Conversation
side2k
force-pushed
the
rust-toolchain-1.93.1
branch
from
6eac845 to
015828f
Compare
Deploying kolme with
|
| Latest commit: |
55c21e2
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://001c7e96.kolme.pages.dev |
| Branch Preview URL: | https://rust-toolchain-1-93-1.kolme.pages.dev |
- integer types have is_multiple_of() method since Rust 1.87: https://doc.rust-lang.org/stable/std/primitive.usize.html#method.is_multiple_of - clippy required replacing Error::new(ErrorKind::Other) with [Error::other](https://docs.rs/borsh/latest/borsh/io/struct.Error.html#method.other) - replace std::mem::replace(...) with [Option::replace()](https://doc.rust-lang.org/stable/core/option/enum.Option.html#method.replace)
side2k
force-pushed
the
rust-toolchain-1.93.1
branch
from
5cf23ad to
b14d608
Compare
| let kolme = Kolme::new(app, code_version, store).await?; | ||
| let kolme = Kolme::new(app, code_version, store) | ||
| .await? | ||
| .set_tx_await_duration(Duration::from_secs(15)); |
There was a problem hiding this comment.
This also looks suspicious: What was the earlier duration and why is it being changed now ?
There was a problem hiding this comment.
Sorry, I missed the description of the PR.
increase tx timeout in postgres tests, 10s was not enough - I'm not sure why, but its seems to be something related to either Rust or dependencies update
Can you look into the root cause of the change? If it's just the GitHub runners, that's fine. Otherwise, we need to know what it is so we can decide how to proceed.
There was a problem hiding this comment.
I did look(with a limited timeframe though), and wasn't able to find the culprit. Locally it works, on CI it fails with a timeout.
There was a problem hiding this comment.
Okay, some questions:
- Is the issue currently reproducible on the main branch with 1.86 rust version ?
- What happens in the CI if you skip the dependency upgrade ?
There was a problem hiding this comment.
Is the issue currently reproducible on the main branch with 1.86 rust version ?
As far as I can tell from action logs on other branches - it is not.
What happens in the CI if you skip the dependency upgrade ?
Will check that and update you.
There was a problem hiding this comment.
Reverted:
- timeout increase
- dependency bump
- and also reverting relevant code changes(replacing Keypair::from_bytes() with Keypair::try_from() due to deprecation)
(see commit list for more details)
tests succeded two times in a row.
I guess we can stop here?
There was a problem hiding this comment.
I guess we can stop here?
Agreed, I'm happy we don't have to increase the timeout to make the tests pass.
Looks like one of the dependency is causing the issue, you can address that in a future PR if you have the time. :-)
There was a problem hiding this comment.
Yeah, I'm going to create a ticket in Jira for the backlog, so the knowledge won't be lost
There was a problem hiding this comment.
Here it is: https://fpcomplete.atlassian.net/browse/KOL-85
2 participants
rust-toolchain.tomlis leaking into the containerlocalosmosisimage's wasmd (up to version 31.0.1) is not compatible with bulk memory instructions in contracts, and Rust 1.93.1 seems to not be able to turn that feature off, despite the relevant flagsclippyupdate, some - by the rest of toolchain update (e.g., integer types now have is_multiple_of() method since Rust 1.87)