reth: TransactionLookup stage hanging
Describe the bug
Database sync is stalled at progress=none on TransactionLookup. ForkChoice messages are stopped and the validator reports the EL is offline (reth is paired with my backup beacon node).
Steps to reproduce
This node performed sync to stage 4 on the beta1 and hung. I upgraded to beta3, the sync started progressing again, with ForkChoice messages, until it hung here on TransactionLookup. The message “local database is uninitialized” is there when I run reth db version
smaller_truncated_reth_upload.log
Running: Fedora Server 38 Lighthouse v5.1.0 reth 0.2.0-beta.3 mev-boost 1.7
Node logs
I am attaching my log
Platform(s)
Linux (x86)
What version/commit are you on?
0.2.0-beta.3
What database version are you on?
Current database version: 2 Local database is uninitialized
What type of node are you running?
Full via --full flag
What prune config do you use, if any?
No response
If you’ve built Reth from source, provide the full command you used
No response
Code of Conduct
- I agree to follow the Code of Conduct
About this issue
- Original URL
- State: closed
- Created 3 months ago
- Reactions: 1
- Comments: 18 (3 by maintainers)
I have the same problem, I also updated from v0.2.0-beta.1 to Beta.2 and Beta.3.
It takes just over an hour and does not increase progress, it show “stage_progress=None”
Thanks
Confirmed that with the v0.2.0-beta.4 , the synchronization completes correctly. There was no need to update to beta5.
Thanks!
~hung again on TransactionLookup, checkpoint = 0 :(~
edit: Well, it jumped from checkpoint=0 to 100% complete and finished the pipeline!
hey guys, wanted to report that for my beta 4 syncing reth node, the TransactionLookup stage stopped hanging and I am past this stage now after I realize I haven’t upgraded my lighthouse to a 4844 compliant version (pretty dumb i know…)
note that, my backfill sync target block is 19426523, which is a bit before 4844 goes live.
so what i would suggest to people having the same problem is to check if your consensus client is the latest version, if not upgrade it and restart reth to see if it syncs through the tx lookup stage now.
now the question is, does during TransactionLookup stage execution client must need to communicate consensus client in this stage, and if the consensus client is stalled due to it being a pre-4844 version it cause reth sync also hangs?
I too have this issue. The lovely folk in Telegram asked to post trace logs from net::session, so here they are. Reth Version: 0.2.0-beta.4 Lighthouse version: v5.1.2-5ce1619
reth.log
Hey, I have the same issue at 0.2.0-beta 3, it seems that the stage is stuck and nothing gets done after calculating transaction hashes is completed (I do not know how many transactions in total in Ethereum mainnet, therefore actually not sure if all transaction hashes have been calculated).
Here are my logs captured at the end of calculating transaction hashes:
Regarding
reth db versionIt shows that:I also tried
reth stage drop tx-lookupandreth stage drop merkle, which it syncs the merkle stage again but still stuck at TransactionsLookup.I am now upgrading to 0.2.0-beta 4 to see if it still happens, will provide update later.
update: the issue is still here despite upgrading to beta 4