CI Tests failing due to foundry toolchain possibly using an older foundry build

[0xTimepunk]

Hi,

During our test suite with foundry-toolchain/github actions the following error appears:

Test result: ok. 5 passed; 0 failed; finished in 1.08s
[269](https://github.com/superform-xyz/super-vaults/actions/runs/4113946632/jobs/7100701243#step:6:270)
2023-02-07T12:37:21.274109Z ERROR sharedbackend: Failed to send/recv `basic` err=GetAccount(0x7fa9385be102ac3eac297483dd6233d62b3e1496, 
[270](https://github.com/superform-xyz/super-vaults/actions/runs/4113946632/jobs/7100701243#step:6:271)
(code: -32000, message: cannot query unfinalized data, data: None)) address=0x7fa9385be102ac3eac297483dd6233d62b3e1496
[271](https://github.com/superform-xyz/super-vaults/actions/runs/4113946632/jobs/7100701243#step:6:272)

[272](https://github.com/superform-xyz/super-vaults/actions/runs/4113946632/jobs/7100701243#step:6:273)
Running 1 test for src/aave-v3/test/AaveV3ERC4626Reinvest.t.sol:AaveV3ERC4626ReinvestTest
[273](https://github.com/superform-xyz/super-vaults/actions/runs/4113946632/jobs/7100701243#step:6:274)
[FAIL. Reason: Failed to get account for 0x7fa9385be102ac3eac297483dd6233d62b3e1496: 0x7fa9385be102ac3eac297483dd6233d62b3e1496]

which appears to be related to foundry-rs/foundry#3874 but was fixed apparently in a nightly build at the end of December as commented here: foundry-rs/foundry#3874 (comment)

It seems to us that the toolchain is possibly not using the latest version with this fix?

Our CI is supposedly using foundry "nightly" build:

Thank you

[PaulRBerg]

This might be related to the recently added cache - have you tried manually deleting the cache by following the instructions in the README?

[0xTimepunk]

I went to the mentioned caches page and couldn't find any cache.

I added a cache: false to the CI file. Also the issue seems random, because when you try sometimes re-running the CI job it passes without errors.

[onbjerg]

This is likely because the RPC you are talking to either times out, or does not have the data available. Remember that non-archive nodes only support accessing state for the n latest blocks.

[0xTimepunk]

This is likely because the RPC you are talking to either times out, or does not have the data available. Remember that non-archive nodes only support accessing state for the n latest blocks.

We are using quicknodes for all the chains we need the service and all of our nodes are archive nodes. So unless they are timing out, which I would find odd, I would rule out that.

Is it possible that what timeouts is because of foundry itself, expecting to get return data faster than what is possible?. Doesn't seem related foundry-toolchain actually.