Seed phrase does not recover all on chain funds

My node crashed and won’t give an umbrel login screen. I can SSH in but my SSH password has defaulted to “umbrel” instead of the node password. I have my seed phrase copied down and have verified that it matches the contents of the “state.json” file on the crashed node.

I tried to recover the funds in Blue wallet but it did not recover all of the transactions and nothing newer than two years ago.

Using new hardware, I created a replacement Pi4, 8MB, 2TB SSD, synced the blockchain and input the seed phrase. It recovered transactions up until 9 months ago but not any more recent transactions. I was even able to recover funds from the one channel that was open when it crashed. I also deleted the LND app and reinstalled it, re-scanned, and got the identical result.

Any suggestion on other things I can try? There’s a lot of transactions missing and a significant amount of bitcoin.

From the information you provide, it seems that your node crashed and created new partitions on your SD. This is something that happens frequently when SDs fail and we could check it if it allowed you to access via web (umbrel.local) because it would ask you to configure umbrel as if it were the first time.
Since you cannot access via web, you can confirm this by connecting your SD to your computer and mounting it. If more than 2 partitions appear, then what I told you happened.
If this is the case and you have not done anything else, you can try the following.
Flash the SD. Try to use another one for this test without altering the one that originally had the RPI that failed. This way, you can always go back.
With the new flashed SD, start the RPI. If everything went well and you did not do anything else that could modify/corrupt what was installed on your disk, it will load with everything as you had configured it.
Read the above carefully. If there is something you are not sure about, ask before trying.

Thanks for the reply.

Previous to posting, I tried flashing a new SD card and using it to restart the node. I used the same software version on a new SD card.

The SD card from the crash is still available.

I just tried looking at the drive in disk utility on a Macbook. It shows it to be one unreadable partition of unknown format consuming the entire 1TB. Not sure where to go from there.

I can SSH into the crashed node. Is it possible to examine the partition and file structure from SSH to determine if it has been corrupted?

I was surprised putting the seed phrase into Blue Wallet didn’t recover the on-chain funds fully.

Even more surprised that using the seed phrase into a new Pi4 setup also didn’t recover recent transactions. It basically shows no transactions prior to nine months ago. The recovery process did recover sats from the one lightning channel that was open at the time of the crash.

How is that possible?

Let’s see…

You say:

Previous to posting, I tried flashing a new SD card and using it to restart the node. I used the same software version on a new SD card.
The SD card from the crash is still available.

Just to be clear. This means it didn’t work properly. Is that so?
What version of umbrel did you have installed?

I just tried looking at the drive in disk utility on a Macbook. It shows it to be one unreadable partition of unknown format consuming the entire 1TB. Not sure where to go from there.

Is your SD 1TB?. According to what you said, you were using a 1TB SD and a 2TB SSD?

I can SSH into the crashed node.

Are you doing this with the original SD or the new one?

I’m asking all this because I’m trying to clarify your situation as much as possible. There are things that don’t add up.

On the other hand, “viewing” the file structure via SSH is not necessarily going to tell us if your installation has been corrupted.

As for Blue Wallet, there are a lot of users on the forum in the same situation. Check out some examples:

I can’t help you much there because what I’m trying to do is get you back to having “normal” access to your umbrel. If we can do that, you’ll have a better chance of solving problems.

1 Like

I can see how it might be confusing…

Responses:

Just to be clear. This means it didn’t work properly. Is that so?

It did not work.

What version of umbrel did you have installed?

I’m not certain but I downloaded, flashed and and installed whatever was the latest version on April 19, 2024 without any difficulty. I replaced the 4MB Pi4 with and 8MB Pi4 at the same time.

Is your SD 1TB?. According to what you said, you were using a 1TB SD and a 2TB SSD?

In the crashed node, the only storage device is a single 1TB SSD. It was definitely getting tight on space. I was getting ready to either spin up a new node with 2TB or buy a Home from Umbrel. I had shut down all channels except a small Lightning Watch channel before the crash. I started all this 4 years ago so 1 TB was acceptable.

Are you doing this with the original SD or the new one?

There are two complete systems.

I have the original crashed node with it’s original 1TB SSD. I flashed a new SD card with the same version of UmbrelOS I installed almost a year ago. That box does not allow me to connect and bring up Umbrel but I can SSH into it. I even was able to see the contents of the store.json file to verify I am using the correct seed phrase. Oddly, the SSH password has reverted to “umbrel”. In the past I have used my node password to SSH. I saw a post suggesting the password “umbrel” and it worked.

Then, in an effort to save my on-chain bitcoin, I decided to use the seed phrase in Blue Wallet. It failed to recover 90%+ of my funds and no transactions newer than about 2 years ago.

So I then I created a completely new node from parts I bought a while back to upgrade my node. Its a new Pi4 8MB, new 2TB SSD, new SD card, new case and power supply. I downloaded the most recent version of UmbrelOS and spun up a new system. Synced the blockchain and used the LND recovery process. This got me back most funds but failed to recover anything newer than 9 months ago. The recovery process did successfully close the one open Lightning Watch channel and returned those funds. I also uninstalled LND, reinstalled, reran the recovery process to get a new scan and got the identical results.

I hope that helps. I appreciate the questions and opportunity to clarify.

I’ll look into some of those BlueWallet posts there’s a couple I have not yet read.

Thanks again

So, I think we can forget about recovering the first node, the one that “got screwed”. You have a new one up and running and the urgent thing is to recover the lost funds.
As I’ve seen, there are a lot of people with this kind of problem with blue wallet and not just from umbrelos.
Anyway, let’s forget about the blah, blah, blah… Try this and see if it works for you.

1 Like

So in addition to using the Blue Wallet method, I also tried the Sparrow Desktop method.

Both of these methods resulted in identical results. They recovered some early transactions but nothing after September of 2022 and only about a tenth of the funds recovered by the replacement node.

The all new replacement node has about 60 to 70 percent of the Bitcoin that was on the original node when it crashed. The transactions there do not include the last 9 months.

What changed two years ago. Or 9 months ago?

Are there any sources of expertise that I can hire to help me with recovering those additional bitcoins? Are there valuable files that could assist on recovering funds from the old node?

mmm… I’m sorry you still can’t get it.

I can’t help you further in that regard. I’ve never used lnd. I don’t need it.

However, I’m curious to know how to solve a problem like the one you have because there are so many users who have the same problem with blue wallet/lnd and not just running on umbrelos.

As far as I can understand, it seems that blue wallet (I still can’t understand why, if it’s a bug of some kind) only scans the first 20 addresses whereas your Node generates a new address on every reboot, so total funds may appear to be different. You can check it out at this MyNode link.

I would recommend that, if you haven’t already, you open a ticket on the blue wallet githum.

By the way, check the open tickets… There are people with the same problem as you.

If you achieve your goal, please tell us how you achieved it.

1 Like

Just to clear few things.
You said you have had only one channel during crash, which was later closed with SCB? There were no other channels in waiting or pending states during the crash?
Assuming both answers are yes, then all your funds 100% must be in your onchain wallet, but hard to tell why LND recovery is not picking them up. If number of these onchain transactions were unrelated to channel opening and closing, then that could explain why backup file is not highlighting them, but wallet rescan should fix that.
Much faster and easier is to import wallet to sparrow, but you must import both native segwit and taproot paths. Sparrow is unable to handle different address types in single wallet, so you must treat each address type as separate wallet in sparrow. Then just set large enough gap limit in both (500 should be high enough) and done. All your funds will appear.

I really suspect you have imported only the segwit path. (LND switched to p2tr by default in august 2022 and if you havent been updating often, perhaps it could explain why you miss transactions only since 9months ago and not 2,5years?)
Fyi bluewallet and electrum dont support taproot so obviously it wont see funds in p2tr addresses. Only sparrow does support singlesig taproot atm. I thought everyone knows this.

1 Like

There is promising information in what you are saying.

To clarify on the processes I’ve attempted and the results of recovering my bitcoin:

I first tried putting the seed phrase in Blue Wallet. It recovered funds from April of 2021 up to September 2022. Nothing more recent.

Then I used the recommended procedure in “Restoring your Umbrel node wallet” document that was suggested to import the wallet into Sparrow Desktop (procedure #5). I converted the AEZEED wallet pass phrase and pasted the zprv Base 58 into Sparrow. I got a result identical to the Blue Wallet recovery attempt.

I built a new RPi4 node. New Pi4 8MB, new 2TB SSD, new micro SD card and latest Umbrel flash. I synced the new node and used the LND recovery process, including the channel backup file for the one channel that was open at the time of the crash (2 million sats). This recovered significantly more bitcoin and included transactions up to May of 2024. It also closed and returned the 2M sats from the channel that was open at the time of the crash.

So, it seems your timeline for the P2TR is important. Is there another process I can run to recover those? The thing you mentioned about Sparrow and Taproot is not something I know about.

There are still significant funds missing as I had been doing occasional buys, meaning to move them to cold storage and didn’t…in time.

I’d appreciate any advice you can provide.

To make these funds “visible” in LND, add reset-wallet-transactions=true to lnd.conf and reboot*
*1) you have to remove this line later or it will rescan on every reboot. 2) default recovery window is 2500 so if you would have ie 9000 onchain txs, this alone would not cover them and LND would also have to be started with custom flag together with reset wallet line in config, (ie “lncli unlock --recovery_window 10000”).
LND rescan 2021-2025 with default setting could take idk, I guess prob no less then a day on RPi4… In sparrow you see everything right away.

Import of p2tr path to sparrow is nearly identical as p2wpkh which you are already familiar with.
In settings of newly created wallet, select taproot (P2TR) from “script type” menu → in keystore section go “new or imported software wallet” > paste your xprv to the master private key (bip32) window and in next step make sure derivation starts m/86 and not m/84 (there is minor difference between older and newer sparrow versions in how this is displayed. Newer one is more streamlined, perhaps less confusing so I would suggest that) + import > at the bottom of settings page select advanced + increase gap limit to 500 (excessive, but wont leave space for missing anything) and close > apply…

PS: Unfortunately I haven’t yet figured how to spend both p2tr + p2wpkh inputs in single tx outside of LND. Taproot signing works differently and combining signatures in externally crafted PSBT containing p2tr inputs together with p2wpkh (or other) inputs won’t form valid tx for some reason. So if you would for example insist on emptying / sweepall of your LND wallet in single transaction, you need synced LND… In sparrow, you have to sweep in two steps, one tx sweeping p2wpkh and second tx sweeping p2tr… Detail worthy of being aware of imho.

2 Likes

Your advice on a second P2TR run was successful!

I now have two wallets on Sparrow and plan to move it all to cold storage while I evaluate my next node hardware/software solution.

Thank you for sharing your knowledge and guidance. It was super kind of you to help.