Tracking: Enable zebra-scan to run as a separate independent process from zebrad

[mpguerra]

Motivation

In order to keep private key material outside of zebra, we need to decouple zebra-scan functionality from zebrad.

Design

Some diagrams for the design can be found in https://github.com/oxarbitrage/zebra-grpc-scan-spec/blob/main/zebra-scanner-process.md

We want to get started with the finalized state (reading from the database) part of the blockchain:

• The scanner will communicate with Zebra for the finalized part of the blockchain with a ReadStateService::init method: Add an init function for a standalone ReadStateService #8520
• We need a replacement for ChainTipChange so we can use it in zebra_scanner::spawn_init: Add a current_tip() or similar to ReadState #8580
• Create a new binary for the new process: Create a new binary zebra-scanner #8581
• Get rid of scanner startup code from zebrad: Remove scanner starting from zebrad #8582
• Move or remove the possibility of keys in the scanner: Move or remove the possibility of keys in the config file for scanning #8584
• Testing: Testing the new zebra-scanner binary #8585
• Documentation: Documentation changes for the new zebra-scanner binary #8586

Non finalized state:

• Add a function for syncing blocks from Zebra's non-finalized best chain to a ReadStateService in another process #8576
• Use sync::init_read_state_with_syncer() in the zebra-scanner #8633

We should also evaluate if it makes sense to keep the zebra-scan crate inside the zebra repo as it is now or whether we need to move it to another repo.

Related work

This a step towards enabling zcashd deprecation by replacing the zcashd built in wallet with a zebra backed equivalent as per zcash/wallet#7 .

[oxarbitrage]

Related:

• Run the scanner in an isolated environment #8141
• Create a zebra-db crate #7937
• feat(zebra-db): Split low level database code inside zebra-state #7926

[mpguerra]

We've finished all the required tasks here