scale. -> #Amazon (Trainium & Inferentia) AWS uses these chips to make training and inference cheaper for customers, giving enterprises more choice beyond standard GPUs. -> #Apple (M-series) A success story in consumer devices, though datacenter AI compute is a very different challenge. * Why #NVIDIA, #TSMC, #ASML still lead - NVIDIA’s pace of execution - TSMC’s unmatched fabs - ASML’s monopoly on EUV machines It seems, Big Techs are not trying to kill NVIDIA. They are rather building in-house chips to lower costs and fine-tune performance for their own platforms. For raw #AI compute, NVIDIA and AMD still set the pace.

10