bsc icon indicating copy to clipboard operation
bsc copied to clipboard

geth does not work properly after upgrading to v1.1.12

Open KyloYang888 opened this issue 2 years ago • 6 comments

Hi team, geth does not work properly after upgrading to v1.1.12. Can anyone help me? Thanks~

Start command: geth --config=./config.toml --datadir=/data --cache=18000 --ws --tries-verify-mode none

Logs: image

Block synchronization no longer continues and the computer gets stuck

KyloYang888 avatar Aug 02 '22 04:08 KyloYang888

@KyloYang888 From logs, it was syncing from old blocks (you can see it from ago). Before using new version, did it catch up to the latest block?

forcodedancing avatar Aug 02 '22 09:08 forcodedancing

@KyloYang888 From logs, it was syncing from old blocks (you can see it from ago). Before using new version, did it catch up to the latest block?

Hi @forcodedancing . Thanks~ Before the upgrade, the nodes were able to keep up with the latest blocks properly. It was only because v1.1.12 stopped itself a few times that it fell behind the main network.

KyloYang888 avatar Aug 02 '22 16:08 KyloYang888

@KyloYang888 From logs, it was syncing from old blocks (you can see it from ago). Before using new version, did it catch up to the latest block?

Hi @forcodedancing . Thank you again! I reproduced the problem in the following way.

  1. v1.1.11 is working properly:

Command: geth --config=./config.toml --datadir=/data --cache=18000 --rpc.allow-unprotected-txs --ws --txlookuplimit=0 --diffsync --persistdiff --pipecommit --snapshot=false

image
  1. Then upgrade to v1.1.12:

git checkout v1.1.12 make geth

(Use the same mainnet.zip file as v1.1.11) Command: geth --config=./config.toml --datadir=/data --cache=18000 --ws --tries-verify-mode none image Block synchronization no longer continues

(When I use the new mainnet.zip file) image Block synchronization no longer continues

  1. Use v1.1.11 - Node is working properly

KyloYang888 avatar Aug 02 '22 17:08 KyloYang888

@KyloYang888 Can you remove the flag --tries-verify-mode none (--snapshot=false if use v1.1.11), then start the node. After running for a while, you can start it with geth --config=./config.toml --datadir=/data --cache=18000 --ws --tries-verify-mode none (do not add --snapshot=false).

Fastnode depends on snapshot for syncing, so we need to make sure snapshots are generated before turning on --tries-verify-mode none.

forcodedancing avatar Aug 03 '22 02:08 forcodedancing

@KyloYang888 Can you remove the flag --tries-verify-mode none (--snapshot=false if use v1.1.11), then start the node. After running for a while, you can start it with geth --config=./config.toml --datadir=/data --cache=18000 --ws --tries-verify-mode none (do not add --snapshot=false).

Fastnode depends on snapshot for syncing, so we need to make sure snapshots are generated before turning on --tries-verify-mode none.

Hi @forcodedancing . I tried to start the node as you suggested, but it still doesn't work properly.

Version: image

Block synchronization no longer continues: image

KyloYang888 avatar Aug 03 '22 03:08 KyloYang888

Did it just hang there? Can you change the log level to debug or trace (in config file) to collect more info, to see it hangs on where? Thanks

forcodedancing avatar Aug 03 '22 11:08 forcodedancing

Hi @KyloYang888 I will close this issue for inactivity of it. It you still have issues, please reopen it or create a new one. Thanks.

forcodedancing avatar Aug 19 '22 02:08 forcodedancing