I use Truenas Scale as a backup machine ( version 25.4.1 ). Normally copying 1.5 Tb (some 60.000 files) - my library of books - takes 1.15 hours, and a normal average om my 10 Gb server connection is 250 Mb per second. This morming, suddenly, the average speed is 50 Mb per second or less.
Copying to a ZFS with 4 disks in a paired ( 2 + 2 ) dataset ( 4 Seagate 16 Tb SATA drives). Truenas tells me that the dataset is 71.9 % full, so still amply below the 80 % limit for ‘optimal pool performance’ … 50 Mbb/sec is far from optimal, in my view…
yes, I do; just as a means of testing TrueNas as well; I am the only one using it, and when working on my normal station it doesn’t hurt to make that daily copy;
but, as I said, since yesterday, half way through the process throughput dropped from the normal 230 MB/sec to kilobytes per second, before finally stopping before its regular end… The original files are on 2 SSD’s in raid 0; possibly there are some patches that are difficult to read on these SSDs and it is possible that these influence the copying process… I just copied the whole library again to a local HD and will start the copy over the 10 Gb ethernet line to Truenas now. See if the results are different when copying from the HD.
