Chris Hager

I think it makes sense to have the header with tx root as part of the submission, rather than forcing the relay to compute the header after submission. Wonder whether,...

should this PR point to the branch of #154 instead of `main`? not sure

Noticed that the tests fail: https://github.com/flashbots/suave-geth/actions/runs/7471328612/job/20331422338?pr=154#step:9:2848

When I do `make devnet-up` i get the following log output which seems all is good: ``` [+] Running 11/11 ✔ Container redis_db Started 0.8s ✔ Container sig-provider Started 0.7s...

Why does the frontend container need 1-2 minutes to start? This seems unreasonably long 🤔 We'd also appreciate if we could collapse the seven blockscout containers down to a single...

Trying an index on `(inserted_at, relay)`: ```SQL CREATE INDEX CONCURRENTLY mainnet_data_api_payload_delivered_insertedat_relay_idx ON mainnet_data_api_payload_delivered (inserted_at, relay); ```

Added some more logging in #31 Looks like the slowness is in this step, when querying top builders per relay: https://github.com/flashbots/relayscan/blob/eb0bb8797491a3e73a0d9abc369683fd36e3a6b6/services/website/webserver.go#L186-L196 And the DB query: https://github.com/flashbots/relayscan/blob/eb0bb8797491a3e73a0d9abc369683fd36e3a6b6/database/database.go#L124-L135 Logs: ``` time="2024-02-24T10:50:27Z" level=info...

it's only taking about 11 seconds nowadays, after updating the indexes ``` time="2024-05-29T08:22:57Z" level=info msg="updating 7d stats..." time="2024-05-29T08:22:57Z" level=debug msg="- loading top relays..." time="2024-05-29T08:23:00Z" level=debug msg="- got top relays" duration=3.681666368s...

Ideally also warn if the last slot in the database is more than 30 min old

Related work was done in #44 (show time since last update)