Actually getting somewhere!

Quick update after my previous update on this while server situation!

THE SERVER IS STABLE!!!!

for now..

OMFG finally.. xcp-ng has made my life soooo much easier, xen orchestra makes managing my VMs a breaze..

I had a few hickups with the host system dying when trying to build a windows VM with gpu passthrough, Managed to work my way around it by building the VM without passthough first then adding it in later. When a lot smoother then.

Hive Witness node is currently replaying, it will probably be a few days because it's I/O intensive and I'm on a single 14TB HDD with 3 VMs on it, Disk performance tops around 66MB/s at the moment (according to the stats) which will make the replay take a lot longer than my previous NVMEs

Blurt is also syncing up as a full RPC node.

and the windows VM is running in the background just doing it's thing while I wait for a fake monitor dongle so it actually uses the GPU correctly.

I've got to trim down the RAM allocated to Hive when it's done replaying as 128GB is overkill. 32GB is enough it seems, Same with it's shared_memory file 128GB is overkill I'll stick it down to 40GB ish later.

I've got a few more VMs to build, Mainly a Node.js VM, for price feeds and anything else node.js. Probably run the alloyxuast test site there while I'm developing it.

I need a proxy VM the uses nginx at my like entry point so I can direct the correct traffic to the correct VM. Would save me having to port forward a lot which requires the network to reboot after every change oof.

So yeah.. Some good news finally. Now! I don't know if it's fully stable as it's only been online 20 hours.. The real test will be when the GPU passthrough really starts getting used.. As the IOMMU stuff might be iffy and cause host crashes.. or it might just work.. time will tell.