It would be great for the Platform chainhook nodes to have some replication - two Chainhook nodes processing blocks and keeping up with the stacks/btc chains, but only one processing predicates and reading/writing from the redis store at a time.
Then, if one of the nodes has a problem handling a reorg or falls behind chaintip, we can swap out the active environment while still processing the same predicates.
It would be great for the Platform chainhook nodes to have some replication - two Chainhook nodes processing blocks and keeping up with the stacks/btc chains, but only one processing predicates and reading/writing from the redis store at a time.
Then, if one of the nodes has a problem handling a reorg or falls behind chaintip, we can swap out the active environment while still processing the same predicates.
See this comment for more details.