Helios4 Support

[gprovost]

On 6/30/2018 at 1:42 PM, iNTiGOD said:

Also the fans seem to be running at full speed. Is there a way to fix this (I tried the steps earlier in the thread to no avail).

 @iNTiGOD which build / image version are running ? 

Maybe you can refer to the following blog post to check if the files on your system show to correct settings.

[iNTiGOD]

On 7/2/2018 at 2:04 AM, gprovost said:

 @iNTiGOD which build / image version are running ? 

Maybe you can refer to the following blog post to check if the files on your system show to correct settings.

 

I'm using Helios4_Debian_Jessie_4.14.20-OMV_3.0.97.img.xz

 

....Just an update..

I've followed the linked blog post but changing the values doesn't seem to make any difference.

I have updated the kernel and running the command "uname -a" gives the output "Linux helios4 4.14.18-mvebu #22 SMP Fri Feb 9 10:41:38 CET 2018 armv7l GNU/Linux".

For the 4.14 kernel i changed the values to

# Helios4 PWM Fan Control Configuration
# Temp source : /dev/thermal-cpu
INTERVAL=10
FCTEMPS=/dev/fan-j10/pwm1=/dev/thermal-cpu/temp1_input /dev/fan-j17/pwm1=/dev/thermal-cpu/temp1_input
MINTEMP=/dev/fan-j10/pwm1=55 /dev/fan-j17/pwm1=55
MAXTEMP=/dev/fan-j10/pwm1=95 /dev/fan-j17/pwm1=95
MINSTART=/dev/fan-j10/pwm1=50 /dev/fan-j17/pwm1=50
MINSTOP=/dev/fan-j10/pwm1=50 /dev/fan-j17/pwm1=50
MINPWM=/dev/fan-j10/pwm1=50 /dev/fan-j17/pwm1=50

but this seems to have made not difference upon rebooting.

 

Running sudo fancontrol

Loading configuration from /etc/fancontrol ...

Common settings:
  INTERVAL=10

Settings for /dev/fan-j10/pwm1:
  Depends on /dev/thermal-cpu/temp1_input
  Controls 
  MINTEMP=55
  MAXTEMP=95
  MINSTART=50
  MINSTOP=50
  MINPWM=50
  MAXPWM=255

Settings for /dev/fan-j17/pwm1:
  Depends on /dev/thermal-cpu/temp1_input
  Controls 
  MINTEMP=55
  MAXTEMP=95
  MINSTART=50
  MINSTOP=50
  MINPWM=50
  MAXPWM=255

Not sure where to go from here.

 

I have also installed the "Helios4_Debian_Jessie_4.4.110-OMV_3.0.96.img" image instead.

Ran sudo fancontrol and the output is below (I would have thought the results below mean the fans could be turned off.

 

sudo fancontrol
Loading configuration from /etc/fancontrol ...

Common settings:
  INTERVAL=10

Settings for hwmon0/pwm1:
  Depends on hwmon2/temp1_input
  Controls 
  MINTEMP=45
  MAXTEMP=65
  MINSTART=1
  MINSTOP=0
  MINPWM=0
  MAXPWM=255

Settings for hwmon1/pwm1:
  Depends on hwmon2/temp1_input
  Controls 
  MINTEMP=45
  MAXTEMP=65
  MINSTART=1
  MINSTOP=0
  MINPWM=0
  MAXPWM=255

Could it also be that both fans are aftermarkets from ebay? Bad fans? They are 4 pin but perhaps they don't support PWM.

[gprovost]

@iNTiGOD Have you try with one of the stock fan that was part of the kit ? To see if it's an issue related to the new fan you bought or it's just system config.

 

Also did you check the existence and content of /etc/udev/rules.d/90-helios4-hwmon.rules ?

[iNTiGOD]

On 7/4/2018 at 5:10 PM, gprovost said:

[mention=9429]iNTiGOD[/mention] Have you try with one of the stock fan that was part of the kit ? To see if it's an issue related to the new fan you bought or it's just system config.
 
Also did you check the existence and content of /etc/udev/rules.d/90-helios4-hwmon.rules ?

 

No I didn’t use one of the kit fans silly me.
Will try that.
The .rules file was there already.
Will report in a couple days

 

Further testing indicates it is indeed the fan. The single stock fan works a charm.....

Thanks again for your help.

 

[hatschi1000]

Hi all,

I hope this is the right place to discuss this problem I have had with using the helios4. I seem to have stumbled upon a considerably strange issue with using the helios4 NAS (...once I even got to the point of booting, but read on).
After receiving and carefully following the instructions for physical assembly of the device, I came to the point of having the included SanDisk SD-Card flashed with the prebuilt OMV image, I plugged it and all the other required cables in and connected the helios4 to mains electricity. The power indicator LED8 came on and the fans came up to full speed - and that's it, nothing booting, no flashing system heartbeat LED1.
As suggested in the Hardware section of the instructions, I had already connected my 2 HDDs (HGST Deskstar 2TB) to the board. It actually took me quite a while figuring out why the system does not even boot: it appears, that the board might be unable to maintain its own electricity supply with even _one_ HDD connected to power via the molex - SATA power adapter. Once I disconnected the HDD power (leaving the data still connected) it suddenly booted after replugging mains and the heartbeat LED came to life. After some basic setup and finding an adapter to test powering the HDD (only one HDD again, I unfortunately only have an adapter able to power one and not two or even more) via an external supply, the system booted and detected the connected drive. Upon trying to re-plug the power connection onto the onboard molex - SATA adapter, the system instantly freezes and did not react to any input via serial console.
In the beginning I also tried using a different OS image (Debian 9 Stretch Kernel 4.14.20, provided in the download section, build date 17/02/2018) and different SD-Cards with the same result: the system did not boot or start flashing the heartbeat LED1.

I honestly have absolutely no clue as to why the SBC can't maintain its power but I really hope there is a solution to this seemingly power-related problem with being able to power the HDDs from onboard, as otherwise without drives the NAS cannot really be used as a NAS.

[gprovost]

Hi @hatschi1000, sorry to hear you are facing issue with your setup.

 

From your description it looks like a case of surge current from disk spin-up. This could happen if you use 4x HDD that have very high inrush current at spin-up (e.g old generation of HDD, or some 7200rpm Desktop HDD). This issue can be addressed with spin-up staggering technique by using Power-Up In Standby feature of HDD (PUIS mode). What is strange here is that your setup only has 2x HDD and you said you also observed the issue with only one HDD connected, so the issue could be something else.

 

So here couple of questions to try to narrow down the issue (Let's focus on 'the only one HDD connected' use case) :

 

1. What is the exact model of HGST HDD you use (please provide model number)

 

2. Have you tried with both SATA power cables ? just to be sure it couldn't be a cable issue.

 

3. What do you see on serial console when you power up ?

 

4. While system is powered up, and nothing is coming up as you described, have you tried to press the reset button (U16) and see if something coming-up on the serial console ?

 

5. Can you power up the system without the HDD connected (power cable), once OS is up try to connect the HDD is see what happened (Does the system freeze) ?

 

 

 

[hatschi1000]

Hi @gprovost and thanks for your quick reply!

It could indeed be the spin-up current, according to the datasheet (pdf) I've found at the HGST website the disks I am trying to operate require up to 1.2A @ +5V and 2A @ +12V - theoretically combined 2*30W = 60W at startup - which seems, according to a comparison-sheet by a dude over at FreeNAS, like quite a lot.

Let me answer your questions:
 

17 hours ago, gprovost said:

 

1. What is the exact model of HGST HDD you use (please provide model number)

 

The exact model is called HGST HDN724040AL. I decided to opt for them instead of the classic WD Reds as HGST drives were sort of considered "the tanks" among HDDs in general (lower failure rates I were told, that stats might be of interest for some people considering which drive (manufacturer and/or model) to buy, note though: these are the statistics for Q4 2017 as well as 2017 in total, at this page there are quite a lot more observation periods available).
 

 

17 hours ago, gprovost said:

 

2. Have you tried with both SATA power cables ? just to be sure it couldn't be a cable issue.

 

Indeed I did try both SATA plugs on each cable with both cables with both molex-headers on the board with both drives I have got here - already tried to narrow down the issue as good as possible (16 tries in total)
When connecting the SATA power cable while the SBC is running, the drives also don't spin up at all (seems like most of the time they don't even start spinning up as they generally have to me a notable noise)
Also notable: if the system heartbeat is illuminated in the moment the power connection from the board to the drive is established it remains on permanently, same goes for the case when it is off in that moment.
 

 

17 hours ago, gprovost said:

 

3. What do you see on serial console when you power up ?

When powering up without any drives attached it works fine (boot-log when no HDDs are connected is attached)

If there is a single drives' SATA-power connected to the helios4, the system either does not boot at all and the drive does not start spinning up, or it stops fairly soon after starting to spin up the drives, which then also start slowing down again - looks like they then lack power as well.
If the drive starts spinning up and the system starts booting, the serial console (and presumably also the entire system) freezes at about line 20 (varies partially by some lines, I guess sometimes it freezes earlier, sometimes a bit later) of the attached log and that's it then, the mains power has to be disconnected for it to start again.
 

 

17 hours ago, gprovost said:

 

4. While system is powered up, and nothing is coming up as you described, have you tried to press the reset button (U16) and see if something coming-up on the serial console ?

If one drive is connected at powering up (note: serial console remains clear) pressing the CPU-reset button does not change anything: the power indicator remains on, the heartbeat and the OS remains off.
 

 

17 hours ago, gprovost said:

 

5. Can you power up the system without the HDD connected (power cable), once OS is up try to connect the HDD is see what happened (Does the system freeze) ?

It is possible to power up the system without any drives attached, which is also how I tested for the 2nd question (always connected the HDD once the system was properly booted and responsive). It instantly freezes, no heartbeat any more, the LED1 remains as it was the moment the connection to the drive was established.

 

 

As of right now it seems it's likely a issue with the spin-up surge current required by the drive(s), I'm gonna look whether I can activate the PUIS mode for my HGST drives - any advice is still very welcome!

 

Attachment:

helios4-no-hdd-startup.log

 

 

A short update (nr. 1): there seem to be multiple ways to prevent the drive from spinning up rightaway (PUIS as well as "SATA pin 11"), but right now after all the testing the board does not even boot without any HDD attached anymore (neither a system heartbeat, nor any serial output) - lol.

[gprovost]

@hatschi1000 Any chance you can try with another HDD model, just for test purpose. Because it is not normal it doesn't even work with one HDD. Starting to feel something could be wrong with your AC/DC power brick.

 

Issue would only arise if transient current on 12V rail goes above 5.8A and on 5V rail goes above 4.5A simultaneously. So even though Hitachi drives are known to be unusually hungry on the 5V rail, I don't think it's possible one HDD would draw more than 4 Amps at spin-up.

 

Just to be sure, you don't have any USB devices connected to the board while doing the test ?

 

Can you double check that the AC/DC power brick is 12V / 8A ? (No reason it isn't, but well we never know)

[hatschi1000]

@gprovost I could if it booted at very least without any HDD connected, but since pretty much right after the various tests and a couple more tries with booting with/without a drive attached yesterday evening (CET) it unfortunately does not even so: no heartbeat, no boot, no logs generated (I double checked) once mains is connected.
 

 

2 hours ago, gprovost said:

 

Just to be sure, you don't have any USB devices connected to the board while doing the test ?

There was never any USB device connected at any time, neither a drive nor anything else
 

 

2 hours ago, gprovost said:

 

Can you double check that the AC/DC power brick is 12V / 8A ? (No reason it isn't, but well we never know)

Indeed the power brick is 12V/8A (model yczx-1268 manufactured by ???).

 

I got no idea why it wouldn't startup anymore, the power connection seems to work as LED8 is illuminated as well as the right hand yellow LED from the network interface.

From some (admittedly rather amateurishly) measurements I could find that the PSU draws ~ 3W when nothing is connected; the device used as an temporary external PSU via a different molex - SATA adapter cable I've had on hand (male molex to female SATA, the linked device also has a molex output; using the SATA interface on it would have been rather pointless) instead of the included female molex to female SATA cable requires ~25W for several seconds to spin up _one_ of the drives and for the running operation of that single drive between 9 - 10 watts.

[gprovost]

On 7/13/2018 at 6:23 PM, hatschi1000 said:

I could if it booted at very least without any HDD connected, but since pretty much right after the various tests and a couple more tries with booting with/without a drive attached yesterday evening (CET) it unfortunately does not even so: no heartbeat, no boot, no logs generated (I double checked) once mains is connected.

 

You mean the board is not working anymore at all ? If yes, then please PM me and let's figure out if we need to process to an exchange. Something is definitely not quite right either with the board or the PSU.

[iNTiGOD]

FIY my PWM fan issue has now been resolved. I used 2x AVC DS07015R12L 70mm fans. They work a treat (almost silent).

 

I have hit another snag however. I can't seem to set a static IP only dhcp. I have followed the wiki but after i "sudo reboot" it seems the static settings are lost. The Helios4 welcome screen shows no IP and opening the armbian-config shows

"Warning. Configurator can't work properly without internet connection.         Press CTRL C to stop to stop or any key to ignore and continue."

I have taken a look into /etc/network/interfaces and the file is as below

 

# armbian-config created
source /etc/network/interfaces.d/*

# Local loopback
auto lo
iface lo inet loopback

# Interface eth0
allow-hotplug eth0
no-auto-down eth0
iface eth0 inet static
        address 192.168.1.100
        netmask 24
        gateway 192.168.1.1
        dns-nameservers 8.8.8.8

 

I must be missing something here.

 

... a quick update ...

I put the armbian config back to dhcp, then logged into OMV and manually added eth0 with static settings and it seems to have taken?

[gprovost]

@iNTiGOD It seem you resolved your issue ?

 

Based on the content of /etc/network/interfaces, it seems you are running an old build where armbian-config tool was outputting something not quite right for the first 2 lines of eth0 settings.

 

allow-hotplug eth0

no-auto-down eth0

should be replaced by

 

auto eth0

allow-hotplug eth0

 

Have you upgraded your system ? Because in Debian Stretch network will be handle by NetworkManager.

[iNTiGOD]

On 7/19/2018 at 3:51 PM, gprovost said:

Have you upgraded your system ? Because in Debian Stretch network will be handle by NetworkManager.

Yes I upgrade to OMV 4.

[rifanoz]

On 7/13/2018 at 5:23 PM, hatschi1000 said:

I could if it booted at very least without any HDD connected, but since pretty much right after the various tests and a couple more tries with booting with/without a drive attached yesterday evening (CET) it unfortunately does not even so: no heartbeat, no boot, no logs generated (I double checked) once mains is connected.

I have same problems. no heartbeat, no boot. see my picture Helios4

Please help me.

Edited July 24, 2018 by rifanoz

[Igor]

1 hour ago, rifanoz said:

Please help me.

Can you attach microUSB serial console and paste the output? If any.

[gprovost]

@rifanoz as Igor suggested please post here the console output.

 

Which image did you write on the sd card ? Have you followed the install instruction ? (https://wiki.kobol.io/install/)

[rifanoz]

1 hour ago, Igor said:

Can you attach microUSB serial console and paste the output? If any.

I have follow instructions on https://wiki.kobol.io/install/, when I press enter key and wait 5 minutes but picocom not respond.

 

[rifanoz]

1 hour ago, gprovost said:

@rifanoz as Igor suggested please post here the console output.

 

Which image did you write on the sd card ? Have you followed the install instruction ? (https://wiki.kobol.io/install/)

I use the latest OS Images

Debian 9 - Stretch (by Armbian) Build date : 04/07/2018) ... not work.

Ubuntu 18.04 - Bionic (by Armbian) Build date : 04/07/2018 ... not work.

 

[gprovost]

@rifanoz when is the issue appeared ?

 

FYI the boards go through a FAT (Factory Acceptance Test) before being sent out. Which means each board had a test run with all interfaces tested. So I guess the issue didn't appear on day one ?

[rifanoz]

6 minutes ago, gprovost said:

@rifanoz when is the issue appeared ?

 

FYI the boards go through a FAT (Factory Acceptance Test) before being sent out. Which means each board had a test run with all interfaces tested. So I guess the issue didn't appear on day one ?

I receive the machine at Mar 23, 2018. Then I keep it on warehouse.

Yesterday I tried install it. I try to use Debian and Ubuntu, It's work and serial console too. But I have network problem. I can not set IP Address and access via ssh.

So, I write image again on new sd-card and reset machine via U16 push button.

I turn on machine again then the problem is appeared. I replace with the old sd-card, It still not work.

So... What must I do? Please help.

[gprovost]

@rifanoz Ok noted. So even after a full power cycle, which means switch off the power adapter for 2min (time capacitors discharge) and switch back on, still don't make the board output anything on serial port ?

[rifanoz]

6 hours ago, gprovost said:

@rifanoz Ok noted. So even after a full power cycle, which means switch off the power adapter for 2min (time capacitors discharge) and switch back on, still don't make the board output anything on serial port ?

It's still not work.

[gprovost]

@rifanoz ok, please send me a pm, and we see how we can proceed to an exchange.

[rifanoz]

17 hours ago, gprovost said:

Ok noted. So even after a full power cycle, which means switch off the power adapter for 2min (time capacitors discharge) and switch back on, still don't make the board output anything on serial port ?

Based your suggestion and after I am in desperation ...

I turn off the machine, remove the cooler and CPU from socket, let them couple hours on my table.

I reassemble the machine... Turn on...  Tadaa... The machine live and heartbeat again... Serial port connect again.

Thank guys 

[gprovost]

3 hours ago, rifanoz said:

Based your suggestion and after I am in desperation ...

I turn off the machine, remove the cooler and CPU from socket, let them couple hours on my table.

I reassemble the machine... Turn on...  Tadaa... The machine live and heartbeat again... Serial port connect again.

Thank guys 

 

That’s a good news. I should have effectively adviced to try that. Might have a been a bad board to board connection. 

 

@hatschi1000 you don’t want to try the same thing before returning the item ?

[hatschi1000]

On 7/25/2018 at 8:21 AM, gprovost said:

@hatschi1000 you don’t want to try the same thing before returning the item ?

 

I'm gonna give it a try and report back

 

EDIT: bad luck, mine still refused to boot after unplugging the SoC and replugging it after a while
@gprovost looks like I'll have to take you up on the offer of sending it in to you, more in the PM

Edited July 27, 2018 by hatschi1000

[JakeK]

On 6/7/2018 at 7:54 AM, gprovost said:

I'm still having this same issue. I updated to the newest Debian 9 release posted earlier this month and then installed OMV 4. I ran great for a day or so then I powered it off for about a week. Now I get this:
 

Quote

openmediavault 4.1.8.2-1 (Arrakis) NAS ttyS0
Copyright (C) 2009-2018 by Volker Theile. All rights reserved.

To manage the system visit the openmediavault web control panel:

No network interface(s) available

Quote

dmesg output: 

[    3.472840] mdio_bus f1072004.mdio-mii: MDIO device at address 0 is missing.
[    3.473084] mvneta_bm f10c8000.bm: Buffer Manager for network controller enabled
[    3.474435] mvneta f1070000.ethernet eth0: Using device tree mac address 00:50:43:XX:XX:XX

 

Quote

[   13.312342] mvneta f1070000.ethernet eth0: could not attach PHY
[   13.312347] mvneta f1070000.ethernet eth0: cannot probe MDIO bus
[   18.650579] NFSD: starting 90-second grace period (net c0b47780)
[   19.025539] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[   19.027649] mvneta f1070000.ethernet eth0: could not attach PHY
[   19.027654] mvneta f1070000.ethernet eth0: cannot probe MDIO bus

 

I there anything I can do to fix this?

 

Thanks - Jake

 

On 6/7/2018 at 7:54 AM, gprovost said:

Hi guys, sorry for the lack of follow-up lately, was completely caught up in another venture but now it is over. So the focus will come back 100% on Helios4 support and the second campaign.

 

@JakeK I think the problem you mentioned is something that we have encountered before at rare occasion and it seemed to be linked to the Marvell ethernet driver under Kernel 4.14. Could be a bring up sequence issue that makes the Ethernet PHY not probed properly by the SoC. Will investigate and revert to you.

 

@nemo19 It's true that your CPU temps seems a quite high. Meanwhile the SoC die is designed to operate without issue up to 115 degres. Couple of things you could do to help troubleshoot along with what Igor suggested.

1. Can you check the thermal pad is well positioned between the SoC and the heatsink. For that you will need to unscrew the heatsink.
2. Can you explain why the load is constantly 70% ? Is it because of RAID resync or something else ?
3. Any chance you can keep a console open on the serial until the crash you described occurs again ?
4. Can you share your /var/log/messages and /var/log/syslog history, unless you wipe everything already ?

 

 

 

 

 

 

 

[gprovost]

@JakeK Full power cycle doesn’t solve the issue ?

[JakeK]

Actually, that did fix it. I'm not sure why. Thanks for the reply. 

[gprovost]

We have been investigating the issue of the network interface that is sometimes not properly initialized (as reported by @JakeK) and we found the issue. During u-boot board initialization, the network PHY is supposed to be reset by toggling a GPIO (GPIO19). Unfortunately in our u-boot implementation the PHY reset call was happening too early when the SoC pin muxing wasn't completed yet, which means the GPIO pull up and pull down wasn't physically happening.

 

We have added the fix to our u-boot repo  : https://github.com/helios-4/u-boot-marvell/commit/15c179624b28ddab7d212a0ef0571bcec91cf2ed

 

@Igor Any chance you can trigger a build of just Helios4 u-boot and publish the u-boot .deb in the armbian repo ? This way everyone can easily get the fix by doing an upgrade. Thanks.