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

Ensure that L1 Events received from the relayer are in a deterministic order #1772

Closed
MitchTurner opened this issue Mar 22, 2024 · 0 comments · Fixed by #1779
Closed

Ensure that L1 Events received from the relayer are in a deterministic order #1772

MitchTurner opened this issue Mar 22, 2024 · 0 comments · Fixed by #1779
Assignees

Comments

@MitchTurner
Copy link
Member

We are now writing the merkle root of all the events received via the relayer to the header of each block. It is important that the order of those events is the same for all nodes verifying blocks so that the merkle root is the same.

Related to #1769

MitchTurner added a commit that referenced this issue Mar 28, 2024
closes #1772

***

Had to modify `MockDB` to also maintain order.

This code only works if the Relayer DB maintains order. In production it
is stored as a `Mappable` with value being `[Event]`:

```
impl Mappable for EventsHistory {
    /// The key is the height of the DA.
    type Key = Self::OwnedKey;
    type OwnedKey = DaBlockHeight;
    /// The value is an events happened at the height.
    type Value = [Event];
    type OwnedValue = Vec<Event>;
}
```
So it should be good.

This approach is flakier than having the relayer reader in the executor
ensure the order, but it could add a lot of extra execution cost to the
executor code if we put the sorting on that side. We are already making
assumptions about what the writing code is doing though. For example, if
an event is missing from L1, correct order doesn't matter. So why not
add more assumptions :P

***

We require that all `Logs`include the `log_index` so we can sort them.
So I added a test
`relayer__if_a_log_does_not_include_index_then_event_not_included` to
check the sad path. **I'm not positive this is the behavior we want.**
It is erroring in the `sort_events_by_log_index` method but I'm not
exactly sure where that bubbles up to: `start_and_await` still returns
an `Ok(Starting)` not an `Ok(StoppedWithError)`. That might be a bug in
our `Service` code, or might be meant to be used differently. Not sure.
So I just check that the `Log` isn't included in the test...

~I had to modify `MockMiddleware` to only let you take the batch once.
Before it would error and then just try again and call the same value
over and over and never exited. Making some assumptions there too.~

---------

Co-authored-by: Green Baneling <[email protected]>
crypto523 added a commit to crypto523/fuel-core that referenced this issue Oct 7, 2024
closes FuelLabs/fuel-core#1772

***

Had to modify `MockDB` to also maintain order.

This code only works if the Relayer DB maintains order. In production it
is stored as a `Mappable` with value being `[Event]`:

```
impl Mappable for EventsHistory {
    /// The key is the height of the DA.
    type Key = Self::OwnedKey;
    type OwnedKey = DaBlockHeight;
    /// The value is an events happened at the height.
    type Value = [Event];
    type OwnedValue = Vec<Event>;
}
```
So it should be good.

This approach is flakier than having the relayer reader in the executor
ensure the order, but it could add a lot of extra execution cost to the
executor code if we put the sorting on that side. We are already making
assumptions about what the writing code is doing though. For example, if
an event is missing from L1, correct order doesn't matter. So why not
add more assumptions :P

***

We require that all `Logs`include the `log_index` so we can sort them.
So I added a test
`relayer__if_a_log_does_not_include_index_then_event_not_included` to
check the sad path. **I'm not positive this is the behavior we want.**
It is erroring in the `sort_events_by_log_index` method but I'm not
exactly sure where that bubbles up to: `start_and_await` still returns
an `Ok(Starting)` not an `Ok(StoppedWithError)`. That might be a bug in
our `Service` code, or might be meant to be used differently. Not sure.
So I just check that the `Log` isn't included in the test...

~I had to modify `MockMiddleware` to only let you take the batch once.
Before it would error and then just try again and call the same value
over and over and never exited. Making some assumptions there too.~

---------

Co-authored-by: Green Baneling <[email protected]>
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 a pull request may close this issue.

1 participant