Resolve hash error for FTM-1511-03209 artifact

[emikelsons]

This is addressing #4956

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 84.87%. Comparing base (7f0c3c7) to head (2b745b6).
Report is 11 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #4957   +/-   ##
=======================================
  Coverage   84.87%   84.87%           
=======================================
  Files         697      697           
  Lines       93932    93932           
=======================================
  Hits        79724    79724           
  Misses      14208    14208           

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[benlorenz]

Thanks!
I have created #4959 to allow running the extra-long tests for PRs with the corresponding label. If we merge that one first and close+re-open this PR it should allow us to verify this and check the tests.

[benlorenz]

The extra-long tests were successful but seem to be taking a bit longer than before. Previous runs on master were 60-70 minutes, this time it took 85 min. Not sure if this is expected with the updated date files.

[emikelsons]

I saw that the downloading of the FTM artifact failed once before it eventually worked

[benlorenz]

I saw that the downloading of the FTM artifact failed once before it eventually worked

That is expected, julia first tries the pkgservers which mirror all artifacts that are referred to by released versions of packages. And if that fails it will try the original url.
So once there is an Oscar release with this artifact (with the correct hash) the download should work on the first attempt.

[fingolfin]

Let's merge this. If it the performance regression is permanent we'll revisit this.