TCB13

received my nvme board and ssd this week, here is my iozone results.
FriendlyArm Nanopi M4v1 4GB board NVME board Armbian Buster Desktop Kernel 5.4.28 Adata XPG SX8200 Pro 1TB Tom's hardware favorite
iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
                                                                                                 random    random                                 
                  kB   reclen       write     rewrite        read      reread        read       write
          102400          4     97170    133733   155927   156368     61290   133851                                                          
          102400        16    254791   325694   337906   339478   134043   323214                                                          
          102400      512    669920   744732   662990   674603   558387   670433                                                          
          102400    1024    658770   710258   677015   705205   627108   776480                                                          
          102400   16384   820732   824027   727671   731739   722866   821623
Power consumption is up from 5.5-6.5watts for 2 and with 1024 and 16384 for 3 watts up to 9.5watts.
Not bad for a lame SBC. I use with Postgres, Timescaledb and Grafana in Docker containers.

Try bigger files. This test does files of 64k. Try files of 100MB 100times.
dd if=/dev/zero of=test bs=100MB count=100 conv=fdatasync I use the gnome-disks utility to benchmark. It shows the performance with big files, but also the latency what is important for small files. That way you get numbers that can be understood.

My results with a NanoPi M4v2 + 8A custom power supply + Samsung NVMe 250 GB EVO 960 PRO:
 
$ dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB, 1.0 GiB) copied, 2.72603 s, 394 MB/s
 
Kernel 4.4.213-rk3399
 
Slow isn't it ? I wonder what can be done to improve this.
 

I like to further add regarding Sabrent 256GB Rocket NVMe i have tested this write speed and read speed same time.
Read 390mb give or take few mb and around 380 write.
 
So i would assume this is the upper limit on the pci2.1 on the m4 of around 780 something considering the NVMe is rated 1050mb write 3100 read. Still alot faster then the ssd i tested through usb3.

It seems, that there is no driver for fan in armbian. So I wrote a little program in C to control fan on the hat. This nice program adjust the fan speed on cpu temperature accordingly. It starts the fan as 60 °C with 30% power. At 70 °C up to 60% and at 80 °C to 90%. You can adjust it within the code easily.
Compile it with "gcc fancontrol.c -o fancontrol". I configured the program as service. If you want it too, do the following steps:
Create a file fancontrol.service with the following content:
[Unit] Description=Fan control [Service] Type=simple User=root Group=root WorkingDirectory=/usr/bin ExecStart=/usr/bin/fancontrol [Install] WantedBy=multi-user.target do the following steps
sudo cp fancontrol /usr/bin sudo cp fancontrol.service /etc/systemd/system sudo systemctl enable fancontrol sudo systemctl start fancontrol  and that's it.
It works perfekt for me.
Have fun.

Linux m4-2 5.4.13-rockchip64 #19.11.8 SMP PREEMPT Wed Jan 22 22:53:38 AEDT 2020 aarch64 aarch64 aarch64 GNU/Linux
 
this is a home built image with X
 

Now I have get the next crash processes (see the log).
Kernel 5.4.13
CPU 1,42/1,8Ghz governor is set to "conservative"
What I doing. I open the webgui from a nextcloud (docker) installation.
 
 
log_crash-20200191.txt

Hi,
First, I use a mean well S-100F-12 (12 Volt 8.5 A) 
https://www.meanwell-web.com/nl-nl/ac-dc-enclosed-power-supply-output-12vdc-at-8-5a-s--100f--12
This one I had some time and I wasn't using it.
I power the NanoPi true the 4-pin connector on the sata hat with a connector used from an old PC power supply.  (All 4 wires connected)
 
 
I had some spare time this week-end and did some test with my NanoPi.  
The problem seems to be more noticeable when I use midnight commander and copy files from usb to the sata hat.
It also shows up when copying files from network to sata true a samba share on the NanoPi.
 
Slowing down the CPU to 1800 ,1600 and 1400 did not make a difference.
 
Temperature found with htop was approx. 42-40°C at max CPU and at 1600 - 1400 about 36°C
I have a large slow turning fan on the NanoPi and a 50mm direct above the sata hat.
 
What I found strange is, that when I started from a cooled down NanoPi the NanoPi worked longer before failing than when is just rebooted and started again. (Or it is my imagination :-))
 
After another tip on https://forum.openmediavault.org/index.php/Thread/30512-Nanopi-crashes-on-file-transfers/ I restarted and used kernel 4.4.
The CPU is now running at 1800 with a temp of 30 a 35°C and I don’t seem to have this problem anymore.
 
At this point I'm not sure it's a power problem or a temperature problem but I'm planning to spend some more time on it this weekend.
There is also still a possibility that I received a faulty NanoPi or hat . . .
 
Any further input about this is always welcome.
 
 

Since I set the CPU frequency to 1.4/1.8Ghz I have no problems.
Look at this topic
 
 

Erueka! Now the wifi is working perfectly.
 
Did fallow part of this instructions https://askubuntu.com/questions/621452/something-keeps-changing-my-wireless-regulatory-domain
 
 

It was the same with me, too. CPU message on the console and nothing works. I use the 5.4.8 kernel.
Currently I use a power supply with 5.3V/4A on pin4/6
https://www.amazon.de/gp/product/B01HRR9GY4/ref=ppx_yo_dt_b_asin_title_o06_s00?ie=UTF8&psc=1
But I'm not entirely satisfied with that either.
I am currently preparing a rack mounting of the NanoPi and will then use this power supply.
https://www.meanwell.com/webapp/product/search.aspx?prod=RS-25
That's stable at 5A.

I didn't observe any crashes because of high load but what I encountered when compiling code were messages on the console indicating (indirectly) that there is something odd with the CPU (sorry, I don't remember the text). I tried to fix that using more powerful power supply, but that did not help. Not even the 12V supply via the SATA-hat. What did help was setting the governor for the CPU frequency to "conservative". Since I did that, I never got any strange messages anymore. So my conclusion was, that the board might have problems handling spikes or large changes in power consumption.
But of course, that's not a prove against crashes...

@Matthias the original issue with tx_delay/rx_delay (as tested with iperf) was fixed in all kernels back then.
The issue with tx programmable buffer length that exhibited itself with Samba is fixed only in 5.x for now.

Could you possibly test it with 4.x on M4v2?

Interesting, I also had problems copying files using the 5.4 kernel on my NanoPi M4V2. I compiled it myself and copies files from the sd-card to and external hard drive connected via the SATA-hat. I used file systems on the hard drive that were able to detect corrupted data on the drive (ZFS and BTRFS) and every time I did a scrub after copying the files, failed CRC-checks within the data were reported, some of them even at the same position on two redundant drives (mirror).
 
I'll try again with the legacy kernel (4.x). Would be greate if it works in this configuration. @piter75: Did you apply your fixes for ethernet also to the legacy kernel? Then it would be much easier to get the data onto the device in the first place.

for the LEDs you can try the following:

ls -l /sys/class/leds
(here as example from a NanoPi Neo2):
lrwxrwxrwx 1 root root 0 Feb 20 21:50 nanopi:green:status -> ../../devices/platform/leds/leds/nanopi:green:status
lrwxrwxrwx 1 root root 0 Feb 20 21:50 nanopi:red:pwr -> ../../devices/platform/leds/leds/nanopi:red:pwr
 
then echo none to the trigger of the led-name: 
 
echo 'none' > /sys/class/leds/[name_of_the_led]/trigger
 
so it should look like this:
 
echo 'none' > /sys/class/leds/nanopi:green:status/trigger
 
You only have to check the led-names of your board and add the commands in /etc/rc.local before "exit 0"

Thank you once again!

Probably ...
To confirm, check it by doing "grep APPARMOR /boot/config-*" .

I try it in kernel 4.4.167.
Thank you for your help.
I am very excited that it can work very well.
But I also try it in the latest kernel 4.20,but get 'don't support the operation'.
I don't know why.
But I can have a good work in the 4.4.167.
Thank you very much!!!

Most Armbian images do not have this problem as well.
 

It should be fixed, but our fix could also be broken - RK3399 has handull of small troubles which is why we actually don't provide end user support yet. When you are approaching with a problem, you always need to provide logs (armbianmonitor -u).

I didn't encounter them by myself, but I think somebody brought up troubles when copying large files via USB3 drives. Check older posts in this subforum.

Problems are outside user space levels otherwise we would notice such problems on other boards, but we don't. Kernel is the king and the troublemaker  

You are using something what was designed for developers and for testing. Neo2 does not have stable builds with mainline kernel yet.
 
This particular error shows that daily automated system is broken and it doesn't provide consistent updates. I am working on fixing it for last two days ... try once again in a week, but beware that you are using unstable kernel, which update might break your system completely. Rather don't upgrade until stable builds are out ...

You are using modern mainline development kernel and those images are experimental. AFAIK this feature is not ported / developed / forgotten to add.