richsoft

Members
  • Posts

    5
  • Joined

  • Last visited

Everything posted by richsoft

  1. I can also confirm that this is working in 6.8-rc1 for me, without any config edits (I have removed them).
  2. This is excellent! I was moving towards this solution too, but I ended up downgrading to 6.6.7 because of the array speed issue and never tried it out (HW decoding works in 6.6.7, so that was a bonus!). Like you, I would not need to pass through any PCIe devices, as the J4105 board I am using only has one slot, and that is in use for a SATA board. I was wondering if you tried 'intel_iommu=igfx_off'? This is another setting that i found when searching for a solution. It might allow other devices to be passed through to VMs, but removing the graphics from IOMMU might allow HW decoding to work. If/when I can upgrade again, I will be able to do some more testing on this issue.
  3. I am also back on 6.6.7, and the performance is back to what is was prior to 6.7.x. I have never changed the NCQ setting, and so it is set to 'Auto'. cat /sys/block/sdX/device/queue_depth = 1 for all spinning disks, but is 31 for the (single) SSD cache drive. I don't know if this was the case before upgrading to 6.7.x, or if this was set during the upgrade and the new config survived the downgrade. However, as it is set to 1 at the moment, and the performance is fine (with 6.6.7), then I am not sure this is the underlying problem. Tunable (md_write_method) is set to 'Auto', but i do have the 'CA Auto Turbo Write Mode' plugin installed that turns on Turbo Write when all disks are spinning (and greatly increases my write speed to ~110Mb/s when transferring over SMB). I did not think about turning this off when having issues in 6.7.2, to see if this was a problem or not. I don't really want to re-upgrade to test things at the moment, until there is some acknowledgement/movement on this issue, as 6.6.7 is working for me.
  4. I'll also add my experience of this into the mix, although I haven't done any measurements to quantify my findings. I've been struggling to pin down a Windows 10 VM performance issue for a while, which I thought was network related, but appears that it may have been a symptom of this issue. I downgraded to 6.6.7 this evening, and my problems seem to have been resolved. It would appear that the terrible, apparent, download speed may actually have been the VM struggling to save the file to the array (directly to an uncached share, not to the SSD where the VM is located). Although running speed checks seemed to show a good internet download speed (around 70Mb/s), when downloading a file in a browser I was getting speeds in the Kb/s range. Downloading to the array was probably in itself exaggerating other general performance issues now realise I was seeing with general file share usage, which also seems to be much better now. As a bonus, Plex can hardware transcode again on my J4105, which also stopped working in 6.7.x :)
  5. This is also affected me with a Asrock J4105. It used to work, but, I think, stopped with the 6.7.0 update. It still does not work after updating to 6.7.2.