I’m running the Docker image on the smallest Hetzner VMs, with 5 concurrent groups and 40 shared rsync threads per container, and 12 containers per server. Star

I’m running the Docker image on the smallest Hetzner VMs, with 5 concurrent groups and 40 shared rsync threads per container, and 12 containers per server. Start one container, do docker top on it to make sure it’s pulling, then start the others one by one, taking a few seconds between each to avoid overwhelming the CPU. I’ve got 6 of those little VMs going, and have rolled up 4GB and 2800 groups worth in 6 hours. After they settle down, they’re more memory than processor intensive. I’ve considered playing with the settings a bit, but thought it was more important to get a bunch of them running on a couple different VMs at different sites. If I were really feeling fancy, I’d write a nice deployment definition for orchestrating this with microk8s...