In 2018 after deciding that I hated ProxMox, that Ganeti was dead (and it was at the time), that Harvester didn’t exist yet, that OpenStack was way too complex, and that I was interested in going the Kubernetes/container route (sorry I’m still a VM guy), I decided to write my own self-hosted hyperconverged infrastructure manager. I based it on what little I knew of how Nutanix worked, with a lot of ideas from Ganeti too.

And I named it after drain pipe on a whim at Home Depot.

https://github.com/parallelvirtualcluster

5 years later I have 16 production clusters, including my own homeproduction (but not including my testing cluster), mostly through finding a niche for it with my employer, and I spend a solid 25% of my free time working on it. It’s not quite at a “1.0” release I’d be comfortable with random people using yet, but it’s getting close enough for me to start talking about it on social media!

Owncloud.

I personally never caught the Nextcloud hype, and stuck with the original. So far I’ve heard (and seen, having tried it twoce) nothing but trouble from Nextcloud while my Owncloud install continues to be rock solid for going on 10 years (regularly updated, of course!).

How do you solve the problem of runaway memory usage? Even monitoring a few dozen hosts, memory usage would grow to many GB and continue to grow indefinitely until it OOM’d, and from my reading Influx has no way to prevent this.

I second CMK.

A TICK stack is unwieldy, Grafana takes a lot of setup, and all of this assumes you both know what to monitor and get stats on it.

CMK by contrast is plug and play. Install the server on a VM or host, install thr agent on your other systems, and you’re good to go.