Skip to content

CI tests fail in PRs #88

Closed
Closed

Description

@ErikBjare

I haven't looked into it, but it seems to be a recurring issue.

Activity

  1. added
    ciRelating to Continuous Integration
    on Apr 1, 2021
  2. changed the title [-]CI is broken[/-] [+]CI tests are broken[/+] on Apr 1, 2021
  3. ErikBjare commented on May 4, 2021

    @ErikBjare
    MemberAuthor

    In #58 I ran the tests locally (after a small fix), and they passed.

    Not sure why they failed in the PR, perhaps due to the Infura token being kept secret? The PR is merged now anyway, we'll see if the tests pass on master.

    Edit: Looks like it passed 00c77d6, must be a secrets thing.

  4. changed the title [-]CI tests are broken[/-] [+]CI tests fail in PRs[/+] on May 4, 2021
  5. ErikBjare commented on May 5, 2021

    @ErikBjare
    MemberAuthor

    Having issues again with token -> token swaps in test_make_trade:

    FAILED tests/test_uniswap.py::TestUniswap::test_make_trade[1-0x6B175474E89094C44Da98b954EedeAC495271d0F-0x0D8775F648430679A709E98d2b0Cb6250d2887EF-1000000000-None-does_not_raise]
    

    The error:

    Runtime Error: invalid JUMP at 72e826e5ec1e970f37af4b41635535232c58bcb7795871b2738347fd8c1314ca/2a1530c4c41db0b0b2bb646cb5eb1a67b7158667:39
    

    I thought I fixed that in #58, but apparently not for good.

  6. ErikBjare commented on May 5, 2021

    @ErikBjare
    MemberAuthor

    Ah damn, just realized that's for v1... Explains a lot.

  7. ErikBjare commented on May 5, 2021

    @ErikBjare
    MemberAuthor

    Wasted way too much time on this today. Considering just dropping support for v1 as it's pretty dead anyway, or at least commenting out the specific testcase for v1.

    What do you think @shanefontaine?

  8. shanefontaine commented on May 5, 2021

    @shanefontaine
    Collaborator

    I am for dropping V1 support. Looks like V1 only has 330 transactions in the last 24 hours and decreasing daily volume -- not enough to warrant ongoing support, IMO.

  9. ErikBjare commented on May 23, 2021

    @ErikBjare
    MemberAuthor

    I fixed the remaining failing tests for v1 in f782df5. Might keep support around for another version or so.

    The issue in #88 (comment) turns out to have been caused by there not being 1000 ETH in the pool from which the InsufficientFunds test tried to buy 1000 ETH with DAI 🤦‍♂️

  10. ErikBjare commented on Dec 21, 2021

    @ErikBjare
    MemberAuthor

    Things seem to work fine now, closing.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    ciRelating to Continuous Integration

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions