I will take all of that into consideration and thanks for the reply. Right now my customers sit comfortable on a Contabo server located in the US. The reason they chose me to host is simple though. They're my friends lol. But they (only 3 of them) get a good deal from me and they can actually come talk to me face-to-face about any issues. I would say more of a technical liaison for their small-business. They throw me a few $100 a month and I build their website, host it, update it, and maintain it. That being said, I will most definitely be looking into all of the prosumer options from hetzner (more than probably) in the near future, but, I still would like to know a good partitioning scheme for the moment using the hardware I have. partition /dev/sda --> lvm --> vg0 ? should I use the 2x3TB hdd's for iso images and storage of snapshots? Or should i install proxmox on one and use the ssd for VM's or for swap? Thanks in advance guys/gals!
Like I already said, I would alteast use raid1 everywhere. If your system disk fails every configuration is lost, every VM is offline and you need to freshly install your OS. It takes hours or days to setup everything again with all the firewall rules, setup of the mail server, setup of monitoring tools, .... A buisness website that is down for hours or days is bad for the reputation and for the buisness loosing customers.
Same for the VM storage. It isn't that bad if you create daily backups but you still loose all data created that day and it takes some time so restore all VMs from backups. I don't know if you are running some webshops or something like that but loosing a day of data and so all webshop orders, is really bad. People paid stuff and you will never deliver it because you lost the orders.
Use raid1 for redundancy and any disk may fail without anyone being able to notice it (but you hopefully notice it because you use some kind of monitoring tool like zabbix, grafana or whatever).
And if you aren't willing to use raid1 I would think you are also not prepared and already have bought a spare disk that is laying around so you could directly replace it if it fails? In that case you need to buy one first so the replacement will take even longer...
And if you already got a spare one, put it in and use it as a raid1. Sooner or later a drive will fail and that way you atleast don't loose any data.
And even if none of your drives may totally fail your data still might silently degrade over time. Google for "bit rot" to understand what I mean. Write something to a SSD or HDD, wait 1 year and read it again. The data won't be 100% the same because bits will flip over time. Against that it is usefull to use a advanced filesystem like ZFS that is checksumming all data and will compare all data to the checksums on a monthly schedule. That way it can find rotten bits. If you now also got some parity data (like using a raid1) it can automatically repair those rotten bits and repair your data. If you don't got a raid1 it will tell you what files are corrupted but it can't do anything to repair them.
So again a good point why to use drive redundancy. And backups won't help you against bit rot, if you for example are using LVM, because you don't know if a file is corrupted or not when you are backing it up. So you might think everything is fine but you overwrite your old backups with new backups and your healthy file is replaced with a corrupted one.
And you can't store snapshots on another disk. Its always part of the same disk that you want to snapshot. So the HDD would only be usefull for backups/isos. And snapshots aren't really helping because if you loose a disk without raid1 you also loose all your snapshots so you only got your backups and these are not the newest because backups can take hours if you don't use PBS with SSDs so you aren'T creating them every hour or every several minutes.
