We currently have a manual process to process StateMarketDeals, extract deals that are expected to be retrievable, and update Spark's database. This process does not support recently introduced DirectDataOnboarding deals (see FIP 76).
Let's implement a new solution that will observe actor events (see FIP 83) and update Spark's database of retrievable deals in real-time.
We need at least the following data for each deal:
miner_id
client_id
piece_cid
piece_size
To link deals to allocators, we need to maintain the list of client<->allocator links by listening to on-chain actor events emitted by the AddVerifiedClient method. We should build this component as part of this task.
eta: 2024-09-30
We currently have a manual process to process StateMarketDeals, extract deals that are expected to be retrievable, and update Spark's database. This process does not support recently introduced DirectDataOnboarding deals (see FIP 76).
Let's implement a new solution that will observe actor events (see FIP 83) and update Spark's database of retrievable deals in real-time.
Resources:
We need at least the following data for each deal:
To link deals to allocators, we need to maintain the list of client<->allocator links by listening to on-chain actor events emitted by the
AddVerifiedClient
method. We should build this component as part of this task.Subtasks