es-node did NOT alarm when failed to download the BLOB

When L1 RPC was misconfigured (did not use --subscribe-all-data-subnets), the BLOB couldn't be downloaded after Fusaka upgrade, but es-node didn't alarm, simply printed an error message, e.g., http://65.21.133.53:4200/eth/v1/beacon/blob_sidecars/13164810?indices=3

We should alarm operators by crashing the es-node or sending an email to notify them after retrying several times.