feat: route commands to correct persistence service #9435
Conversation
Rjected
added
C-enhancement
Rjected
force-pushed
the
dan/use-higher-level-sender
branch
2 times, most recently
from
9f2ae54 to
7083c5e
Compare
|
This is RFR now, merge conflicts resolved with recent tests, but the |
Rjected
force-pushed
the
dan/use-higher-level-sender
branch
from
9a5f979 to
bc38020
Compare
| // Write state and changesets to the database. | ||
| // Must be written after blocks because of the receipt lookup. | ||
| let execution_outcome = block.execution_outcome().clone(); | ||
| execution_outcome.write_to_storage(&provider_rw, None, OriginalValuesKnown::No)?; |
There was a problem hiding this comment.
btw, I find this api a bit horrible cc @joshieDo
There was a problem hiding this comment.
yeah i agree, should be trait/type impl
noted
| let static_file_service = | ||
| StaticFileService::new(provider_factory, static_file_service_rx, database_handle); | ||
| std::thread::Builder::new() | ||
| .name("Static File Service".to_string()) | ||
| .spawn(|| static_file_service.run()) | ||
| .unwrap(); |
There was a problem hiding this comment.
hmm, why are those two separate threads?
There was a problem hiding this comment.
mainly for organization, but they could be combined
There was a problem hiding this comment.
Going to leave combining them into a followup, should not be hard
| { | ||
| /// This is the main loop, that will listen to database events and perform the requested | ||
| /// database actions | ||
| pub fn run(mut self) { |
There was a problem hiding this comment.
Do we need a loop here? according to spawn_services this run method is going to be run on a separate thread, not polled?
There was a problem hiding this comment.
yep, this will be handled by the while and sender is not expected to be dropped, nvm
There was a problem hiding this comment.
this has also made me realize that the current run loop implementation in EngineApiTreeHandlerImpl was wrong, fixed it and updated the test here 0b100ee
feat: add spawning for the persistence task rm duplicate key fix compilation fix compilation again fix docs fix test receipts for single default block
Co-authored-by: Federico Gimenez <[email protected]> Co-authored-by: Matthias Seitz <[email protected]>
Rjected
force-pushed
the
dan/use-higher-level-sender
branch
from
177ff25 to
f0e2538
Compare
|
going to merge this, and start working on some tests for existing code + scope out remaining work |
|
remaining work includes the "on reorg truncate static files" right? |
This refactors the persistence related services, renaming the
Persistenceservice to the database service. This is because it mainly writes to the database. This also introduces a newPersistenceHandletype, which sends to the static file or database service, depending on the command. This type should be used for consumers of the db / static file services.TODO: