Murko: continue thaw if redis is down

[jacob720]

Fixes #810

Requires DiamondLightSource/dodal#1750

Instructions to reviewer on how to test:

1. Check tests pass and thaw happens when redis connection fails.

Checks for reviewer

• Would the PR title make sense to a user on a set of release notes

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.30%. Comparing base (c1346e4) to head (f75054e).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1437      +/-   ##
==========================================
+ Coverage   92.28%   92.30%   +0.01%     
==========================================
  Files         142      142              
  Lines        8092     8107      +15     
==========================================
+ Hits         7468     7483      +15     
  Misses        624      624              

Components	Coverage Δ
i24 SSX	78.56% <ø> (ø)
hyperion	97.93% <ø> (ø)
other	98.09% <100.00%> (+<0.01%)	⬆️

🚀 New features to boost your workflow:

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

[DominicOram]

Thanks. This works but I think it might be cleaner if it was handled inside the MurkoCallback and the MurkoResultsDevice. They can both just do nothing and warn, then the MurkoResultsDevice can keep the positions all at 0 and the plan would never have to know there's a redis issue

[DominicOram]

Could: Would be better to put this in a parametrize

[jacob720]

I ended up deleting this test

[DominicOram]

I should add my reasoning here. Basically, what if we wanted to make use of the MurkoCallback again? We would have to put a similar check everywhere we wanted to use it, whereas if we make the callback itself handle it then we don't have to worry about it. Also, we would have less places we're trying to make redis connections, which is nice

[jacob720]

That makes sense, I'll refactor

[DominicOram]

Thanks, couple of minor things, take them or leave them

[DominicOram]

Could: It might be better to re-check the connection on start. For other callbacks we only create them once when we start up and never recreate them. I know we currently don't for MurkoCallback but if we did want to this would mean that we would have to restart the whole env to pick up redis coming back up

[DominicOram]

Could: I don't think it's that clear with this patch that you're definitely testing the right outcomes. Instead I would do something like:

        omega_put = get_mock_put(smargon.omega.user_setpoint)

        assert omega_put.call_args_list == [
            call(360.0, wait=True),
            call(0.0, wait=True),
        ]

        omega_put.reset_mock()

and remove the patch of _rotate_in_one_direction_and_stream_to_redis all together