So big picture, I have geth 1.10.3 running with lighthouse and a validator in production and it’s running just fine. However, sometime prior to 4 August, I need to upgrade geth to 1.10.4 (or later) to be london-fork compliant. I have a second “test” server running erigon and lighthouse (with no validator) . The version of lighthouse on the production server points to the test server as a secondary eth1 endpoint. Both eth1 nodes are fully synced.
My plan is to stop the geth process, which will cause production lighthouse to switch to using the erigon node. I can the upgrade geth to be london-fork compliant. I assume once geth is back running and resynced, the production lighthouse will switch back (If my assumption is wrong, I guess I can restart production lighthouse between epochs to switch it back).
The problem: The test version of lighthouse cannot connect to the rpcdaemon erigon process, and if the test version cannot connect, then I’m sure prod will fail too. I get an error from both lighthouse and rpcdaemon logs once per minute, every time it tries to connect.
lighthouse: 14:45:52.827 WARN Error connecting to eth1 node endpoint action: trying fallbacks, endpoint: [http://localhost:8545/](http://localhost:8545/), service: eth1_rpc
rpcdaemon: 14:45:52 toprack rpcdaemon: WARN [07-11|14:45:52.837] Served conn=[::1]:52276 method=net_version reqid=1 t=15.011724484s err=”context canceled”
My googlefoo is pretty good, but I can’t seem to find any info on this. Does anyone recognize these errors that can suggest config changes? I pulled erigon from git ~16 days ago.
Any help is appreciate,