sota

you find a solution? My VMs are now dead in the water.

bump for someone else having this problem.

Glad it was something simple. I had a feeling based on your description of being thorough in checking your unBALANCE logs and actual transactions, that it was something like this.

did you search the actual disk mount points and not the aggregated shares for the missing data?

I can't believe no one has ever done a study on that. Hell, someone write me a script to do the following and I'll buy 2 brand new disks and abuse them to failure in a "scientific" test. disk 1: spin up and constantly write for 5 minutes, spin down for 5 minutes disk 2: constantly write for 5 minutes, idle for 5 minutes or set the time windows for whatever we think would be the most useful.

NEGATIVE. the AR variants don't plug into PCIe slots. They plug into something called "Flexible smart host bus controller" that appeared in the Gen9 servers. you need the NON-AR variant.

also check your power settings in iLO.

... and if anyone else who's in or is seeing this thread, doesn't go and make an immediate backup of their config...

also be prepared, those could be SMR disks, which are bad news if you're needing to write tons of data to them periodically. No idea how they'll respond to a parity build.

now go make a backup of your important stuff.

you're not in NJ are you? and if you're thinking something's amiss with the hardware, strip it all the way down to single processor, single memory chip, no expansion boards, single power supply, no drives plugged in. I had mine delivered with a pair of memory sticks that turned out to be faulty, but because it's ECC memory the system was quietly dealing with the problem. Took a bit to figure it out. Product ID 669257-B21 System ROM P73 05/24/2019 System ROM Date 05/24/2019 Backup System ROM P73 08/02/2014 iLO Firmware Version 2.70 May 07 2019 Smart HBA H240 Firmware Version 7.00

I made a bootable USB stick out of the ISO. there's even an HP branded tool/software to it. I'll grant you it was a bit disappointing when HP stuffed everything behind a paywall basically, but you can find stuff if you look for it. I didn't have as much apparent difficulty you seem to be having, getting the H240 to work properly. Looking at my archives, it's possible I used P19473_001_spp-2019.09.0-SPP2019090.2019-0905.39.iso instead, but I thought I only used that to get the updated BIOS/Firmware for the server, iLO, and potentially the controller. It's been a while since i did this, and the memory wasn't retained.

I used 864794_001_spp-2016.04.0-SPP2016040.2016_0317.20.iso to switch to HBA mode. if you're still having problems getting it done, try and find that. also, see my sig.

HP H240 or other H-series controller works great. The P-series controllers are more for actual RAID work.

given the size of drives available today, if you're needing 30 drives in an array, I think you need to take a hard look at how much crap you have.

I got it and a pair of old Dell servers, from a guy who wanted to retask them as-is for home video storage. We're talking P4 processor in each, 1GB ram, Win2k3 x32 ancient boxes. One had the controller for the MD1000 in it, and the latter has 15 750GB drives installed. I tried to use the 10TB disks when they first came in, but it said NO BUENO. Found out from research that the controller itself in the server can't handle disks > 2TB. I haven't yet bothered to find a controller I can use, as I don't need another 15 drives yet

I have one of these laying around. problem is the PERC6e controller it came with won't support disks > 2TB apparently, and I've been too lazy/disinterested to hunt down an appropriate controller and cable combo to make it work with 10TB+ disks.

also what parity write strategy are you using? RMW or Turbo Write? the former would definitely be ugly.

I think we'd both like to change the time as well.

I was *JUST* about to open a thread about this myself. In my case it's the weekly notification" Unraid Status: 16-03-2020 12:20 AM Notice [CUBE] - array health report [PASS] Array has 8 disks (including parity & cache) Note the time? I have my systems setup so I get emails, and those emails are forwarded to my cell phone via texts. Oh Dark Thirty text ping noises are getting annoying

knight rider, eh? wanna rsync to mine for the time being?

Yea, I'd have choked on that too. You got enough working disks back to recover everything. And didn't I send you a link for a guy with a bunch of used disks for sale?

good chance you're SOL, as that card is old.

you probably spent more in toilet paper from the bricks you must have been... well, you know. Now get a backup copy of that done already. 17 TB is s drop in the bucket these days. out of curiosity, what was the damage to the 5th drive that a PCB swap couldn't fix.

that worked. thanks.