This is a companion discussion topic for the original entry at https://level1techs.com/video/building-ultimate-devops-workstation-part-1-nvme-raid
Thats what we are talking about
I am glad you decided to pursue this series! I remember the last video you did on DevOps-ey stuff and it was 10/10.
Really excited to see the rest of the series. I’m curious to know if you’re going to dive into the workflows you spoke about earlier: Vagrant/Virtualbox, Ansible, Docker, Kubernetes, unit testing, integration testing, etc.?
I am considering a very similar development machine (sans RAID 0 NVME). I use all of the tools you mentioned plus a few more and minus one or two Wanting to see performance first hand before I spend the cash. Of course, all things considered, I’ll probably make it rain regardless
When you offload the testing/building to containers, are you hosting the container tech in a virtual machine or are you running it locally?
The series is literally just performance
What do you think I should show next?
The last vid in the series is booting from zfs and being able to roll back the whole fs
Mainly I just like the speed of docker-compose and ansible. I haven’t found a good way to demo a kubernetes workflow that spins up containers for doing integration tests though.
The idea of an Ansible and Docker-Compose speed run sounds awesome lol. I am looking forward to your ZFS boot, snapshot, and roll back episode as well.
Maybe load test the pods and see how the auto-scaling and health checks perform Also provisioning more worker nodes with a script or configuration management tool would be awesome.
Incrementing work loads on the existing compute nodes would probably be the best way! Stress/load testing and then maybe running all of your unit and integration tests.
Maybe write a Bash or Terraform script to perform a full migration from your existing development box to the new one? That is something I plan on doing when I build the new box.
Explaining how and when PCI-E 4 nvme is useful and helpful. Seems like a step in the right direction to me. A little tired of tech tubers stating the that pci-e 4 is not that useful.
I have a particular workload in mind that could fairly easily be run as a benchmark, outlined in the quoted post.