Currently we are running the aggregator and relayer setups in the same VM. We should make the setup for those two compatible for metrics tracking. We could have separate grafana and prometheus instances, for example. We could also add the relayers as a separate profile to the aggregator docker compose.
Then, we need to add a relayer metrics dashboard and add grafana jobs for the relayer instances. In that point, it's important to keep things simple at first - it's easy to try and do crazy metrics tracking on panels and lose a lot of time on metrics that are not needed. Cover the basics - i.e. a panel for each metric with labels that indicate fields - and wait for other panels to be actually necessary before doing them.
Currently we are running the aggregator and relayer setups in the same VM. We should make the setup for those two compatible for metrics tracking. We could have separate grafana and prometheus instances, for example. We could also add the relayers as a separate profile to the aggregator docker compose.
Then, we need to add a relayer metrics dashboard and add grafana jobs for the relayer instances. In that point, it's important to keep things simple at first - it's easy to try and do crazy metrics tracking on panels and lose a lot of time on metrics that are not needed. Cover the basics - i.e. a panel for each metric with labels that indicate fields - and wait for other panels to be actually necessary before doing them.