visualize app pie chart other bucket should apply correct filter on other bucket

[liza-mae]

└- ✖ fail: "visualize app pie chart other bucket should apply correct filter on other bucket" 23:58:17 │ pie chart 23:58:17 │ other bucket 23:58:17 │ should apply correct filter on other bucket: 23:58:17 │ 23:58:17 │ Error: expected [ 'win 8', 'win xp', 'win 7', 'ios', 'Missing', 'Other' ] to sort of equal [ 'Missing', 'osx' ] 23:58:17 │ + expected - actual 23:58:17 │ 23:58:17 │ [ 23:58:17 │ - "win 8" 23:58:17 │ - "win xp" 23:58:17 │ - "win 7" 23:58:17 │ - "ios" 23:58:17 │ "Missing" 23:58:17 │ - "Other" 23:58:17 │ + "osx" 23:58:17 │ ] 23:58:17 │
23:58:17 │ at Assertion.assert (node_modules/expect.js/index.js:96:13) 23:58:17 │ at Assertion.eql (node_modules/expect.js/index.js:230:10) 23:58:17 │ at Context.it (test/functional/apps/visualize/_pie_chart.js:125:28) 23:58:17 │ at 23:58:17 │ at process._tickCallback (internal/process/next_tick.js:188:7) 23:58:17 │

Failing on cloud 6.5.1

[elasticmachine]

Pinging @elastic/kibana-app

[LeeDr]

Screenshot from recent failure;

And debug log;

[00:03:01]                   └-> "before each" hook: global before each
[00:03:01]                   │ debg TestSubjects.find(pieSlice-Other)
[00:03:01]                   │ debg findByCssSelector [data-test-subj~="pieSlice-Other"]
[00:03:01]                   │ debg isGlobalLoadingIndicatorVisible
[00:03:01]                   │ debg TestSubjects.exists(globalLoadingIndicator)
[00:03:01]                   │ debg existsByDisplayedByCssSelector [data-test-subj~="globalLoadingIndicator"]
[00:03:01]                   │ debg awaitGlobalLoadingIndicatorHidden
[00:03:01]                   │ debg TestSubjects.find(globalLoadingIndicator-hidden)
[00:03:01]                   │ debg findByCssSelector [data-test-subj~="globalLoadingIndicator-hidden"]
[00:03:01]                   │ debg in findAllByCssSelector: path.slice
[00:03:01]                   │ debg pieData.length = 6
[00:03:01]                   │ info Taking screenshot "/var/lib/jenkins/workspace/elastic+kibana+6.x/JOB/kibana-ciGroup11/node/immutable/kibana/test/functional/screenshots/failure/visualize app  pie chart other bucket should apply correct filter on other bucket.png"
[00:03:02]                   │ info Current URL is: http://localhost:5620/app/kibana#/visualize/create?type=pie&indexPattern=logstash-*&_g=(refreshInterval:(pause:!t,value:0),time:(from:%272015-09-19T06:31:44.000Z%27,mode:absolute,to:%272015-09-23T18:31:44.000Z%27))&_a=(filters:!((%27$state%27:(store:appState),meta:(alias:!n,disabled:!f,index:%27logstash-*%27,key:machine.os.raw,negate:!t,params:!(%27win%208%27,%27win%20xp%27,%27win%207%27,ios),type:phrases,value:%27win%208,%20win%20xp,%20win%207,%20ios%27),query:(bool:(minimum_should_match:1,should:!((match_phrase:(machine.os.raw:%27win%208%27)),(match_phrase:(machine.os.raw:%27win%20xp%27)),(match_phrase:(machine.os.raw:%27win%207%27)),(match_phrase:(machine.os.raw:ios))))))),linked:!f,query:(language:lucene,query:%27%27),uiState:(),vis:(aggs:!((enabled:!t,id:%271%27,params:(),schema:metric,type:count),(enabled:!t,id:%272%27,params:(field:machine.os.raw,missingBucket:!t,missingBucketLabel:Missing,order:desc,orderBy:%271%27,otherBucket:!t,otherBucketLabel:Other,size:5),schema:segment,type:terms)),params:(addLegend:!t,addTooltip:!t,isDonut:!t,labels:(last_level:!t,show:!f,truncate:100,values:!t),legendPosition:right,type:pie),title:%27New%20Visualization%27,type:pie))
[00:03:02]                   │ info Browser output is: 
[00:03:02]                   │ info Saving page source to: /var/lib/jenkins/workspace/elastic+kibana+6.x/JOB/kibana-ciGroup11/node/immutable/kibana/test/functional/failure_debug/html/visualize app  pie chart other bucket should apply correct filter on other bucket.html
[00:03:02]                   └- ✖ fail: "visualize app  pie chart other bucket should apply correct filter on other bucket"
[00:03:02]                   │ 
[00:03:02]                   │       Error: expected [ 'win 8', 'win xp', 'win 7', 'ios', 'Missing', 'Other' ] to sort of equal [ 'Missing', 'osx' ]
[00:03:02]                   │       + expected - actual
[00:03:02]                   │ 
[00:03:02]                   │        [
[00:03:02]                   │       -  "win 8"
[00:03:02]                   │       -  "win xp"
[00:03:02]                   │       -  "win 7"
[00:03:02]                   │       -  "ios"
[00:03:02]                   │          "Missing"
[00:03:02]                   │       -  "Other"
[00:03:02]                   │       +  "osx"
[00:03:02]                   │        ]
[00:03:02]                   │       
[00:03:02]                   │       at Assertion.assert (node_modules/expect.js/index.js:96:13)
[00:03:02]                   │       at Assertion.eql (node_modules/expect.js/index.js:230:10)
[00:03:02]                   │       at Context.it (test/functional/apps/visualize/_pie_chart.js:125:28)
[00:03:02]                   │       at <anonymous>
[00:03:02]                   │       at process._tickCallback (internal/process/next_tick.js:189:7)
[00:03:02]                   │ 
[00:03:02]                   │ 

I haven't dug in enough to figure out how this test was supposed to work. The previous test verifies this data; const expectedTableData = [ 'win 8', 'win xp', 'win 7', 'ios', 'Missing', 'Other' ];

and then this test does; PageObjects.visualize.filterPieSlice('Other');

and expects; const expectedTableData = [ 'Missing', 'osx' ];

How does filtering out other also remove those other items? I'll run it locally.

[ppisljar]

clicking on other bucket should create a filter like "not windows8 and not win7 and not ios and not winxp"

[lukeelmers]

I'm looking into #26753 which will likely possibly also resolve relate to this issue.

Edit: turns out it's not related to the issue I was working on after all :(

[stacey-gammon]

Whoa this one is really flaky!

Here's the html output from a recent failure on master:

Kibana.htm.zip

and the screenshot which looks a bit different from Lee's above:

[spalger]

Skipped

master: b03e63b5de 6.x: 010a998bac

[liza-mae]

No longer skipped in master with #28509 merged - next test runs should tell and waiting for back ports to be merged for 6.x and 6.6 will be skipped.

[dmlemeshko]

It should be fixed with #28936 as now we have a more reliable way to handle when visualisation loading is completed. Let's keep an eye on related Kibana Failing Tests for the next couple of weeks.

[liza-mae]

Thanks @dmlemeshko