feat: update double echo to use pending CF

[Freyskeyd]

Description

This PR is the final one regarding the stabilization of the soft-fork playbook for devnet.

It contains various fixes and changes that allow the network to be upgraded without the need of purging it, allowing the continuity of the delivery while upgrading. This PR biggest change is that the broadcast is actively asking for certificates, instead of receiving them from a channel. It allows us to stack incoming certificates in the Storage and process them at node pace. When shutting down, the broadcast will restart from the next certificate still in pending in the Storage.

Changes per crates

topos

• Update on the node up command, the parsing of the boot_peers from genesis is needed in order to define the is-bootnode config value as soon as possible.
• The is-bootnode is a config value but is not available in the config file. Mostly to prevent anyone of defining is-bootnode = true directly, relying solely on the genesis file.

topos-p2p

• Uses a is-bootnode from configuration in place of the old value
• Add new logs in order to detect corner cases such as kademelia error when bootstrapping

topos-tce-broadcast

• Refactor how the task_manager is handling shutdown
• The task_manager will actively request for new pending certificates on interval (currently 1sec), improvements need to be made on this, it'll be part of another PR.
• Add logs to better understand the flow of a certificate or message inside the task manager

topos-tce-storage

• Add a new variant on the InternalStorageError to catch when a certificate is CertificateAlreadyPending
  • The PendingResult also expose a AlreadyPending
• Add a new method on Map trait to iter_at a specific index, implemented in DBColumn
• ValidatorStore exposes a new get_next_pending_certificates allowing to get the next N certificates in pending
• Logs have been added to better understand the flow of the certificate inside the storage
• Inserting a pending certificate that is already present will return the newly define CertificateAlreadyPending
• The ValidatorStore implementation of insert_certificate_delivered will now correctly remove the child certificate from precedence_pool after adding it to pending_pool
• ValidatorPendingTables is properly fetching the current pending_id from the storage when starting, if not, falling back as expected on 1

topos-tce

• Dump values of the storage when shutting down to better understand the final state of the node before it stops
• Improve various logs in different places (app_context/api and app_context/network mostly)

PR Checklist:

• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added or updated tests that comprehensively prove my change is effective or that my feature works

[codecov]

Codecov Report

Attention: 62 lines in your changes are missing coverage. Please review.

Comparison is base (5b0257b) 69.24% compared to head (29ce9bf) 68.75%.

Files	Patch %	Lines
crates/topos-tce-storage/src/validator/mod.rs	67.34%	16 Missing ⚠️
crates/topos-tce-broadcast/src/task_manager/mod.rs	79.16%	15 Missing ⚠️
crates/topos-tce-storage/src/validator/tables.rs	53.84%	12 Missing ⚠️
crates/topos-tce/src/app_context/api.rs	0.00%	10 Missing ⚠️
crates/topos-tce/src/app_context/network.rs	0.00%	5 Missing ⚠️
crates/topos-p2p/src/runtime/mod.rs	50.00%	2 Missing ⚠️
crates/topos-tce-broadcast/src/constant.rs	0.00%	1 Missing ⚠️
...ates/topos/src/components/node/services/process.rs	93.75%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #418      +/-   ##
==========================================
- Coverage   69.24%   68.75%   -0.50%     
==========================================
  Files         220      220              
  Lines       11969    12109     +140     
==========================================
+ Hits         8288     8325      +37     
- Misses       3681     3784     +103     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[dvdplm]

This seems like a pretty big change. Do you mind adding more info in the description so reviewers can understand the motivation and consequences a bit better?

[dvdplm]

Lots of good micro-improvments here, good stuff.

I fixed a few typos and log messages, but the rest of my comments are questions or topics best addressed in separate PRs. Feel free to respond to question at your convenience (not a blocker to merge).

[gruberb]

Just one nitpick about having a const for the amount of pending_certificates we want to fetch with each tick.

Generally, I would like some explanation how the intented flow is. When we re-deploy the network, how is the pending_certificate help us? Do we do partial shutdowns and then the nodes pick up, and then get shutdown?

[gruberb]

👍