Some day, I'll bite the bullet and set up ZFS and ceph with Proxmox and make a real cluster... :(
Some day, I'll bite the bullet and set up ZFS and ceph with
Proxmox and make a real cluster... :(
Its really cool - the ONE thing I was missing before I could try all
these different things out was a PBS. If you don't run a Proxmox
Backup Server, its a worthy cause to install on almost any hardware.
What it does is allow you to create backups of any PVE CT/VMs - and
even other file-based backups of HDDs or any NFS shares/whatnot.
|07p|15AULIE|1142|07o
|08.........
--- Mystic BBS v1.12 A49 2024/05/29 (Linux/64)
* Origin: 2o fOr beeRS bbs>>>20ForBeers.com:1337 (1337:3/129)
+1 for PBS. It took me a while to realise that it was so much better than just NAS backups as it does incremental & ZFS de-deplication also - resultin in far less disk usage on my "backup" NAS.
+1 for PBS. It took me a while to realise that it was so much better
than just NAS backups as it does incremental & ZFS de-deplication also - resulting in far less disk usage on my "backup" NAS.
+1 for PBS. It took me a while to realise that it was so much better th just NAS backups as it does incremental & ZFS de-deplication also - res in far less disk usage on my "backup" NAS.
Ok, you sold me. I'm backing up the same download files over and over
and over again...
paulie420 wrote to poindexter FORTRAN <=-
+1 for PBS. It took me a while to realise that it was so much better th just NAS backups as it does incremental & ZFS de-deplication also - res in far less disk usage on my "backup" NAS.
Ok, you sold me. I'm backing up the same download files over and over
and over again...
It's so cool - if you have a backup if:
I think I'm going to run PBS as a VM under Proxmox instead of a
bare-metal install. My only spare system at the moment is a Pi 4, and
it sounds a bit complicated. If I needed to, re-installing Proxmox,
changing the basic settings and installing PBS shouldn't be too hard.
I'm backing up to NFS, not locally to PBS.
so we both have a FINAL 3-2-*1* backup offsite... I just need to find someone else who runs a 100TB+ NAS. :P
paulie420 wrote to poindexter FORTRAN <=-
I use my NFS for file storage; like on all my laptops ~/Downloads, ~/Documents, etc etc - I syncthing to the NAS... then I backup the
entire NAS to my PBS. So I guess ONE of my backups gets sent to the
NAS, but I try to keep a proper 3-2-1 backup system - its getting so
big, however, that I'm gonna have to buy more hardware in order to get
an offsite backup.
One really cool thing I've been looking into - is linking up with
someone else who has the same sized NAS as I do, and doing the remote backups with them. I give them access to some share on my NAS/PBS, and they do the same for me, and we send encrypted backups offsite; sharing the hardware we run so we both have a FINAL 3-2-*1* backup offsite... I just need to find someone else who runs a 100TB+ NAS. :P
so we both have a FINAL 3-2-*1* backup offsite... I just need to find someone else who runs a 100TB+ NAS. :P
How much data is on that 100TB nas backup?
Restoring from it would be a pain though, while I'm 1000MBs down, I'm
only 50MBs up at the moment, hoping to change that to 250down/100up soon...
Thought I'd mention it :)
If you want to rent out space on your NAS, I might have a business
model for you. :) I'm looking into AT&T fiber, and once I'm off of
Comcast with their bandwidth caps, am looking at backing up to the
cloud somewhere as another layer of protection. I've only got around
3TB I'm backing up.
deon wrote to poindexter FORTRAN <=-
Have you heard of tahoe-lafs?
I played with it a few years ago - and its design is perfect for this.
It's effectively raid (or erasure coding) across unreliable targets.
IE: Everybody provides capacity, and data stored is split, encrypted
and erasure coded across everybody's drives (so no single person has a full file - and the data is encrypted anyway).
If a node goes down, your data is still recoverable from other targets
- upto the parity setting. IE: If you set parity to 3, then 3 systems
can go offline and your data is still accessible.
That would be an interesting experiment, if a group of people got
together and offered x amount of disk space and bandwidth into a storage pool.
| Sysop: | Winzlo |
|---|---|
| Location: | Minnesota, USA |
| Users: | 11 |
| Nodes: | 16 (0 / 16) |
| Uptime: | 495936:51:11 |
| Calls: | 82 |
| Files: | 1,070 |
| D/L today: |
27 files (11,920K bytes) |
| Messages: | 286,952 |