SDCSRM-388 Fix flaky unit test timing

[AdamHawtin]

Motivation and Context

This particular unit test was intermittently failing in the build pipeline when the result wasn't within the arbitrary time window. I spent some time looking at it but couldn't see any good reason this test would run slower than any others, so I'm concluding for now that it's not a particular issue with this function or test performance, and is more likely down to performance hiccups in the build environment, and we simply don't test this timing in many (if any) other places to see the issue. For now, I am fixing it by checking the timing differently, simply checking the timestamp is within the time window of the test.

What has changed

• Fix flaky unit test timing

How to test?

Check the change. Nothing to test locally, the issue only seems to present in remote build environments.

Links

https://jira.ons.gov.uk/browse/SDCSRM-388

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 82.30%. Comparing base (16c9bbf) to head (a2b1836).

Additional details and impacted files

@@             Coverage Diff              @@
##               main      #62      +/-   ##
============================================
- Coverage     86.05%   82.30%   -3.75%     
+ Complexity      216      210       -6     
============================================
  Files            45       45              
  Lines           667      667              
  Branches         33       33              
============================================
- Hits            574      549      -25     
- Misses           81      102      +21     
- Partials         12       16       +4     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[dbanks91]

Your fix looks good.

Might be worth just raising a card so in future we can do performance tests in build environments to see why we might get performance hiccups?

[AdamHawtin]

Might be worth just raising a card so in future we can do performance tests in build environments to see why we might get performance hiccups?

@dbanks91 I think it's something for us to keep an eye on, but most of the build jobs have been reliable despite these slow downs, so unless it causes us any proper issues I don't think it's worth any more of our time at the moment.