Start9 Node not booting back up after aupdating Operating System -- Please Help

I recently updated the OS for my Start9 node and it turned back on but for the longest time (over an hour) the Services page showed yellow dots beside apps (LND, Thunderhub, etc) instead of going to green. I tried restarting again and the same thing but this time it said “gateway timeout”. There was only a white light on the front of node instead of a white and blue light as usual. The back showed a sold green light to left of LAN cable and a flashing yellow light right of LAN cable. After researching this using AI, I decided to turn off node and let cool down and then turn back on. From this, there was no sound of booting up and again only a white light was on at front of node (with same solid green and flashing yellow lights at back near LAN). Please help!

I should also add that when I try to type in my .local address in browser it says “Server not found” and when I search for the IP using Angry IP it doesn’t show as it used to before this.

UPDATE for everyone’s information (in case this happens to someone else):

I was able to flash the StartOS found at Release start-os v0.4.0 · Start9Labs/start-technologies · GitHub onto a USB drive and then use it to reinstall the StartOS operating system with data preservation (very important). So after a few stressful hours I was able to recover my node fully updated to v0.4.0 with everything fully intact and no losses of Bitcoin.

1 Like

Can you tell me more about your server?

Same thing happened to me. Two Server Pure.

Update appeared to fail on my first test box but maybe I didn’t wait long enough. Tried the usb flash drive method and I was delighted to find the option to preserve my data. It works now.

Second production box ran over night and failed the same way so I’m unfucking it the same way.

That is not working for me. I have the original Server Pure I purchased two years ago. My older Server One had zero issues. Both boxes had roughly same number of services (8 / 6). The Server Pure literally just died on booting after restart. I have a monitor and keyboard attached to it, and it won’t even boot up through the USB image [ x86_64 / AMD64-slim (FOSS-only) ]. I can get inside recovery shell. I can get to a boot menu, but never boots.

  • Edited -
    I was able to re-install from USB boot drive (after selecting a port in the back). I removed the USB, rebooted and am now left here where I will stop:
    Server Pure
  1. Upgrading from 0.3.5.1 to 0.4.0
  2. Used the official x86_64/AMD64-slim installer
  3. Selected “preserve existing data”
  4. Installer completed
  5. On reboot, drops into BusyBox initramfs with:
    Target filesystem doesn't have requested /sbin/init

I had an issue trying to boot via ventoy. Maybe an issue with EFI booting? I had to use isoimagewriter on a dedicated flash drive

1 Like

Thanks guys. Go ahead and continue your upgrades using this guide and a USB thumb drive: Update to StartOS 0.4.0 - StartOS

We’ll look at what’s going on with the Pures.

2 Likes

I’m not sure if this helps: When I start the usb boot restore process it has had to convert the filesystem from ex4 to btrfs on each box. So whatever the OTA update was doing never appears to complete the fs conversion step.

1 Like

Thanks brother. My final assessment on my Server Pure is:
The Btrfs filesystem mounts successfully, and /root/media/embassy/next contains what appears to be a complete Debian root filesystem.

The init chain is intact:

next/sbin -> usr/sbin
next/usr/sbin/init -> /lib/systemd/systemd
next/usr/lib/systemd/systemd exists

However, /root itself contains only media, and the boot process fails with:

Target filesystem doesn't have requested /sbin/init

It appears that run-init is attempting to execute /root/sbin/init, while the complete staged root remains under /root/media/embassy/next. This suggests the initramfs may not be switching/pivoting the staged root into /root before calling run-init.

Cleaned this up for the Start9 engineers.

Hi Stu – I have the Start9 Pure (32GB RAM / 4TB Storage).

Why is everyone called Bruce?

Anyway, there’s currently an issue with Server Pure OTA v040 upgrades. We’re working on it. But if you have a Pure, and you started the process, you’re now forced to continue manually with a USB:

1 Like

Thanks Stu.

See my note above, which speaks to the USB install failure after the OTA start.

Tried again this morning and this is the error from https://start.local
StartOS launch error:

Database Error

error returned from database: role "root" does not exist

OK, I’ve spent some time researching and conferring. It doesn’t look good. Your last point, and the repeated issues aren’t specific to v040 but rather a response to some sort of drive corruption that will have surfaced from the drive reformatting.

Your later message, with the database error, I’d need more info on, because it seems your trying to do something beyond booting the USB and immediately getting that. What you’d need to try it completely wiping the drive rather than preserve, confirm that works. Depending of what services you were running on this on v0351, you might need to factory reset on 0351, restore from backup and try the migration again. If you keep getting hiccups, then it it’s likely going to be a drive issue.

1 Like

Thank you Stu. Will give that a try in a bit.

This looks to have worked @StuPleb. Bow I am faced with Start Fresh or Recover. I paused here, assuming Start Fresh, but wanted your confirmation first. The only data backed up is pre v0.4.0.

Re-read your post and trying recovery now.

Thanks,

Bruce

@StuPleb Factory Reset got me back to 0.3.4~3. When I attempt to upgrade to 0.3.5.1 I get an I/O failure. The error is No Such File or Directory.

I’m not sure why you flashed v0.3.4~3. Don’t do that.

Just to clarify, I didn’t manually flash 0.3.4.1~3. I performed a Factory Reset as instructed, and the recovery image restored the system to 0.3.4.1~3. I haven’t manually installed any older image.

FYI - This is a two-year-old Server Pure. After the factory reset, it automatically came back on 0.3.4.1~3. I didn’t manually flash an older image.

There are no instructions on this page or anywhere, that tells me which image to use for Server Pure?
startos-0.3.5.1-39de098_aarch64-nonfree.iso

  • startos-0.3.5.1-39de098_aarch64.iso

  • startos-0.3.5.1-39de098_raspberrypi.img.gz

  • startos-0.3.5.1-39de098_x86_64-nonfree.iso

  • startos-0.3.5.1-39de098_x86_64.iso

For discontinued hardware like Server Pure, the factory reset documentation could explicitly state which StartOS image to flash (x86_64 vs aarch64, nonfree vs standard) and what version the recovery process is expected to restore. That would eliminate a lot of uncertainty.

Ok @StuPleb I am back at 0.3.5.1. I took a calculated elimination guess on the iso to use. All good. I think I’ll just keep this box at 0.3.5.1 lol