feat: encapsulate SlicedProcessedDatabase behind ProcessedDatabase trait#2186
Draft
dkales wants to merge 1 commit into
Draft
feat: encapsulate SlicedProcessedDatabase behind ProcessedDatabase trait#2186dkales wants to merge 1 commit into
dkales wants to merge 1 commit into
Conversation
dkales
force-pushed
the
dk/db_loading_encaps
branch
from
5ccc93b to
87e470a
Compare
Introduce a ProcessedDatabase trait covering the per-slice record-movement operations (load_single_record_from_db/s3, write_at_index, preprocess, prefetch_chunk, prefetch_subset) and implement it for SlicedProcessedDatabase. The GPU engine context is passed in via &ShareDB, which already carries code_length and the device manager, so the previous loose n_shards/code_length parameters collapse away. The operation bodies move out of ShareDB / the free write_db_at_index fn into the trait impl; alloc_db (construction), load_full_db (test-only composite) and register_host_memory stay inherent/on DeviceManager. ServerActor's InMemoryStore impl and the dedup/insertion paths now drive the store through the trait. No behavioral change; pure refactor. ServerActor remains concrete over the single GPU implementation (generic/dyn dispatch deferred).
dkales
force-pushed
the
dk/db_loading_encaps
branch
from
87e470a to
59c0cad
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Introduce a ProcessedDatabase trait covering the per-slice record-movement operations (load_single_record_from_db/s3, write_at_index, preprocess, prefetch_chunk, prefetch_subset) and implement it for SlicedProcessedDatabase. The GPU engine context is passed in via &ShareDB, which already carries code_length and the device manager, so the previous loose n_shards/code_length parameters collapse away.
The operation bodies move out of ShareDB / the free write_db_at_index fn into the trait impl; alloc_db (construction), load_full_db (test-only composite) and register_host_memory stay inherent/on DeviceManager. ServerActor's InMemoryStore impl and the dedup/insertion paths now drive the store through the trait.
No behavioral change; pure refactor. ServerActor remains concrete over the single GPU implementation (generic/dyn dispatch deferred).