Michael_P

Dunno, I don't use 'em - should be the same tho, i'd think

FWIW, you're not alone - I see the same thing. If I restart the SABNZB docker I can hit max throughput ~60-70MB/s but soon drops to 30MB/s. When I was running SAB under a VM it maxed out my connection no problem for the entirety of the DL. I haven't figured it out, either. There does seem to be reports of others having the same issue, tho.

Dead giveaway is reallocated sectors and un-correctable sectors 5 Reallocated_Sector_Ct PO--CK 001 001 010 NOW 64880 (0 5) 197 Current_Pending_Sector -O--C- 099 099 000 - 656 198 Offline_Uncorrectable ----C- 099 099 000 - 656 Anything more than zero is cause for concern and for me is a candidate for immediate replacement - doubly so in the case of Seagate drives, which for me at least, have never failed gracefully allowing data to be retrieved (always lose a few files here and there, if not the whole drive's contents).

Fair enough, plenty of horsepower there then. I run mine on a Atom C2550 and it barely breaks a sweat.

If you're already in a Unifi ecosystem, why bother with a pfsense box?

sdn and sdi don't look happy and should be replaced, sdl looks about the same vintage so is likely to follow suit. I wouldn't trust any of those as Seagates have always failed with data loss for me (saved by backups of course)

1- yes, but you need at least 1 disk for the array 2- yes, it's exactly how I moved from WHS 2011 (which I still have running now in a VM on my Unraid server, still great for backing up Windows clients)

OK here's what I did to fix it: Left blank the Storage Path in more settings Added an import folder share Added a path variable Name: Temp Container Path: /tmp Host Path: /mnt/user/appdata/photoprism/temp Not filling up docker image file anymore

Yeah, the photos are going to the correct share but whatever it's doing is filling up the docker image. I've tried setting a /tmp path variable which helps during the actual upload, but when it indexes docker gets maxed What should I be setting as the container and host paths?

How do I get PhotoPrism to stop filling up my docker image when uploading pictures?

FWIW- the drive is still covered under warranty in the US even if you've taken it out of the enclosure, tho the warranty will no longer cover the enclosure if it fails... Drives looks fine, tho - have enough power going to them?

Start with the OS and go from there

That's pretty directly lol - leaving 22 80 and 443 is terribad - that's like saying you locked up your house but went ahead and left the front door, garage door, and back door wide open. Wipe clean and start fresh, once it's compromised it can't be trusted

Your Unraid box was exposed to the internet?

Your Unraid server shouldn't be exposed to the internet, anyway, so why bother?

Don't use those. You'd need them too read 100% correctly to rebuild the array, should you ever need to. There's a reason why it's a Pre-Fail notice lol, it's failing.

FWIW- in the US they still have to honor the warranty (and they do)

I use the PR1500LCDRT2U now for the server/network I have always had a surge suppressor in the main panel, and another in the sub-panel. Fat lot of good they do lol. I suspect the first surge was a near lighting strike that induced a charge on my security camera's video line, back to the security DVR (which fried) and sent a surge down the connected HDMI to my AVR which then passed it along to my projector. In all, about a $5,000 hit - tho both the projector and AVR were replaced/repaired under warranty. No idea where the second one came from, it only made my HDMI ports go screwy this time. When I replaced my security system with POE cameras, I connected the switch via fiber to the rest of the network, so no chance of it passing it beyond the cameras.

Anecdotally, I stay away from APC after a couple of bad experiences - one of the BR1500Gs I had protecting my server failed spectacularly, one of the power transistors exploded (literally) filling the room with smoke as it continued to try and operate. Luckily, someone was home to yank the plug. The replacement (brand new) BR1500G failed in less than a year. I do still have a few NS1080's protecting desktops and AVR, tho. No issues with those so far, except surges have taken out the AVR, twice. Haven't been able to specifically blame the UPS for not stopping the surge(s), tho, as it could have come in from a different path. I put a LE1200 behind the UPS and the AVR still got popped a second time. YMMV

My Cyberpower PR1500LCDRT2U works fine with apcupsd

It wouldn't be twice the cost, it would be more like 5-10 times. Logistics is a key factor driving costs for all products - lithium batteries in that kind of density could never be shipped air, and limited in weight when shipping ground. Special circuitry would need to be incorporated into the design and a bunch of testing performed before it's even allowed to ship ground. It's just impractical for consumer goods. And that's before you factor in the cost of the battery itself.

Cost, and they're a pain in the rear to ship

If the UPS uses MOVs for suppression, those do wear out

Yeah, they have gotten expensive, I paid $85 new back in 2018! In a nutshell, VT-d allows for directed I/O, which allows hardware to be assigned to VMs The Mellanox cards work fine, but you'll need at least an x4 PCIe slot free Depending on the number of drives, you may need to keep the controller on an x8 lane or your parity checks will suffer - and I'd stay way from that marvel controller on the board.

The CPU will be the limiting factor, here: No VT-d means no pass-through (this is the downside of the early K processors) PCIe is limited to 1 x16, or 2x8, or 1x8 and 2x4 If you source a different non-K variant, your setup should be fine - get an Intel expander, no slot needed as it can be powered by a molex connector - res2sv240