| 1 |
> I think q0 is the "administrative" queue, and the other 8 are ordinary |
| 2 |
> queues. (Sorry, I read that somewhere, but can't remember where). |
| 3 |
> |
| 4 |
|
| 5 |
Good to know. |
| 6 |
|
| 7 |
> |
| 8 |
> > .... but on your system i'd say there should be some queues on other |
| 9 |
> > interrupts to they can be serviced by other cores, so that doesnt look |
| 10 |
> > right. |
| 11 |
> |
| 12 |
> It wasn't right. |
| 13 |
> |
| 14 |
> > Do you have MSI enabled? Bus options -> PCI Support -> Message Signaled |
| 15 |
> > Interrupts (MSI and MSI-X) |
| 16 |
> |
| 17 |
> I didn't have MSI enabled, but do now. I now get 5 queues (nvmeq[0-4]) |
| 18 |
> on four interrups, and the interrupts are spread over the four cores |
| 19 |
> (I've got a 4 core Athlon). |
| 20 |
> |
| 21 |
> > If your system is not too old you may get more interrupts or a better |
| 22 |
> > spread with that enabled. |
| 23 |
> |
| 24 |
> Yes, that is the case. From a "warm start", I was able to copy my 1.4GB |
| 25 |
> file in 2.76s. This is a bit more like it! |
| 26 |
> |
| 27 |
|
| 28 |
Nice! |
| 29 |
|
| 30 |
> |
| 31 |
> > You look like you're getting SATA speeds, but since you have the nvme |
| 32 |
> > device, i guess that implies you havent fallen back to SATA. |
| 33 |
> |
| 34 |
> I don't think the SSD has a SATA interface. |
| 35 |
> |
| 36 |
|
| 37 |
That was a guess on my behalf - i know that M2 disks can come in SATA or |
| 38 |
NVMe, but I dont know if NVMe drives can fallback to SATA. |
| 39 |
|
| 40 |
|
| 41 |
> |
| 42 |
> > Could well be older hardware or less PCIe slots/lanes. |
| 43 |
> |
| 44 |
> My $ hdparm -tT /dev/nvme0n1 speeds haven't improved since enabling MSI |
| 45 |
> in the kernel. |
| 46 |
> |
| 47 |
|
| 48 |
Ah well, its only a simple synthetic benchmark. |
| 49 |
|
| 50 |
Adam |
Archives