Add turn_off_exodus_output option to the PetscJacobianTester

[lindsayad]

This option defaults to True. This option helps eliminate race conditions
that can occur when a test developer writes both PetscJacobianTester
and Exodiff tests.

Closes #12975

[moosebuild]

Job Documentation on 152051c wanted to post the following:

View the site here

This comment will be updated on new commits.

[moosebuild]

Job Pthreads Test on 152051c : invalidated by @lindsayad

[tophmatthews]

This certainly works, but I wonder if there is a better way. For example, it's not just exodus files that can cause the writing error correct? Can't csv's do it as well? Also, this won't suppress output if there is a named output block

[Outputs]
  [./someout]
    type = Exodus
  [../]
[]

One way to suppress all outputs by doing

[Outputs]
  active = ''
  [./someout]
    type = Exodus
  [../]
[]

But then if an object is has outputs = someout, then the input file will fail with The output object 'someout' is not a defined output object.

The best way may be to do exodus = false, csv = false, and add enable = false to all named outputs. That way you don't get any errors

[Outputs]
  exodus = false
  csv = false
  [./someout]
    type = Exodus
    enable = false
  [../]
[]

Not sure if that is possible though.

[YaqiWang]

Not sure if possible, but looks to me the simplest is to remove [Outputs] block completely.

[tophmatthews]

Not sure if possible, but looks to me the simplest is to remove [Outputs] block completely.

That will still error with The output object 'someout' is not a defined output object

[YaqiWang]

I mean not using active but removing the Outputs block or turning it off entirely.

[permcody]

Nice - probably should have done this a long time ago.

[permcody]

Don't know how I glazed over the comments on this ticket, but I did. I think Topher and Yaqi's comments are valid, but we don't have a way to do this in MOOSE just yet. I think the proper way to do this would be to write new code in MOOSE that allows us to turn of Output globally, but there might be some work involved there. I believe this PR is an incremental improvement but we still have a ways to go.

[YaqiWang]

@permcody Thanks for the explanation. It does make me feel not ignored ;-)