We have now identified the issue causing the AppHash errors during the last few days! The issue was caused by bad combination of mempool configurations that are not compatible with EVMOS. This then caused some validators to have different AppHashes for blocks.
We have made a new release v0.5.3 that forces the NoOp mempool configuration for the node.
These are the steps to fix your node:
Stop your node
Update to binary v0.5.3
Do a
wardend rollback --hard
Start your node
Cosmovisor
sudo systemctl stop wardend
cd $HOME
rm -rf bin
mkdir bin && cd bin
wget https://github.com/warden-protocol/wardenprotocol/releases/download/v0.5.3/wardend_Linux_x86_64.zip
unzip wardend_Linux_x86_64.zip
chmod +x wardend
sudo mv $HOME/bin/wardend $(which wardend)