Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat!: implement FillSquare method for the testnode #866

Merged
merged 8 commits into from
Oct 13, 2022

Conversation

evan-forbes
Copy link
Member

This PR

  • adds a modified FillSquare method for the testnode that only submits a single large message
  • modifies the default configs for a testnode's mempool to accept large transactions
  • modifies the PostData method to accept a flag that determines if the process blocks

The issue with the old approach was that it was submitting too many transactions for a block with such short block times. The testnode was unlikely to get all of the transactions included in a single block, particularly when waiting to submit each transaction until the previous one was accepted into the mempool.

@evan-forbes evan-forbes added the testing items that are strictly related to adding or extending test coverage label Oct 13, 2022
@evan-forbes evan-forbes self-assigned this Oct 13, 2022
@codecov-commenter
Copy link

codecov-commenter commented Oct 13, 2022

Codecov Report

Merging #866 (74cbca3) into main (be5406b) will increase coverage by 0.16%.
The diff coverage is 65.21%.

@@            Coverage Diff             @@
##             main     #866      +/-   ##
==========================================
+ Coverage   25.59%   25.75%   +0.16%     
==========================================
  Files          75       75              
  Lines        9273     9287      +14     
==========================================
+ Hits         2373     2392      +19     
+ Misses       6677     6675       -2     
+ Partials      223      220       -3     
Impacted Files Coverage Δ
pkg/prove/proof.go 83.56% <ø> (-0.12%) ⬇️
testutil/testnode/node_interaction_api.go 60.00% <60.00%> (-0.23%) ⬇️
pkg/da/data_availability_header.go 81.14% <100.00%> (+2.45%) ⬆️
testutil/testnode/full_node.go 75.29% <100.00%> (+0.29%) ⬆️
x/payment/types/wirepayfordata.go 68.69% <0.00%> (+2.33%) ⬆️
pkg/shares/non_interactive_defaults.go 100.00% <0.00%> (+5.17%) ⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

Copy link
Member

@rach-id rach-id left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, only lefts comments to understand more

testutil/testnode/full_node_test.go Show resolved Hide resolved
testutil/testnode/node_interaction_api.go Outdated Show resolved Hide resolved
}
maxShareCount := squareSize * squareSize
// we use a formula to guarantee that the tx is the exact size needed to force a specific square size.
msgSize := (maxShareCount - (2 * squareSize)) * appconsts.SparseShareContentSize
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

[Non Blocking][Question]
The (2 * squareSize) refers to the size that the transaction will take, aside from the message?

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Assume squareSize=16 so maxShareCount=256:

msgSize := (maxShareCount - (2 * squareSize)) * appconsts.SparseShareContentSize
msgSize := (256 - (2 * 16)) * appconsts.SparseShareContentSize
msgSize := (256 - 32) * appconsts.SparseShareContentSize
msgSize := (224) * appconsts.SparseShareContentSize

I don't think the transaction is guaranteed to take (2 * squareSize) but I think it is an approximation for the other shares in the square (transactions, intermediate state roots, evidence).

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah, what Rootul said. In order to get the square size we want, we just have to make sure that we're using some number of share in between maxShareCount and minShareCount

maxShareCount := squareSize * squareSize
minShareCount := maxShareCount / 2 

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks a lot for your help 🙏

I think it is an approximation for the other shares in the square (transactions, intermediate state roots, evidence).

Another question, are ISRs and evidence currently added to blocks?

minShareCount := maxShareCount / 2

Why is that? is it to accommodate ISRs and evidence?

testutil/testnode/full_node_test.go Outdated Show resolved Hide resolved
rootulp
rootulp previously approved these changes Oct 13, 2022
Copy link
Collaborator

@rootulp rootulp left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🚀 you're a beast, great work!

testutil/testnode/full_node_test.go Outdated Show resolved Hide resolved
testutil/testnode/full_node_test.go Show resolved Hide resolved

// FillBlock creates and submits a single transaction that is large enough to
// create a square of the desired size. broadcast mode indicates if the tx
// should be submitted async, sync, or block. (see flags.BroadcastModeSync). If
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

[question] should this be flags.BroadcastMode per https://github.com/cosmos/cosmos-sdk/blob/7781cdb3d20bc7ebac017452897ce1e6ab3903ef/client/flags/flags.go#L114

Suggested change
// should be submitted async, sync, or block. (see flags.BroadcastModeSync). If
// should be submitted async, sync, or block (see flags.BroadcastMode). If

Note: block mode was removed in cosmos/cosmos-sdk#12659. Since celestia-app currently uses cosmos-sdk v0.46.0 we can still use block here but note to future selves that we'll need to remove it when we upgrade cosmos-sdk to a release with that change

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah, blocking is just convenient atm and we will have to fix this in the future across the board.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

}
maxShareCount := squareSize * squareSize
// we use a formula to guarantee that the tx is the exact size needed to force a specific square size.
msgSize := (maxShareCount - (2 * squareSize)) * appconsts.SparseShareContentSize
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Assume squareSize=16 so maxShareCount=256:

msgSize := (maxShareCount - (2 * squareSize)) * appconsts.SparseShareContentSize
msgSize := (256 - (2 * 16)) * appconsts.SparseShareContentSize
msgSize := (256 - 32) * appconsts.SparseShareContentSize
msgSize := (224) * appconsts.SparseShareContentSize

I don't think the transaction is guaranteed to take (2 * squareSize) but I think it is an approximation for the other shares in the square (transactions, intermediate state roots, evidence).

func (s *IntegrationTestSuite) Test_FillBlock() {
require := s.Require()

for squareSize := 16; squareSize < appconsts.MaxSquareSize; squareSize *= 2 {
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

[question] out of curiosity, why start squareSize at 16? My guess is that the estimation logic for non message-shares may break for squareSizes < 16.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

good catch, that was a remnant of the old implementation.

57744bf

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

also #867

Co-authored-by: CHAMI Rachid <chamirachid1@gmail.com>
@evan-forbes
Copy link
Member Author

we should cut a release for v0.7.0 after merging this imo

@evan-forbes evan-forbes merged commit c9dce39 into main Oct 13, 2022
@evan-forbes evan-forbes deleted the evan/implement-fillnode branch October 13, 2022 09:38
rach-id added a commit to rach-id/celestia-app that referenced this pull request Nov 16, 2022
This PR 
- adds a modified FillSquare method for the testnode that only submits a
single large message
- modifies the default configs for a testnode's mempool to accept large
transactions
- modifies the `PostData` method to accept a flag that determines if the
process blocks

The issue with the old approach was that it was submitting too many
transactions for a block with such short block times. The testnode was
unlikely to get all of the transactions included in a single block,
particularly when waiting to submit each transaction until the previous
one was accepted into the mempool.

Co-authored-by: CHAMI Rachid <chamirachid1@gmail.com>
Co-authored-by: Rootul P <rootulp@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
testing items that are strictly related to adding or extending test coverage
Projects
No open projects
Status: Done
Development

Successfully merging this pull request may close these issues.

4 participants