Fix possible spec failure due to array ordering

[carbonin]

Introduced in bf66236

Example failure: https://travis-ci.org/ManageIQ/manageiq/jobs/431577113

  1) EvmApplication.queue_status groups zone together
     Failure/Error:
       expect(described_class.queue_status).to eq(
         [
           {
             "Zone"   => zone1.name,
             "Queue"  => "generic",
             "Role"   => nil,
             "method" => "X.x",
             "oldest" => tgt_time.strftime("%Y-%m-%d"),
             "count"  => 1,
           },
       expected: [{"Zone"=>"Zone 9", "Queue"=>"generic", "Role"=>nil, "method"=>"X.x", "oldest"=>"2018-09-21", "count"..."=>"Zone 10", "Queue"=>"generic", "Role"=>nil, "method"=>"X.x", "oldest"=>"2018-09-21", "count"=>1}]
            got: [{"Zone"=>"Zone 10", "Queue"=>"generic", "Role"=>nil, "method"=>"X.x", "oldest"=>"2018-09-21", "count...e"=>"Zone 9", "Queue"=>"generic", "Role"=>nil, "method"=>"X.x", "oldest"=>"2018-09-21", "count"=>1}]
       (compared using ==)
       Diff:
       @@ -1,10 +1,10 @@
       -[{"Zone"=>"Zone 9",
       +[{"Zone"=>"Zone 10",
          "Queue"=>"generic",
          "Role"=>nil,
          "method"=>"X.x",
          "oldest"=>"2018-09-21",
          "count"=>1},
       - {"Zone"=>"Zone 10",
       + {"Zone"=>"Zone 9",
          "Queue"=>"generic",
          "Role"=>nil,
          "method"=>"X.x",
     # ./spec/lib/tasks/evm_application_spec.rb:233:in `block (3 levels) in <top (required)>'

Backport status will be the same as #17987

[bdunne]

👍 LGTM

Although, it was introduced by the way strings with numbers are sorted and bf66236 just exposed that...

[carbonin]

Although, it was introduced by the way strings with numbers are sorted

Ah. I see. Because "Zone 10" is sorted before "Zone 9" ...

Is the array from the method supposed to be sorted? Should that be part of the test? I'm sure we could come up with another way to fix spec so that we continue to test that aspect of the method as well ...

[bdunne]

Is the array from the method supposed to be sorted?

I'm not sure, maybe @kbrock knows?

[miq-bot]

Checked commits carbonin/manageiq@c9aba72~...e78ff10 with ruby 2.3.3, rubocop 0.52.1, haml-lint 0.20.0, and yamllint 1.10.0
1 file checked, 0 offenses detected
Everything looks fine. 🍪

[carbonin]

Okay, pushed another commit for ordering that actually makes the spec closer to the original by creating the zones with the names they had before @bdunne 's change. That should make the ordering consistent.

[bdunne]

Merged for now since it's the same as it used to be (and removes the random test failures). It would still be good to know if it's supposed to be sorted or if the string sorting is good enough and we can switch to match_array in the future.

[carbonin]

Oh, yeah, forgot to actually link the line that does the sorting. It's definitely supposed to be sorted by zone name.

[kbrock]

ugh. late to the game - this looks great

[simaishi]

@carbonin We're backporting #17987 to Gaprindashvili. Please add gaprindashvili/yes if this can be backported to Gaprindashvili as is.

[simaishi]

Turned out this test doesn't exist in Gaprindashvili, removing gaprindashvilil/yes