mina-indexer

mina-indexer copied to clipboard

Consolidate logic for precomputed block filenames with format `<network_name>-<state_hash>.json`

2

Isaac-DeFrain

robinbb

Need tooling to compare our ledgers against the AN checkpoint ledgers

1

trevorbernard

Store raw JSON bytes instead of the serialized `PrecomputedBlock` bytes

5

`BlockStore` function `add_block` currently takes a `&PrecomputedBlock` and we store the bytes from `serde_json::to_vec()` in the db. We should store the `contents` of a `BlockFileContents` instead.

Isaac-DeFrain

We need docs to show users how to use the image we produce. Relates to #669

jhult

Make `prune_interval` more clear to users

6

'prune_interval' is the multiple of the height of the transition frontier at which tree pruning occurs. The default is 10. Perhaps this is too high? Or maybe no matter? Why...

robinbb

Moving from `flake-utils` to `flake-parts` will enable us to split out flake definitions in a more reusable and composable fashion. One key benefit is you also get type checking.

trevorbernard

Database column family names should use a consistent scheme

1

```rust const COLUMN_FAMILIES: [&'static str; 14] = [ "blocks", "blocks-global-slot-idx", "lengths", "slots", "canonicity", "commands", "mainnet-commands-slot", "mainnet-cmds-txn-global-slot", "mainnet-internal-commands", "events", "ledgers", "snarks", "snark-work-top-producers", "snark-work-top-producers-sort", ]; ``` Improve the naming consistency for projections,...

trevorbernard

Log when doing re-orgs in database

3

I recommend logging start and end time. We probably also want to periodically output a log statement about the progress.

jhult

Isaac-DeFrain