geth follow invalid fork in ETH chain after 15050000 block

Question

I was using an old version ( 1.10.13 ) of geth to run my full node. I didn't update my node to new version before block#15050000. and notice that it stop syncing at 15050010. when I check I found that all blocks after 15050000 are invalid, ( different block hash than blocks accepted by etherscan.io ). I upgrade my client to eth-1.10.20 and then using debug.setHEAD("0xE5A50F") to go back to block no. 15049999, but again it syncs block to invalid chain and stops.

How can I force by node to only connect to trusted nodes or nodes with the valid chain?

Answer 1

I have the exact same symtom with you, stuck at 15050010. Assume you know the Gray Glacier hard forks but I also didn't get it you have updated geth to 1.10.20, which meet requirement as mentioned 1.10.19.

However, I just switched to prebuilt docker version (1.10.23) instead of build it myself with exisiting data, it syncing block 15050305 at the moment.

Some indicator is I saw some new log from geth at bootup which I didn't seen before, like: "Consensus: Beacon (proof-of-stake), merging from Ethash (proof-of-work) "

Answer 2

Hard shut-down and restart geth worked for me.

I was having the same issue. I tried (unsuccessfully) to shut down geth with SIGTERM or SIGINT (graceful shut-down) and it was just acknowledging the signal without doing anything else, so after a while I just sent SIGKILL.

After restart it positioned itself on the right chain and went on with the syncing.