helmsecure / armbian-images Goto Github PK

First, thanks for fixing the disk io issues. I now have my helm running the docker.io/mailserver/docker-mailserver:latest container. It is also running a bind9 nameserver with DNSSEC set up. I am now able to get this score https://en.internet.nl/mail/hosteng.au/872677/

However when I try to run a shutdown -r or shutdown -h, the helm never reboots. Even 30 minutes after this command is executed its still pingable but the ssh daemon is down as I get connection refused f I try to connect to it.

The only reliable way I have found to restart it sync;sync;sync;reboot -f
I have not been able to find any way to turn it off with shutdown.

In the example ipsec.conf file there is the line below that references a file that I am not seeing installed by strongswan. Is this file needed or can you post an example of what is in the file, so that it can be created?

leftupdown=/usr/bin/vpn-updown.sh

This is a long video without any edits but it shows all the steps needed to set up a dns server and mail server in containers on a linux machine (works fine on the helm as well).

I am running this setup on my reformatted helm at home. to follow the steps you need your own domain.
The video is almost 2 hours but it contains all the steps with explanations.

youtube video

I flashed my Helm with the correct Armbian image. How do I know if it's on? The front LED isn't working and I'm not able to determine through IP scanner what the IP address is. I've tried some that aren't recognized with a identification of the device, but have not been able to SSH in. Does Armbian use port 22 to SSH?

Thanks.

I've tried asking this question in different ways over the past couple of months. I've seen it asked by others as well. I haven't seen any direct answer in other ways the question has been asked.

The closest I've seen was related to the Docker issue and a workaround to make Docker work was provided.

My situation is that I can't use an outdated kernel on a Production system.

Is there any way to get information to understand what is needed to build or retrieve or be able to upgrade to a newer kernel?

After successfully flashing the device and rebooting it, when trying to log in via SSH on port 22, the response is 'Connection Refused'.

Restarting the device also doesn't seem to fix this.

Is there something that needs to be done in order for it to recognize the full 1TB size?

v2a with 16 G mmc : Trying to boot NVME and have valid gpt for the NVMe drive. However when rootdev=UUID= is changed in armbianEnv, I see that uboot is not following the directive.
Is there a way to have access to the initial uboot SPI. or have a uboot that is pointed to /dev/nvme* ?
(looking into jtag for further debug)

On v2a with 8 G mmc I was able to boot from nvme (both boot and / )

Thank you.

Is there a way to get a GUI desktop on Armbian running on Helm?

Would it be possible to use a USB-C dock to provide HDMI support and potential for a desktop environment?
I have tried an Anker 7 in 1 dock and it detects the card reader and USB ports but the HDMI output is not.
I could well be barking up the wrong tree but the RK3399 has a reasonably strong GPU built in so I was hoping to get access to it via a dock.

Just wondering if there is still ongoing work to port armbian to the first gen helm?
Would love to use it as a low power home server. (And Tailscale node)

I couldn't get my v2 to flash with the instructions and frankly the CPU hardware isn't so compelling that I care that much. So, I did the following:

• Remove the screws off the bottom of the case
• Remove the circuit board screws and lift out board
• Unscrew the NVMe storage device and save it for another project (e.g. a USB enclosure)
• Remove the two batteries for recycling or another project (e.g. many small flashlights use these batteries)
• Trash the white plastic casing
• Recycle the black metal heat sink as scrap metal
• Recycle circuit board in electronic waste

Good morning all of us with Helm server v1 are still waiting patiently for our software so we can not only use our heavy desk weights as it was intended but also have access to our data. Can you please let use know when we can down load the server software for V1 please.

What is the license for this project and the images? How does one generate the images?

I successfully flashed my 2a version according to the instructions, but now it won't power on.

I'm hoping that someone can help me out. Everyday desktop computer I'm using is LinuxMint 21 based on Ubuntu 20.04, running Cinnamon desktop.

I have a Helm v2a and finally found a use case for this device (since I've already found new email that I trust), downloaded the requisite files, and successfully installed rkdeveloptool and dependencies.

All provided instructions on this site were followed TO A TEE for entering maskrom at least 10 times and the device will not enter it. I'm using the power cable that came with the Helm device, a CAT 6 cable that I hijacked from another device currently on the network, and a USB-C cable that I use for many different devices for both charging and data transfer - in other words, I trust that they're up to the task and can rule out the physical layer WRT cabling matters.

Is there a nuance to this technique or difference from the given instructions that is not mentioned?

Is there an alternative method - even setting jumpers? I've opened the case and noticed no backup batteries were installed, and not sure if this would make a difference. And BTW, what type of batteries would be needed for backup power?

Thank you in advance for any assistance!

Can anyone get a kernel newer than 4.4 to run? I tried upgrading to the Rockchip64-22.11.1 Kernel 6.0 version (edge) and now nothing.

Any advice is appreciated.

I will be reflashing tomorrow it looks like.

Hi there! I have the v1 helm and purchased the v2 helm to serve as a new mail server (to be configured for that later). I have not registered or configured the new v2 helm I just received and went right to the Maskrom steps. Unfortunately, I seem not to be able to detect it as a usb device. I've followed the steps and experienced the following:

1. Power device.
2. Wait until front light flashes Green and Blue consistently
3. Hold power button for 25 seconds (front light stops)
4. Remove power cable, Release button
5. Press power button, replace power cable
6. Release power button after 6 seconds
7. Front light does not start, NIC card flashes green/yellow
8. Can not detect USB device on Mac, or by proxy, my ubuntu VirtualBox VM.

For new devices, never registered or configured, are there any additional steps that need to occur?

Any guidance would be helpful and appreciated!

is it possible to use a Mac running OS X 12.6 as the host for connecting/loading the Armbian images?

Thanks,

Russ

I understand that you want to get this whole Helm issue behind you and move on.
Only reason why I pitched in and supported the project was the trust I put on Giri's "integrity".
Let us know if you are unable to release the firmware for V1 and that would be ok. (really)
Making people wait for something you will never release is not moral.

Hi, I have reflashed my V2A helm, and I have been able to reformat the internal nvdimm disk as well. However, the system seems to hang after some time with heavy disk io activity with an SMP exception. I have moved my email to a docker setup, and every time I try to copy the emails from my VM that has been running my email since I migrated it off the helm server.

For me, it’s straightforward to recreate the problem. Run the following command:
dd if=/dev/zero of=/nvme_mountpoint/data.dd bs=1M count=8000

This command never completes for me. The helm just goes unresponsive after around 3GB written.

I got as far as writing the Helm-v2a firmware to my device, and got the following:

twp@surfboard:~/src/rkdeveloptool$ sudo ./rkdeveloptool wl 0 ~/Downloads/Armbian_22.11.1-build-38_Helm-v2a_bullseye_legacy_4.4.213_minimal.img 
Write LBA from file (32%)
Write LBA failed!

I repeated the rkdeveloptool wl 0 command, but it returned Write LBA failed! immediately.

My device lists FCC ID 15C-2AYVZ256G on the bottom, so I'm confident that I chose the right Armbian image.

At this point, lsusb still showed "Fuzhou Rockchip Electronics Company RK3399 in Mask ROM mode" connected to my system. I tried re-running sudo ./rkdeveloptool db ~/Downloads/helm-loader-build-38.bin. This printed "Downloading bootloader..." on my screen and then did nothing for several seconds (where before it returned almost immediately)

At this point I disconnected and reconnected the power supply in the hopes of putting it in maskrom mode again from scratch. However, holding the power button, connecting power and releasing the button after 6 seconds does not show the Fuzhou Rockchip device in my device listing. When power is connected, the lights on the Ethernet port flash, so the device appears to be receiving power, it just doesn't appear to be responding on the USB port.

Does anyone know if there's a way I can reset the device to a state where I can try flashing the firmware again?

As the 4.4 kernel is past end-of-life - can we get an update on any plans for updating the kernel?

Thanks.

I have followed these directions a dozen times or more. Can you add more detail as to what I am doing wrong. For example:

• After the 25 seconds - do I release the button, or keep holding it in and just unplug and then plug in power?
• Do the 25 second and 6 second times need to be precise?
• Do you need to wait for some amount of time between the 25 second event and the 6 second event, or can you do it immediately?
• Does the Helm server shut down by itself after the 25 second event? What state should the power light be in after the 25 second event but before the 6 second event?

On many of my attempts the power will not come back on (or at least the light is not on) and I need to let it sit for a while and then come back and play around with power / no power before it comes back online in normal Helm mode.

=======================================

5. Remove the power cable, press the power button and hold, insert power cable. Keep the power button pressed for 6 seconds after inserting power and then release. This process will put your Helm into maskrom mode.

I get confused at this point can someone tell me how to get to root. ive tried CD: rkdeveloptool i get an error.

2 go into root of rkdeveloptool
3.aclocal
4.autoreconf -i
5.autoheader
5.automake --add-missing
4 ./configure
5 make

when runing aclocal i get the following error anyone know how to fix the configure.ac

/rkdeveloptool$ aclocal
aclocal: error: 'configure.ac' is required

I have not seen any messages from the helm team here since late December. Are you guys still around or is these images now set adrift?

Questions:

After flashing the firmware to Armbian, does the power button actually function? And if not, what is needed to make it do so? The LED on front does not activate, but I can see a green glow through the casing and rear grill holes of an LED on the board that does not shut off after pressing the power button numerous times/ways.

Secondly, I remember that there was an issue with Helm Servers overheating before a fix was issued. After flashing to Armbian I have found my unit to become unresponsive after a period of time and when I went to examine it, the unit was really really hot. Could you tell me what sort of thermal monitoring/throttling was implemented so that I can do so similarly now?

And I guess, it’d be nice if I could find out how to control the front LED as well.

After flashing the helm my NIC port is no longer working. Lights flash but I never get a green link light flashing. Is there a factory Img to flash to return Helm V2A to factory defaults. All indications in terminal appear to suggest the flash took. I tried an external USB C NIC, but still no go. Since initially installing Helm as an email service I had my router assign it a static IP, so I know which IP address it should be asssigned. Any suggestions?

Will the Ambian flash work on the 1st generation of Helm?

The device refuses to boot after updating armbianEnv to disable unified cgroup support (as recommended to enable docker support).
Steps to reproduce:

1. Flash Armbian_22.11.2-build-48_Helm-v2a_bullseye_legacy_4.4.213_minimal.img
2. Log in as root, run through the initial setup process
3. As recommended in #1 (comment), add a new line to the end of /boot/armbianEnv.txt containing: extraargs=systemd.unified_cgroup_hierarchy=0
4. Update the boot image, also as recommended in that post:

root@helm-v2a:~# mkimage -C none -A arm -T script -d /boot/boot.cmd /boot/boot.scr
Image Name:   
Created:      Sat Feb 25 10:01:16 2023
Image Type:   ARM Linux Script (uncompressed)
Data Size:    3326 Bytes = 3.25 KiB = 0.00 MiB
Load Address: 00000000
Entry Point:  00000000
Contents:
   Image 0: 3318 Bytes = 3.24 KiB = 0.00 MiB

5. Reboot the helm
6. The helm never comes back on the network; neither ping nor ssh connection is successful even after several minutes.

This started after I reflashed to pick up the fix for #10; prior to that change I was able to update armbianEnv and use docker.

I've tried running apt update && apt upgrade before regenerating the boot image, but that didn't make any difference. I did this both with the default APT sources and with switching from https://github.com/HelmSecure/armbian-packages to https://github.com/HelmSecure/armbian-packages-staging, as armbian-packages tries to downgrade uboot from 48 to 38.

If there's any other experiments that I can run to help shed light on this, please let me know.

Step 3 of the installation instructions say "Connect your Helm to power and to your router via ethernet". Is it really required to have the Helm connected via Ethernet for the bootstrap process, or can it be bootstrapped over USB and then updated directly via wifi?

There are other projects that use this SoC such as Orange Pi 4.

Could you give a summary of what special drivers/kernel mods/etc. are unique to Helm hardware vs other SBCs with the same SoC?

Could we theoretically load Debian or another distro that supports the right SoC or is there custom work included in the Armbian images provided?

Was any other SBC as reference or even an SoC vendor reference design used when designing the Helm?