Failed to query access key for v0.6.1

[mrLSD]

In github actions have got errors:

Failed to query access key: handler error: [Access key for public key ed25519:5WMgq6gKZbAr7xBZmXJHjnj4C3UZkNJ4F5odisUBFcRh has never been observed on node the

When I tried re-run, it appeared for another test. Some random error.

Example output:
https://github.com/aurora-is-near/aurora-eth-connector/actions/runs/3653185111/jobs/6172860110

[ChaoticTempest]

Is there a specific test on Aurora connector I can take a look at? This error usually comes from the fact that sandbox has no record of the public key being associated with the account. Some common ways this happens is if you do set_secret_key but don't attach the public key to the account via something like account.batch call

[mrLSD]

It's absolutely random. Sometimes all tests pass, and sometimes random tests fail.

There is code for attaching the pub key for Aurora eth-connector tests util:
https://github.com/aurora-is-near/aurora-eth-connector/blob/master/eth-connector-tests/src/utils.rs#L101

[ChaoticTempest]

This is super weird: I've ran those tests (both mainnet and testnet profiles) locally on my M1 macbook a few times and they've always passed. Are you sure this isn't an issue with something on the instance the tests are being ran on? Is this passing all the time locally for you guys? I did have to change the path of .deploy(include_bytes!("../../bin/aurora-eth-connector-test.wasm")) to include ...testnet-test.wasm to get the testnet profile to work though. Maybe there are subtle environment issues you guys aren't aware of.

Also consider upping workspaces to 0.7.0. There's error message improvements which might give more details about this issue.

[mrLSD]

I never received that error locally. All tests passed successfully. Also M1 chip.

Also sometimes it passed with Github CI:
https://github.com/aurora-is-near/aurora-eth-connector/actions/runs/3624790698

But sometimes it didn't. And it's related to Github CI.

Specially extended errors for that. The issue was discovered here: https://github.com/aurora-is-near/aurora-eth-connector/blob/feat/use-near-standard-nep141-implementation/eth-connector-tests/src/utils.rs#L87

And again, it's related to Github CI. Locally it always works fine.

Unfortunately, we have approved PR: Near-One/aurora-eth-connector#2
and can't merge it because of workspace.rs issues.

That's actions CI: https://github.com/aurora-is-near/aurora-eth-connector/actions

Multiple times re-run CI, multiple times different, and randomized errors. Pretty strange.

But without Github CI pass we can't merge critical PR.

[mrLSD]

Sorry, also answered here to #255. Pretty similar problem - random errors related to workspace.rs

[mrLSD]

Maybe there are subtle environment issues you guys aren't aware of.

@ChaoticTempest But what is exactly? Do not have even ideas.