Patriot Scorch M2 SSD 128 GB showing low performance with hdparm

razorbjm · 02-03-2019, 03:51 AM

Hi all fellow, members.
Newly joined member, hoping to expand my knowledge and perhaps even be able to help some others.

I have a small issue understanding the behavior of my machine. It seems that my Patriot Scorch m2 device is having performance issues.
When the Debian system was newly installed (9.0.4) it was performing with around 550 BM/sec. Below is what I get from hdparm now.

hdparm -Tt --direct /dev/nvme0n1

/dev/nvme0n1:
Timing O_DIRECT cached reads: 34 MB in 2.08 seconds = 16.31 MB/sec
Timing O_DIRECT disk reads: 92 MB in 3.03 seconds = 30.39 MB/sec

Any ideas or suggestions would be appreciated. I am not a very advanced user and I guess there might be something I am missing with the setup on this rig.

Current hardware.

CPU : Intel i5-8400 2.80GHz 9MB BOX
MB : MSI B360M PRO-VH
Storage: Patriot 128GB M.2 SSD PCIe Scorch 2280
RAM : 2x4 GB GOODRAM 2400MHz Iridium CL15 White
PSU : MODECOM MC-500-S88 500W

Current Linux version *just updated it now* before was 9.4 for around half a year.

# lsb_release -a
No LSB modules are available.
Distributor ID: Debian
Description: Debian GNU/Linux 9.7 (stretch)
Release: 9.7
Codename: stretch

Info about the M2 disk.

Code:

nvme id-ctrl /dev/nvme0
NVME Identify Controller:
vid     : 0x1987
ssvid   : 0x1987
sn      : DDC3077C0E4C00000746
mn      : Patriot Scorch M2
fr      : E8FM11.4
rab     : 1
ieee    : 000000
cmic    : 0
mdts    : 9
cntlid  : 0
ver     : 10200
rtd3r   : 124f80
rtd3e   : 16e360
oaes    : 0
oacs    : 0x17
acl     : 0
aerl    : 3
frmw    : 0x2
lpa     : 0x4
elpe    : 15
npss    : 4
avscc   : 0x1
apsta   : 0x1
wctemp  : 354
cctemp  : 358
mtfa    : 0
hmpre   : 0
hmmin   : 0
tnvmcap : 128035676160
unvmcap : 0
rpmbs   : 0
sqes    : 0x66
cqes    : 0x44
nn      : 1
oncs    : 0x1e
fuses   : 0
fna     : 0x1
vwc     : 0x1
awun    : 255
awupf   : 0
nvscc   : 0
acwu    : 0
sgls    : 0
subnqn  :
ps    0 : mp:3.00W operational enlat:0 exlat:0 rrt:0 rrl:0
          rwt:0 rwl:0 idle_power:- active_power:-
ps    1 : mp:2.00W operational enlat:0 exlat:0 rrt:1 rrl:1
          rwt:1 rwl:1 idle_power:- active_power:-
ps    2 : mp:2.00W operational enlat:0 exlat:0 rrt:2 rrl:2
          rwt:2 rwl:2 idle_power:- active_power:-
ps    3 : mp:0.1000W non-operational enlat:1000 exlat:1000 rrt:3 rrl:3
          rwt:3 rwl:3 idle_power:- active_power:-
ps    4 : mp:0.0050W non-operational enlat:400000 exlat:90000 rrt:4 rrl:4
          rwt:4 rwl:4 idle_power:- active_power:-

K.R.

razorbjm

Timothy Miller · 02-03-2019, 08:51 PM

What's the drive usage? Give us the output of 'df -h'. Might just be nearing capacity, most SSD's take a MASSIVE hit in performance past ~80% filled.

razorbjm · 02-04-2019, 10:06 AM

Hello Thimothy,

Thanks for the suggestion. Surely you have a very good point there.

Code:

# df -h
Filesystem      Size  Used Avail Use% Mounted on
udev            3.9G     0  3.9G   0% /dev
tmpfs           786M  9.1M  777M   2% /run
/dev/nvme0n1p2  113G  6.0G  101G   6% /
tmpfs           3.9G     0  3.9G   0% /dev/shm
tmpfs           5.0M  4.0K  5.0M   1% /run/lock
tmpfs           3.9G     0  3.9G   0% /sys/fs/cgroup
/dev/nvme0n1p1  511M  132K  511M   1% /boot/efi
tmpfs           786M     0  786M   0% /run/user/1000

I do get increased performance according to hdparm when doing the below on the biggest partition but it is still not quite in the area I had hoped for.

I want to host an application on this machine and I kinda need the m2 SSD perform better than the below.

Anyone who has any idea where else to look I have read some places about /etc/fstab config file and changing the scheduler but I have not knowingly changed any of this since the install but maybe it is needed?

Code:

# hdparm -Tt --direct /dev/nvme0n1p2

/dev/nvme0n1p2:
 Timing O_DIRECT cached reads:    88 MB in  2.01 seconds =  43.73 MB/sec
 Timing O_DIRECT disk reads: 174 MB in  3.00 seconds =  57.97 MB/sec

K.R.
razorbjm

razorbjm · 02-04-2019, 10:19 AM

Quote:

Originally Posted by Timothy Miller

What's the drive usage? Give us the output of 'df -h'. Might just be nearing capacity, most SSD's take a MASSIVE hit in performance past ~80% filled.

Hi Timothy,

You pose a very relevant question there , see below. further more if I do the same test on nvme0n1p2 which is the biggest partition hdparm shows increased performance. But it is still far from the 500+ MB/sec.

Code:

# df -h
Filesystem      Size  Used Avail Use% Mounted on
udev            3.9G     0  3.9G   0% /dev
tmpfs           786M  9.1M  777M   2% /run
/dev/nvme0n1p2  113G  6.0G  101G   6% /
tmpfs           3.9G     0  3.9G   0% /dev/shm
tmpfs           5.0M  4.0K  5.0M   1% /run/lock
tmpfs           3.9G     0  3.9G   0% /sys/fs/cgroup
/dev/nvme0n1p1  511M  132K  511M   1% /boot/efi
tmpfs           786M     0  786M   0% /run/user/1000

Code:

# hdparm -Tt --direct /dev/nvme0n1p2

/dev/nvme0n1p2:
 Timing O_DIRECT cached reads:    88 MB in  2.01 seconds =  43.73 MB/sec
 Timing O_DIRECT disk reads: 174 MB in  3.00 seconds =  57.97 MB/sec