Q&A: pics on 4.19 minimum requirements

[asmoore82]

I was in the spot of wanting to install my first test cluster on limited hardware. Being a sysadmin and not a dev I didn't want to rollout any of the lighter or automatic options. I wanted to focus on making a true cluster of 3, explore the console, and practice non-disruptive maintenance.

With the limited hardware, this basically became a plan to ignore the documented minimums "for science!" (^_^)d

The first cluster that achieved a working console had 3 VM's each with 12GB of RAM and 20GB disk. Can confirm that's not enough disk:

32GB of physical memory on the box was over-allocated to those 3 VM's, so the next run brought RAM down to 10GB each with 40GB disks. Can confirm that's not enough RAM:

The console was stable though:

And I got to check pending pods with a clear explanation:

It was working well enough that the plan shifted to transplant 1 VM to different hardware, give it 16GB RAM, and then increase RAM on the other VMs one at a time to mimic non-disruptive maintenance - cordoning and draining along the way.

continued...

[asmoore82]

Notifications spiked while the transplant VM was offline:

Here is that VM back with 16GB RAM and the top one preparing to go offline for its increase:

And here is the process completed on all 3, at this point notifications all cleared up (^_^)d