Which is more resilient and more performant: Ceph or ZFS?

There are other threads here that talk about the performance of Ceph vs. ZFS but what I am interested in finding out from the members here is which is better in terms of resiliency and processor efficiency?

The background in terms of why I am asking is I have a 3-node HA Proxmox cluster, that’s also running Ceph 17.2.

It has an Inland 512 GB 2242 M.2 NVMe SSD in each of the nodes and they’re connected to each other over dual GbE.

One of the nodes right now, is showing 100% CPU utilisation and I suspect that it is because the 2nd node completely crashed and auto-restarted itself, so Ceph is trying to heal itself right now.

So, to that end – which is more processor efficient, during the healing/resilver process – Ceph or ZFS? (i.e. would thrash the CPU less during said healing/resilvering process?)

And if I had six identical drives, and the Ceph was using erasure coding (3,1) (m,k) with three of the drives and ZFS was using the other three drives in a raidz configuration, and assuming everything else about the system is the same – which would hit the CPU less during the healing/resilvering process?

As a corollary question to this, in terms of being able to recover from a configuration failure (e.g. config files gets corrupted or something), which would be easier to recover from this type of a failure?

Your help is greatly appreciated.

Thank you.