Hello! Hoping to get some insight here from those that may use something similar.
My current setup is:
1TB SSD drive in Raid1 (via zfs when ProxMox was installed).
1 TB SSD drive in Raid1 (via zfs when ProxMox was installed).
2 TB HDD Drive
2 TB HDD Drive
When I setup ProxMox initially, I chose only the first two disks, and chose ZFS for Raid1. On the other two disks, I selected "-- do not use --".
This appears to be working for what I need, as I can build my VMs and containers and they will store properly on the SSD drives where I'd like them.
However, I would also like to initialize my second Raid1 array using my other 2 2TB drives. This is going to be for storage only, and maybe for monthly backups of the VMs / Containers (this isn't going to be immediate, but figured I'd mention it is eventually the plan, so I want to make sure I'm forward thinking the setup here).
I will have an application server for my Ruby web applications in a VM on that first Array.
However, I'd like to use the second Raid1 array (which I haven't built yet) for storage for usage by some of those VMs.
For instance, one of my Ruby applications hosted on the first Raid array will be using a self-hosted S3 bucket for storage of items the application would reference (perhaps images, videos, or other kinds of files). Now, this S3 bucket will be running in a VM or containerized on that same first Raid1 array right along-side (though in a separate VM/container) the application server. However, I'd like to tell that S3 bucket to actually use the storage of the SECOND Raid1 array for the actual directories and files that the Bucket references. So the S3 bucket only needs a small amount of actual storage for the service itself / configuration / logs on Raid1(a), but I'd like the files themselves on Raid1(b). I'll configure that once the S3 bucket container is up and running. Does this make sense?
Raid1 Array(a) (2 1TB SSD)
VM (Debian Workstation)
VM (Ruby Application Server)
VM (S3 Bucket Server)
VM (DNS Server)
Raid1 Array(b) (2 2TB HDD)
Storage Pool (S3 Bucket Server) <- Can this be put on a slider for scaling use?
Storage Pool (un-used, but assignable as needed)
I apologize if this is super obvious or standard, however I'm only a year or two into the Linux world in general, with about ~20 years in Windows environments, so I still have to lookup all the commands to even navigate Linux here and there. Basically, I've only done initial setup, and have not added any VMs that I plan to as above. When I installed ProxMox, I only initialized those first 2 drives into Raid1, and they seem to be working properly, but I am not sure how I can manually setup another Raid1 with the unmounted / unused storage drives. I am ok with setting this up in the ProxMox console via web interface, or ssh'ing into the server to do it on the terminal. Again, I'm still learning the entire Linux world, so if there's a way to do manually exactly what ProxMox did when I chose those first two drives, I have no idea how to do it, and if it's different than using, say, mdadm on a Debian server.
I am okay with installing the Bucket on the other Raid1 array entirely, if necessary, and I am also ok with re-installing ProxMox completely as I haven't configured anything past a VLAN.
Thanks!
My current setup is:
1TB SSD drive in Raid1 (via zfs when ProxMox was installed).
1 TB SSD drive in Raid1 (via zfs when ProxMox was installed).
2 TB HDD Drive
2 TB HDD Drive
When I setup ProxMox initially, I chose only the first two disks, and chose ZFS for Raid1. On the other two disks, I selected "-- do not use --".
This appears to be working for what I need, as I can build my VMs and containers and they will store properly on the SSD drives where I'd like them.
However, I would also like to initialize my second Raid1 array using my other 2 2TB drives. This is going to be for storage only, and maybe for monthly backups of the VMs / Containers (this isn't going to be immediate, but figured I'd mention it is eventually the plan, so I want to make sure I'm forward thinking the setup here).
I will have an application server for my Ruby web applications in a VM on that first Array.
However, I'd like to use the second Raid1 array (which I haven't built yet) for storage for usage by some of those VMs.
For instance, one of my Ruby applications hosted on the first Raid array will be using a self-hosted S3 bucket for storage of items the application would reference (perhaps images, videos, or other kinds of files). Now, this S3 bucket will be running in a VM or containerized on that same first Raid1 array right along-side (though in a separate VM/container) the application server. However, I'd like to tell that S3 bucket to actually use the storage of the SECOND Raid1 array for the actual directories and files that the Bucket references. So the S3 bucket only needs a small amount of actual storage for the service itself / configuration / logs on Raid1(a), but I'd like the files themselves on Raid1(b). I'll configure that once the S3 bucket container is up and running. Does this make sense?
Raid1 Array(a) (2 1TB SSD)
VM (Debian Workstation)
VM (Ruby Application Server)
VM (S3 Bucket Server)
VM (DNS Server)
Raid1 Array(b) (2 2TB HDD)
Storage Pool (S3 Bucket Server) <- Can this be put on a slider for scaling use?
Storage Pool (un-used, but assignable as needed)
I apologize if this is super obvious or standard, however I'm only a year or two into the Linux world in general, with about ~20 years in Windows environments, so I still have to lookup all the commands to even navigate Linux here and there. Basically, I've only done initial setup, and have not added any VMs that I plan to as above. When I installed ProxMox, I only initialized those first 2 drives into Raid1, and they seem to be working properly, but I am not sure how I can manually setup another Raid1 with the unmounted / unused storage drives. I am ok with setting this up in the ProxMox console via web interface, or ssh'ing into the server to do it on the terminal. Again, I'm still learning the entire Linux world, so if there's a way to do manually exactly what ProxMox did when I chose those first two drives, I have no idea how to do it, and if it's different than using, say, mdadm on a Debian server.
I am okay with installing the Bucket on the other Raid1 array entirely, if necessary, and I am also ok with re-installing ProxMox completely as I haven't configured anything past a VLAN.
Thanks!
Last edited:
