1 |
On Sun, Mar 19, 2017 at 3:36 AM, Alan Mackenzie <acm@×××.de> wrote: |
2 |
|
3 |
> Hello, Gentoo. |
4 |
> |
5 |
> I've just bought myself a Samsung NVMe 960 EVO M.2 SSD. |
6 |
> <snip> |
7 |
> Some timings: |
8 |
> |
9 |
> An emerge -puND @world (when there's nothing to merge) took 38.5s. With |
10 |
> my mirrored HDDs, this took 45.6s. (Though usually it takes nearer a |
11 |
> minute.) |
12 |
> |
13 |
> An emerge of Firefox took 34m23s, compared with 37m34s with the HDDs. |
14 |
> The lack of the sound of the HDD heads moving was either disconcerting |
15 |
> or a bit of a relief, I'm not sure which. |
16 |
> |
17 |
> Copying my email spool file (~110,000 entries, ~1.4 GB) from SSD -> SSD |
18 |
> took 6.1s. From HDD RAID -> HDD RAID it took 30.0s. |
19 |
> |
20 |
> <snip> |
21 |
> |
22 |
|
23 |
I was also hoping for more speed up when i got mine, but of course it only |
24 |
helps with the system is IO bound. Its great for loading VMs. |
25 |
|
26 |
It may be mandatory with NVM, but you can check multiqueue is setup/working |
27 |
with; |
28 |
# cat /proc/interrupts | egrep '(CPU|nvm)' |
29 |
CPU0 CPU1 CPU2 CPU3 CPU4 |
30 |
CPU5 CPU6 CPU7 |
31 |
30: 21596 0 0 0 0 |
32 |
0 0 0 PCI-MSI 1572864-edge nvme0q0, nvme0q1 |
33 |
40: 0 12195 0 0 0 |
34 |
0 0 0 PCI-MSI 1572865-edge nvme0q2 |
35 |
41: 0 0 12188 0 0 |
36 |
0 0 0 PCI-MSI 1572866-edge nvme0q3 |
37 |
42: 0 0 0 13696 0 |
38 |
0 0 0 PCI-MSI 1572867-edge nvme0q4 |
39 |
43: 0 0 0 0 11698 |
40 |
0 0 0 PCI-MSI 1572868-edge nvme0q5 |
41 |
44: 0 0 0 0 0 |
42 |
45820 0 0 PCI-MSI 1572869-edge nvme0q6 |
43 |
45: 0 0 0 0 0 |
44 |
0 10917 0 PCI-MSI 1572870-edge nvme0q7 |
45 |
46: 0 0 0 0 0 |
46 |
0 0 12865 PCI-MSI 1572871-edge nvme0q8 |
47 |
|
48 |
If its not setup there'll be just a single IRQ/core handling all the IO. |
49 |
|
50 |
FWIW |
51 |
# hdparm -tT /dev/nvme0n1 |
52 |
|
53 |
/dev/nvme0n1: |
54 |
Timing cached reads: 9884 MB in 2.00 seconds = 4945.35 MB/sec |
55 |
Timing buffered disk reads: 4506 MB in 3.00 seconds = 1501.84 MB/sec |