February 15, 20251 yr Community Expert If the transfer wasn't large enough it could have been cache to RAM, check the actual array write speeds during an SMB transfer, you can toggle that in the GUI: What speed do you see there?
February 15, 20251 yr Author I always have the speed showing. That's how I know it's going slow. This photo is pretty normal for what I'm seeing all the time. The only thing running is Mover.
February 15, 20251 yr Author 3 hours ago, JorgeB said: Turbo write is not enabled. That is one of the first things you told me to set, which I did.
February 16, 20251 yr Community Expert Looking at that screenshot, it's definitely not working, uninstall the CA Turbo plugin and retest.
February 16, 20251 yr Author That screenshot of the entire array, running Mover, was AFTER I removed the CA Turbo plugin, and verifying that Turbo Write was enabled
February 16, 20251 yr Author But, here's a screenshot after you asked for another one. Turbo write IS enabled. CA Turbo plugin removed. No other activity. Here's another screenshot of the disk settings, as well, and for good measure, yet another diagnostics. neo-diagnostics-20250216-1136.zip
February 17, 20251 yr Community Expert Turbo write is enabled in that screenshot, but it's still slow, there appears to be a hard limit at 25MB/s, like if write cache is disabled, which also doesn't look like the problem, sorry but no more ideas, this would require more drastic testing, e.g., remove one of the parity disks and test its write speed outside the array, if normal test the other one, and the disk be written to.
February 17, 20251 yr Author 2 hours ago, JorgeB said: Turbo write is enabled in that screenshot, but it's still slow, there appears to be a hard limit at 25MB/s, like if write cache is disabled, which also doesn't look like the problem, sorry but no more ideas, this would require more drastic testing, e.g., remove one of the parity disks and test its write speed outside the array, if normal test the other one, and the disk be written to. Both parity drives were individually speed-tested with the DiskLocation plugins Benchmark tool. Both drives were comparable with the identical 12Tb drives in the array.
February 17, 20251 yr Author I just did a test by copying a 2.02Gb file from the array to the array-pool via the file manager in Unraid. 100Mb/s+ I then copied the same file from the array-pool directly to a drive in the array. 100Mb/s+ Next, I waited until there was zero file activity and started Mover. 25-26Mb/s I then copied that same file from the array-pool to the array (/mnt/user/media_library) via the cmd line - 25-26Mb/s It's definitely something in the array or parity...
February 17, 20251 yr Author As for testing the parity drives - If I disconnect one at a time (stop array each time, of course), then run Mover, maybe I can find where the bottleneck is?
February 17, 20251 yr Community Expert 2 hours ago, malaki86 said: Both parity drives were individually speed-tested It only test the read speed, not the write speed, which is the problem here. 2 hours ago, malaki86 said: As for testing the parity drives - If I disconnect one at a time (stop array each time, of course), then run Mover, maybe I can find where the bottleneck is? It may work, assuming the issue is one of them.
February 18, 20251 yr Author Solution I stopped the array, popped the bad parity drive out, and set the shares to not use the array drive with the helium issue. That seems to be the cause of it all because it's running great now. Thank you for all the troubleshooting.
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.