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

protocols/kad/query/peers/closest: Make at_capacity use max #1548

Merged
merged 5 commits into from
Apr 16, 2020

Conversation

mxinden
Copy link
Member

@mxinden mxinden commented Apr 14, 2020

When not making progress for parallelism time a ClosestPeersIter
becomes State::Stalled. When stalled an iterator is allowed to make
more parallel requests up to num_results. If num_results is smaller
than parallelism make sure to still allow up to parallelism requests
in-flight.

When not making progress for `parallelism` time a `ClosestPeersIter`
becomes `State::Stalled`. When stalled an iterator is allowed to make
more parallel requests up to `num_results`. If `num_results` is smaller
than `parallelism` make sure to still allow up to `parallelism` requests
in-flight.
@mxinden mxinden requested a review from romanb April 14, 2020 19:12
assert!(!iter.at_capacity());
}

/// When not making progress for `parallelism` time a [`ClosestPeersIter`] becomes
Copy link
Member Author

Choose a reason for hiding this comment

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

This test is likely a bit over the top. I left it here to showcase the issue in a more in-depth scenario. I would suggest removing it before merging. What do you think?

Copy link
Contributor

Choose a reason for hiding this comment

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

Yes, I don't think it adds much to the other test, especially if the other is turned into a property test. Thanks for writing it though.

Copy link
Contributor

@romanb romanb left a comment

Choose a reason for hiding this comment

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

An interesting edge-case, though unlikely to be hit in practice as num_results for the iterator is taken from the configured replication factor (K_VALUE by default) and I doubt someone configures it to be < ALPHA = 3. Nevertheless, this is a bug (at least after #1536) 👍.

@@ -691,4 +693,78 @@ mod tests {

QuickCheck::new().tests(10).quickcheck(prop as fn(_) -> _)
}

#[test]
fn stalled_iter_at_capacity_use_max_of_parallelism_and_num_results() {
Copy link
Contributor

Choose a reason for hiding this comment

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

Wouldn't this make for a good property test? The property being the following: "For any ClosestPeersIter, if it is in State::Stalled then it must be at capacity only if num_waiting = max{num_results, parallelism}" ?

assert!(!iter.at_capacity());
}

/// When not making progress for `parallelism` time a [`ClosestPeersIter`] becomes
Copy link
Contributor

Choose a reason for hiding this comment

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

Yes, I don't think it adds much to the other test, especially if the other is turned into a property test. Thanks for writing it though.

Remove in-depth test and make smaller test property based.
@mxinden
Copy link
Member Author

mxinden commented Apr 16, 2020

Property based test is a good idea. Would you mind taking another look @romanb?

protocols/kad/src/query/peers/closest.rs Outdated Show resolved Hide resolved
Co-Authored-By: Roman Borschel <romanb@users.noreply.github.com>
@mxinden mxinden merged commit 77a34c0 into libp2p:master Apr 16, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants