Wip/mango 1945/fix failing tests

[kishorevenki]

• This fix addresses the issues mentioned in https://radixiot.atlassian.net/browse/MANGO-1945 .
• Two tests from TrendLogObjectTest and one test from TrendLogObjectMultipleTest were failing.
• The intent of these tests are to verify that data are collected at periodic interval as per Log-Interval property.
• While TrendObjects are collecting the data for every minute, the test incorrectly change the log interval to one hour. Collection of records and hence the record-count cannot be accurate if the log interval property is changed while TrendLog objects are collecting the live data. Further such step is not the standard test requirement.
• Tests are fixed by removing the steps of chage of Log-Interval property while collecting the data.

[sonarqubecloud]

Quality Gate passed

Issues
11 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[terrypacker]

I checked out the code and dug around a little. Instead of removing some of the coverage can we just move those 1hr test cases into separate tests? I get a little nervous when I see the failing code of a test just deleted. I think I understand why it is failing but since we are using a WarpClock there is probably a few ways to make this work without removing the code.

From what I can tell this is really the code you removed that you are concerned with:

        // Advance the clock to the new polling time.
        final int minutes = (62 - clock.get(ChronoField.MINUTE_OF_HOUR)) % 60;
        clock.plus(minutes, MINUTES, 0);

I'm thinking if you just set the time to a nice round number at the start of the test or account for why the above could be off by 1m2s ever.

However I would be ok with just having a separate tests for the 1hr polling interval.

[kishorevenki]

Adding a separate test for one hour polling interval would be better.