If you have a hdd drive “T” and you have a mount point on that drive for another hdd drive “TT”, when writing plots to both drives simultaneously, will you encounter any type of throughput bottleneck like you would if you were trying to write two plots to the same physical hdd drive at the same time?
I would have guessed not but there seems to be something slowing it down.
Thanks in advance.
The drive mounted to “TT” should perform solely based on its physical connection, as well as its own physical characteristics.
If it would normally get 250 MB/s while connected to a port, then regardless of its OS mount point, it should perform at 250 MB/s.
Mechanical drives do slow down as they fill up their space. I believe that they initially write to the outer portion of their platters, and work their way inwards as space fills up, with the inner portion having less real-estate per rotation, and therefore lower read/write speeds.
Have you tested the drive with a simple drive letter and compared its performance to mounting it as you described?
It could be general I/O related bottlenecks. On my old clunker box, writing to two or three drives at once bogged down quite a bit - I never got the full write speeds if the drives were sharing in SATA land. I believe you are correct in that mount points shouldn’t matter, the computer is smart enough to not push data through the other drive first (unlike a plumbing situation where water would for example).
Thanks for the replies. I’ll do some more testing on it this weekend. The box has a 3950x with 128GB ram but the thought occurred to me that I might be overlooking some other type of bottleneck inherent in using the mount point on the same drive. Both SATA on adaptec 16 port expander.
Your replies helped confirm what I was thinking so thanks! If I reproduce and figure it out I’ll update for future farmers.