Pod scaling ad infinitum!!
summary
Benchmarking a MEAN‑style geospatial workload (MongoDB geospatial dataset, Node.js 8.14 binaries in CentOS containers) on OpenShift 3.11 with RHEL7.6 and MongoDB Enterprise 4.0.2 shows POWER9 systems using PowerVM and SMT4 sustain far higher container packing and throughput: 3.2x container density per core and roughly 2.5x price‑performance versus an Intel Skylake 6148 baseline. Switching Kubernetes QoS classes (Guaranteed/Burstable/BestEffort) raises transactions/sec; workload driven by an emitter automation tool across identical network paths validates the result.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:48Agenda overview
- 3:57IBM and Red Hat partnership
- 7:49Power 9 server models
- 10:17Proof point study objective
- 13:30Price performance ratio
- 17:42Topology and setup
- 21:59Quality of Service classes
- 24:09Q&A session
related talks
32mPack-em or Stack-em: Advantages of Running OpenShift on Power
20mBenchmarking HPC workloads on OpenShift
44mPerformance and Scalability Tuning Kubernetes for OpenShift and Docker
27mThe Kubernetes Hunger Games: Distro Performance in the Edge
05mIgnite - How Fast Was My Kubernetes?
39mHow we learned to love the Data Center Operating System
56mBenchmark Your Cloud-Native Database
47mCloud Native and Sustainable, Reproducible Scientific Computing