MarkP

I believe you're right. I can only think that it is a windows issue. I'll continue to look in that direction.

That wouldn't explain why a newly created VM is speedy. Unless there is an incompatibility with VM's that were created on an older version of unraid. But yes. I did notice it shortly after an unraid upgrade.

Unapologetic bump. These threads report the same issue recently.

Windows 10 VM that was performing flawlessly suddenly is experiencing slow network speeds. Speedtest from the affected VM shows download speeds of 1.5Mbps and upload speeds of 36Mbps. testing from another machine is 300Mbps testing from unraid speedtest CLI utility is 300Mbps testing from newly created Win10 VM is 300Mbps Created new Win10VM and attached vdisk.img from affected VM and speed is 1.5Mbps. Updated virtio drivers to latest drivers. virtio-win-0.1.173.iso

Copied 2TB without issue last night. I think I can safely call the issue resolved. Solution Summary: Marvell Ports on the motherboard were being used. Incorrectly assumed that the 4 marvell ports were the ones that were physically separated into a separate group. On the ASRock EP2C602 Motherboard - the marvell ports are the four closest to the RAM modules. Don't use them.

3 more hours on my parity sync/data rebuild. I think I'd rather do it with parity enabled. Need to feel confident that everything is working correctly. I disabled the marvell controller in bios. Will let you guys know how it goes. THanks for the help.

Well crap. I thought the four ports that were in the separate grouping were the Marvell ports. I stand corrected. I'll rearrange those and try again.

This is my first build. ASRock EP2C602 Motherboard. Dual Xeon E2660V2 CPUs Using the onboard INTEL Sata ports (not the marvell). 64GB ECC Registered RAM One 8TB WD RED for Parity One 3TB WD RED for Data Trying to transfer data from a 3 TB WD RED formatted as HFSPLUS and mounted as unassigned device (using the unassigned device plugin). Using Midnight commander from SSH prompt to perform file copy. SYSLOG is attached. The parity and data drives are new and have been extensively tested and precleared twice. The unassigned device was pulled from my Mac JBOD tower and has been in used for about a year. I have eight of these and was planning to copy them to the array and then add the disk to the array (one by one). This was my second attempt. I was unable to capture the syslog the first time - but it exhibited similar symptoms. I'm getting pretty discouraged. Rebuilding parity after each crash takes so damn long. I'm starting to feel like I made a huge mistake. I really want to like unraid - but I'm not feeling the love right now. syslog.txt