core, trie/triedb/pathdb: adjust pathdb for verkle #28297
Closed
Conversation
This file contains 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
|
From the metrics, fortunately we don't have any clean cache false which is expected. 
|
fynnss
reviewed
Oct 10, 2023
gballet
approved these changes
Oct 10, 2023
rjl493456442
force-pushed
the
pbss-verkle
branch
from
415f5c3
to
c4f2625
Compare
rjl493456442
force-pushed
the
pbss-verkle
branch
from
84332cd
to
b22d3f2
Compare
rjl493456442
force-pushed
the
pbss-verkle
branch
from
a4ccd19
to
46c1ced
Compare
rjl493456442
force-pushed
the
pbss-verkle
branch
2 times, most recently
from
7676345
to
59e7d50
Compare
rjl493456442
force-pushed
the
pbss-verkle
branch
from
59e7d50
to
4fc4949
Compare
|
I agree that we need to get this in, however, later rather than sooner. The type of error that we encountered here: #28595 , we would not have been able to detect that error if this PR had been already merged. |
rjl493456442
force-pushed
the
pbss-verkle
branch
from
4fc4949
to
b122acb
Compare
|
Replaced with #29084 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
This pull request is a prerequisite for landing verkle.
Specifically, verkle and merkle are two independent trees which may be existent at the same time. In order to store tree data separately, we need to explicitly make the isolation in database level, for both key-value store and state freezer.
Besides, verkle uses a slim format which doesn't store the node hash in the parent node. This slim format can significantly reduce the read and write amplification, but in the same time loses some robustness. However, we think it's still worthwhile to at least experiment it to see the performance difference. In order to support it, pathdb will no longer maintain the node hash which is also beneficial by maintaining less stuff in memory.
But one thing is noteworthy, pathdb will return the node blindly where ever it's found without hash comparison(clean cache, dirty cache, diff layer, disk). Originally we have this tiny trick to compare the hash if the node is found in clean cache, and fallback to disk if it's not matched, just in case the clean cache is not properly maintained, but this mechanism is lost in this pull request.
Update: this pull request can snap/full sync the mainnet, we have more confidence that the clean cache is properly maintained and safe to have this change.