Mpool fixes #2847
Mpool fixes #2847
Conversation
elmattic
requested review from
lemmih,
LesnyRumcajs,
jdjaustin,
hanabi1224,
creativcoder,
sudo-shashank and
aatifsyed
as code owners
| @@ -70,10 +73,17 @@ impl MsgSet { | |||
|
|
|||
| /// Add a signed message to the `MsgSet`. Increase `next_sequence` if the | |||
| /// message has a sequence greater than any existing message sequence. | |||
| pub fn add(&mut self, m: SignedMessage) -> Result<(), Error> { | |||
| pub fn add(&mut self, m: SignedMessage, untrusted: bool) -> Result<(), Error> { | |||
There was a problem hiding this comment.
What does untrusted mean? And it looks like you always set it to false.
There was a problem hiding this comment.
I did this for future endpoint (if we implement it). Right now we support only Filecoin.MpoolPushMessage.
But it exists a counterpart untrusted one: Filecoin.MpoolPushUntrusted for public gateways.
See lotus PR: filecoin-project/lotus#3915
Co-authored-by: David Himmelstrup <david.himmelstrup@chainsafe.io>
| // sleep allows for async block to update mpool's cur_tipset | ||
| tokio::time::sleep(Duration::new(2, 0)).await; |
There was a problem hiding this comment.
I highly dislike sleeps in the tests.
- It makes them flaky, yielding failures when run on unexpectedly low-end (or busy) workers.
- It makes tests unnecessarily long. Say the update took 5ms - we are waiting for 2s for nothing.
Can you think of a mechanism that would at least partially mitigate those issues?
There was a problem hiding this comment.
Yes. I don't like it as well and I'm looking for a way to remove those.
I've put 10ms while waiting for something better.
There was a problem hiding this comment.
I've added a new method in Provider so it's possible for TestApi to use await instead of sleep.
Co-authored-by: Hubert <hubert@chainsafe.io>
|
Closing it because of merged in #2740. |
Summary of changes
Changes introduced in this pull request:
Reference issue to close (if applicable)
Closes #2761
Other information and links
Change checklist