Bitcoin Knots Stopped Running

Support and Troubleshooting
1

This is now the second time this has happened to me since I have switched to Knots:

After receiving the same error the first time, I reindexed the chainstate as per the log and patiently waited 12 days for my node to sync to 100%. Once it syncd, I restarted all my apps. Knots stops running and I get the same errors:

Error initializing block database
[error] Aborted block database rebuild. Exiting.
Error initializing block database.

Do I seriously have to go through this whole process again?

bitcoin-log-tail.txt (7.9 KB)

2

what apps are you running?

3

Lightning Node, Datum, Core Lightning, Thunderhub, Electrs, Lightning Terminal, Mempool

4

2025-07-26T14:03:15Z Config file: /data/bitcoin/bitcoin.conf
2025-07-26T14:03:15Z R/W Config file: /data/bitcoin/bitcoin_rw.conf (not found, skipping)

I noticed that it does not find this file when it looks for it.

5

Is there an override I can use?

6

Have you run out of space maybe ?

7

Posted a similar issue , knots stops syncing at 98 percent, and have a 2tb ssd, logs attached

8

Knots have to restart knots often, continues to stop working , anyone have any ideas

thanks

9

I had the same issue with Umbrel, never could figure out what it was.

Since I switched from Umbrel to Start9, I have no problems with Bitcoin Knots anymore. Seems like an issue with the Umbrel specific implementation.

Also be aware that some evil people out of Bitcoin Core are sabotaging Knots, which leads to DoS attacks on Knots on residential internet lines:

1 Like
10

Ya. I suspect that. As of now I am reindexing the entire drive. Moving forward though I am periodically making a backup copy of the blockchain so if it bricks again I can start from a more recent place to avoid the 3 week process.

11

It shouldn’t take you more than 3 days if you are not running another node within the local network.

To increase the download speed, allocate 50%-80% of the unused ram that you see on your dashboard to dbcache in the submenu and restart. For instance, in a computer with 8GB ram, set this to 4096

Other overrides that will speed it up (but do these temporarily in the override section):


par=-1
banscore=10
listen=0
bantime=500

you can find some nodes to connect to manually to boost download sources by finding them at: Bitcoin Network Snapshot - Bitnodes , but you will have to do this several times as discovery is turned off.

You would add them in the override section like:

connect=51.81.56.15:8333

But you should do at least 5 of these especially if they are bandwidth limited

12

But how much ram do you have? As the computer should have at least 8GB of ram.

13

I was having this exact same issue and rebuilt the index at least three times. Finally I allocated more RAM to my VM (I am running umbrel in a Virtual machine).

As soon as I did that I have never had the issue again. I did throw a lot of RAM at it (24Gb to be exact) but I had a lot of spare RAM in the host.

Perhaps this might help the devs to identify the issue.

14

I am using a 8GB pi5. I am using the overrides you suggested. I am not seeing a big difference. Probably because I’m working with 8GB of RAM.

15

What is the memory usage on the desktop dashboard?
And what kind of storage the PI5 is using?
it could be either swap file allocation or disk i/o
What do you get when you run at the terminal:

free -m