If tests are failing w/ 'Internal Salesforce Error' in Summer '18, what will happen to deployed version when Production upgrades? [closed]

Question

I've been building out queueable processes in our Sandbox on Summer '18, and I was able to deploy some of the functionality last week to Production (Spring '18), and now those tests in Sandbox are failing with 'Internal Salesforce Error'.

I have two immediate concerns -

1. Is it likely, then, that the tests (and possibly the processes themselves) will start failing when Production is upgraded, and
2. If the sandbox was upgraded ~ a week ago, and I was able to run tests last week in the upgraded sandbox, what's going on?

I don't see any issues on Trust at the moment for my Sandbox instance; not sure what my next steps ought to be to move forward on these projects, since the similar tests for phase two are also failing w/ the same(?) internal errors.

Answer 1

Before each release, Salesforce runs the Apex Hammer Test, which basically means it runs all of your tests in the current release and the next release, and logs any differences in behaviors (e.g. failures, Internal Server Errors, etc). If there is a difference, these differences are given high priority to be patched before the next release hits.

So, what likely happened is that the AHT ran against your org, the error was logged, a bug was generated, and it was patched. This all happens invisibly behind the scenes so customers don't have to worry about it. As far as I can tell, no customer anywhere ever has had a test that passed in one release and fail in the next purely because of regression; Salesforce is really good at catching those errors.

However, if you're really concerned, you can contact support. The most likely course of action is that they'll forward your case to R&D (Research & Development) to get it fixed, or advise you on how you can fix your code so it won't be impacted in Summer '18.