Optimize for profitability
Key insights:
- Profitability depends heavily on the network and peer quality.
- While block proposals are more lucrative than attestations, they are much rarer.
Check for next action before restarting
To see when your validator is next due to make an attestation or proposal pay attention to the Slot end messages in your logs:
INF 2021-05-31 17:46:11.094+02:00 Slot end
topics="beacnde" tid=213670 file=nimbus_beacon_node.nim:932
slot=1304329
nextSlot=1304330
head=cffee454:38460
headEpoch=1201
finalizedHead=077da232:38368
finalizedEpoch=1199
nextAttestationSlot=338638
nextProposalSlot=-1
nextActionWait=4m35s874ms405us837ns
Specifically, have a look at nextActionWait time.
If you're concerned about missing an attestation or proposal, wait until nextActionWait is greater than 4 minutes or so before restarting Nimbus.
You can also use the nimbus-eth2 API.
For example, to check if your validator has a next Proposal slot assigned, run:
curl -d '{"jsonrpc":"2.0","method":"get_v1_validator_duties_proposer","params":[${HEAD_EPOCH_NUMBER}],"id":1}' -H 'Content-Type: application/json' localhost:9190 -s | jq ".result[]" | grep ${PATTERN_WHICH_MATCHES_VALIDATOR_PUBLIC_KEYS}
Subscribe to all subnets
Launching the beacon node with the --subscribe-all-subnets option increases bandwidth and cpu usage, but helps the network and makes the block production algorithm perform slightly better.
To elaborate a little, without this option enabled Nimbus only listens to a subset of the attestation traffic: in particular, Nimbus doesn't listen to all unaggregated traffic but instead relies on peers to aggregate attestations on the subnets it doesn't subscribe to.
With this option enabled, Nimbus listens to all unaggregated channels (subscribes to all subnets). Practically speaking, this means that when producing a block, Nimbus can "top up" the aggregates that other peers have made with it's own unaggregated attestations. This can lead to better packing in some cases, which can lead to slightly greater rewards.