"This listing was ended by the seller on Tue, Jan 27 at 5:38 AM because there was an error in the listing."
Enterprise SSD "small deals"
- Thread starter ca3y6
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
To be honest I'm also starting to think whether to downgrade future systems to a single SATA / NVMe Drive (with Backup Server turned on 24/7 and backing up every few Minutes) instead of a ZFS Mirror Setup 
.
Knock on Wood, I don't recall having had ZFS Failures, aside from some Crucial M500 / MX100 / maybe MX200 SSDs (those were "early Days"). Never managed to get those fixed with their weird Power Cycling / Power Reset Procedure in BIOS (basically only plug in Power Connector without connecting the SATA Data Connector). But apart from that, knock on Wood once more, nothing.
.Knock on Wood, I don't recall having had ZFS Failures, aside from some Crucial M500 / MX100 / maybe MX200 SSDs (those were "early Days"). Never managed to get those fixed with their weird Power Cycling / Power Reset Procedure in BIOS (basically only plug in Power Connector without connecting the SATA Data Connector). But apart from that, knock on Wood once more, nothing.
Uh ? Single VDEV has the same available Disk Space as a Mirror ...
I'm not claiming it does.
I agree I should have a Backup.
But until I keep having one Issue after the other, Salt/Saltstack keeps having Bugs (last workaround I had to do was Debian Permissions after each Update), I don't have a proper way to setup automated Secrets Management (latest Trend seems to be git with encrypted secrets using e.g. Mozilla sops + Git + AGE Encryption), and and and ... well, there is just so much you can do in 24h
.Originally the Backup Server was supposed to be turned on on demand e.g. once every Week, to keep Power Consumption low.
But at the current SSD and RAM Prices, the cost Calculation seems to be very different.
Having dealt with a ransomware recently, a backup computer on the same network and with the same credentials than the primary server is as good as no backup. These things propagate from machine to machine if they are online.
Oh yeah I'd be in a big **** if that happened. Same credential everywhere both for root and normal User . Once again tied to that Issue with Salt and lack of being able to deploy Credentials.
I could in Theory change everywhere except my Workstations, and store the passwords in Bitwarden, but then it's also how to unlock in that event where I got no SSH / Clevis access at boot. Entering 64 Characters via IPMI Web Interface isn't fun.
. Once again tied to that Issue with Salt and lack of being able to deploy Credentials.I could in Theory change everywhere except my Workstations, and store the passwords in Bitwarden, but then it's also how to unlock in that event where I got no SSH / Clevis access at boot. Entering 64 Characters via IPMI Web Interface isn't fun
.Not if they do
zfs destroy -r ... if they can gain access.I heard
restic in append-only mode was also good against ransomware (since nothing older can be deleted), but all bets are off if the Account running the Program (let alone root) gets compromised.what you can do at least is have a firewall rules to prevent server to server connections of dangerous ports, like RDP on windows or SSH on linux. Likewise, no reason for a server to initiate a connection to your desktop, so firewall rules on your desktop.
I only have Firewall (kinda) set up on 1 Proxmox VE Host (takes Time to make sure that all Whitelisting and Management rules are in place in order to avoid getting locked out).
Nothing on Desktop.
I don't use Windows anymore so that's one less Thing to worry about.
But Linux and SSH with Password Authentication is also not good (especially since Password is always the same).
For backups you might want to consider Stablebit DrivePool. it has some nice features. I just discovered one recently when I accidentally deleted the partition form a 14TB HD when I had it out for testing in another box. After an O shit moment that seemed to last forever as I contemplated on how to best fix this major goof I put the drive back in it's original server and to my complete surprise and joy, DrivePool began to rebuilt the pool automatically after booting up without me doing anything but stare in disbelief. After the rebuild everything was back to the way it was before. And also, you can remove a single drive from a pool and it automatically renders the entire pool read only. To recover you simply plug the drive back in.
EDIT - Forgot too mention the pool was X2, (Duplicated) had it been X1 I would have been SOL. And FWIW, if a drive in a pool fails you will only lose whats on the one drive, the rest of the pool is safe, but as individual drives and not a pool.
EDIT - Forgot too mention the pool was X2, (Duplicated) had it been X1 I would have been SOL. And FWIW, if a drive in a pool fails you will only lose whats on the one drive, the rest of the pool is safe, but as individual drives and not a pool.
It doesn't.
I'm still looking for a decent Linux Backup program but haven't found one yet. . This is why my 2 backup servers are still Windows. I'm not using it yet, but I heardIt doesn't.
restic was a good Option.Deduplication, Encryption, and you can run it with a local Disc as a Target, to a HTTPS
restic Server and I believe over SSH if you want that road.