AI
Why Your GPU Infrastructure Costs 40% More Than It Should
Most AI infrastructure teams spend 35-60% more on GPU compute than they need to. The cause isn’t cloud pricing. It is architecture, and it is fixable.
apache spark
Spark on Kubernetes Reserves CPU It Never Uses. Here’s the Overcommit Fix.
Spark sets executor CPU requests equal to limits, so a Kubernetes cluster reserves twice the CPU it uses and refuses to schedule pending pods. Kubernetes has no native overcommit. Here is the mutating-webhook operator I use to fix it.
apache spark
Apache Spark smoke tests on kubernetes
Introduction Benchmarking remains a critical (and often underestimated) tool when designing or validating large-scale data platforms. While many teams rely on synthetic workloads or production replays, standardized benchmarks still play a key role when comparing architectures, tuning clusters, or validating infrastructure choices. TPCx-HS is one of those benchmarks: designed to Read more
data
Who’s Using Ceph RBD Mirroring for Kubernetes Storage in Production?
Alpha Feature in Production… Good Idea? The most promising approach, journal-based mirroring, offers near real-time replication and faster failover. However, it’s currently an alpha feature in the Ceph CSI driver and relies on rbd-nbd, which introduces significant risks: For documentation on Ceph RBD mirroring see IBM documentation For details, see Read more
ceph
Choosing the Right Storage Backend for Kubernetes PVCs
Most discussions about PVC focus on on-prem deployments, but many of these technologies are equally relevant in the cloud, especially when using managed block storage like AWS EBS, which caps at around 64K IOPS. By contrast, OpenEBS LocalPV (ZFS or LVM) can attach directly to NVMe-backed local disks, unlocking true Read more