-
Notifications
You must be signed in to change notification settings - Fork 202
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Corrupted deposit history detected / Resyncing loop #5145
Comments
Update: clearing the suggested ~/.besu/caches directory didn't work for me: Jun 29 13:47:37 ethVal2 nimbus_beacon_node[184]: ERR 2023-06-29 13:47:37.333+02:00 Corrupted deposits history detected |
Really like to use nimbus as my beacon node as it's so nice on resources, but right now it makes the system run hot. Would it be an idea to go back and try history:archive instead of prune mode or maybe skip the trusted sync? |
Same problem here. Besu version: 24.3 log:
I've tried deleting besu cache, or disabling besu cache with I found the query to get deposits was like the following:
So I tried to compare my besu node's response with infura's. So I wrote a script to compare all the response from 0x10a0000 to the latest, and found no difference at all. That made me to think it is nimbus's problem. |
If could be that the initially imported deposit snapshot was incorrect (doing a new checkpoint sync should fix it).
|
Thanks for your reply.
Another thing that might be interesting: The difference between 'ourDepositsCount' and 'targetDepositsCount' does not stay fixed. The difference is fixed for a given targetCount, but as epochs advance and targetCount changes, the difference between these two will change. I have recorded the following logs:
I'm not very familiar with the machanism, but I think if there are some fixed deposits that are missing, the difference should stay a fixed number. |
I did resynced deposit snapshot. It worked for now. |
Great to hear that this is working now! Do you know whether the previous deposit snapshot was correct? If there is a way to reproduce the issue, we could investigate it more thoroughly. |
Luckily I have a copy of the previous deposit snapshot, in two format. From API:
From DB:
I see the snapshot was quite old. EL block 17403122 is roughly around the time when I setup the node. It seems it was not updated ever. I remember every time when I restart nimbus before, it would sweep the deposits from that snapshot point, and became running quietly after the sweep. It was since recent weeks that I noticed it keep sweeping and showing the error after each time. |
@arnetheduck thank you for looking at the issue, but it closed a bit too soon I think :) Follow of issue #5136 as it got closed.
I do:
then start the node with systemd
After it does the backfill for a few hours all is good for a few minutes and then this happens:
After half an hour it's in sync again and after a few minutes the process starts all over again in a loop.
When I look in the besu log I see this:
Now I found the "Corrupted deposits history" it seems I am not alone: #4669 @ibhagwan (found a solution yet?)
edit: now testing the clear .caches folder from besu.
The text was updated successfully, but these errors were encountered: