Hi. Forgive the newbie question!
I have an 8 x 12TB disk Zfs RAID0 array running on the motherboard's SATA connectors on an AMD X399 motherboard. Read speeds top out at about 800 MB/sec. This is slower than I expected. Each disk can sustain over 200 MB/sec so the RAID0 array should, in theory, deliver 1600 MB/sec. But I'm getting half that.
Is the bottleneck the motherboard's SATA connectors?
The X399 chipset is connected to the CPU via 4 lanes of PCIe gen 3 (which can sustain about 4 GB/sec). But perhaps the X399 just doesn't have enough grunt to handle over 800 MB/sec?
Do I need to upgrade to a proper HBA to get closer to 1600 MB/sec?! Or is the bottleneck ZFS?
I have an 8 x 12TB disk Zfs RAID0 array running on the motherboard's SATA connectors on an AMD X399 motherboard. Read speeds top out at about 800 MB/sec. This is slower than I expected. Each disk can sustain over 200 MB/sec so the RAID0 array should, in theory, deliver 1600 MB/sec. But I'm getting half that.
Is the bottleneck the motherboard's SATA connectors?
The X399 chipset is connected to the CPU via 4 lanes of PCIe gen 3 (which can sustain about 4 GB/sec). But perhaps the X399 just doesn't have enough grunt to handle over 800 MB/sec?
Do I need to upgrade to a proper HBA to get closer to 1600 MB/sec?! Or is the bottleneck ZFS?
Last edited:
)