KJB

Hi all, So after battling with a Gaming VM for over a year through Unraid I have received a VAC ban from BattlEye. Coupled with ongoing performance and passthrough issues I decided to virtualize Unraid. Using Plopk and creating vhd files to passthrough the raw disks has worked flawlessly and it's great to have baremetal performance back. Even the drives are performing at max speeds. My dockers are still running in Unraid without issue. Couple of questions: - I have nested virtualisation enabled (Ryzen cpu) and can start up and use VM's fine however they won't connected to the network, i've tried different adaptors but just can't get them to bridge. Has anyone managed to get bridged networking to work under a nested VM? - Drive's seem to be spinning down fine but is there anyway at all to pass through drive temps or any other SMART data with Vbox under Windows? Cheers

FS2020 is using DX11 so what you are seeing with your cores is mostly accurate. It took a bit of stuffing around but I'm running it on a VM with 4c/8t of a 3700x isolated and passed through and am seeing exactly the same performance through the VM as baremetal. Averaging 40fps in the air with most settings on Ultra and a 1660Ti. Performance takes a fair performance hit on the ground but this is expected. i440 performs slightly better for me than Q35 however does introduce slight stuttering at times. Using hugepages for RAM also seems to help and the more the better (more than 16gb). This helped in my case. https://www.reddit.com/r/MicrosoftFlightSim/comments/ig1mwz/running_msfs2020_under_qemu_utilizes_cpu_poorly/ Happy to share my XML if you want to have a look but there's nothing too special about it.

Hi all, I'm having issues with my network speeds with over my LAN. I'm currently on the latest beta however this issue was present on all previous builds including stable. Unraid seems to be capping any network transfers at 100mbps. I am unable to read or write to my cache drive over LAN any faster that ~10MB/s. iPerf3 results are reflecting this: Network is showing correct at eth0: 1000 Mbps, full duplex, mtu 1500. I have tried the following: -Different CAT6 cables -Different switch port -Connecting directly to router -A second intel NIC card -Disabling NIC flow control and offload in tips and tweaks plugin My WAN speed has a 100mbps limit which Unraid has no issue maximising so I can't really test if this is affected. Any other advice or suggestions? I've living with this for awhile now but transferring large files locally is starting to become very tedious. Cheers unraid-diagnostics-20200723-1157.zip

I'm not really understanding the new Cache Pool feature. User Shares set to Cache Only are now showing as All Files Protected in the Shares tab. I only have the 1 cache drive. Are cache shares now part of parity? When you go down to Disk Shares it is still showing the whole drive is unprotected.

GPU passthrough is broken with current BIOS versions on Ryzen.

Still not working on an Asus Prime board with a 3000 or a 2000 series chip.

I would hold off buying a 3000 series for now. I'm starting to regret buying one. I've had to turn PBO off to get stable temps and stop the chip from over-volting. The whole launch is starting to look like a bit of a mess. https://www.reddit.com/r/Amd/comments/cbls9g/the_final_word_on_idle_voltages_for_3rd_gen_ryzen/

@mattz Just to add I did try the initial BIOS revision that was released for this board. I've also popped a 3700x onto the board which I purchased for my gaming PC and still cannot passthrough either an AMD or NVIDIA gpu. Tried different slots, ACS settings, BIOS settings, the works. This is definitely a board issue and not cpu related. https://www.youtube.com/watch?v=l-WHCCwgyDg

Board is an Asus PRIME X570-PRO. I have tried all released BIOS revisions for the board. This board does let me downgrade. Currently on Version 7010 which is latest and a BETA apparently.

Any suggestions for x570 boards yet? I just upgraded from a b350 due to a broken SATA port and still running a 2600 however I have lost the ability to passthrough any GPU to Win10. Obviously I can't downgrade the bios. VFIO support looks poor at this stage.

I'm running an Asus Prime x570 with a 2600 and GPU passthrough no longer works. Was fine on the B350 it replaced. Obviously I can't roll back my bios so it's a waiting game name.

I replaced the cable and swapped to a different SATA port and has been fine for the past 24 hours. I'm putting it down to a dud cable. Or port. Thanks for the replies.

The only reason i'd suggest the drive is fine is because it was purchased 6 weeks ago and pre-cleared fine before use. I know drives can fail at any time but is it likely within 4 weeks of a pre-clear completing? Parity was done after the new array was set and came up with 0 errors.

Hi all, I had an issue the other day where one of my drives became disabled overnight. Ran a full SMART check and the drive showed no issues, there were no writes being performed at the time so I put it down to a hardware error, changed the SATA cable and re-enabled the drive with the Trust My Array procedure. Runnings through the logs of the drive at the moment I keep seeing these errors. Jun 15 02:45:56 SERVER kernel: ata10: SATA link up 3.0 Gbps (SStatus 123 SControl 320) Jun 15 02:45:56 SERVER kernel: ata10.00: configured for UDMA/133 Jun 15 02:45:56 SERVER kernel: ata10: EH complete Jun 15 02:45:56 SERVER kernel: ata10.00: sense data available but port frozen Jun 15 02:45:56 SERVER kernel: ata10: limiting SATA link speed to 1.5 Gbps Jun 15 02:45:56 SERVER kernel: ata10.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen Jun 15 02:45:56 SERVER kernel: ata10.00: irq_stat 0x48000001, interface fatal error Jun 15 02:45:56 SERVER kernel: ata10: SError: { UnrecovData 10B8B BadCRC } Jun 15 02:45:56 SERVER kernel: ata10.00: failed command: READ DMA EXT Jun 15 02:45:56 SERVER kernel: ata10.00: cmd 25/00:00:78:df:fa/00:01:2a:03:00/e0 tag 20 dma 131072 in Jun 15 02:45:56 SERVER kernel: ata10.00: status: { DRDY SENSE ERR } Jun 15 02:45:56 SERVER kernel: ata10.00: error: { ICRC ABRT } Jun 15 02:45:56 SERVER kernel: ata10: hard resetting link Jun 15 02:45:57 SERVER kernel: ata10: SATA link up 1.5 Gbps (SStatus 113 SControl 310) Jun 15 02:45:57 SERVER kernel: ata10.00: configured for UDMA/133 Jun 15 02:45:57 SERVER kernel: ata10: EH complete Any ideas what is causing this? I'm confident the drive is fine. Should I be looking at changing the SATA port on my MB? Another dud cable perhaps? Looking through the SMART data I have a tonne of UDMA CRC error's. Cheers