UI Flake - Wait timed out

[jwforres]

**************************************************
*                    Failures                    *
**************************************************

1)  authenticated e2e-user new project when creating a new project should validate taken name when trying to create
  - Error: Wait timed out after 3015ms
      at /data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1425:29
    at /data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/goog/base.js:1582:15
    at [object Object].webdriver.promise.ControlFlow.runInNewFrame_ (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1654:20)
    at notify (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:465:12)
    at [object Object].then (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:522:7)
    at Object.webdriver.promise.asap (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:767:11)
    at newFrame.onComplete (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1670:25)
    at [object Object].webdriver.promise.Frame_.notify_ (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1914:5)
    at [object Object]._onTimeout (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1901:11)
==== async task ====
    at [object Object].webdriver.promise.ControlFlow.wait (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1410:15)
    at [object Object].webdriver.WebDriver.wait (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/webdriver.js:646:21)
    at login (/data/src/github.com/openshift/origin/assets/test/e2e/test.js:38:12)
    at [object Object].<anonymous> (/data/src/github.com/openshift/origin/assets/test/e2e/test.js:209:7)
    at /data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/jasminewd/index.js:94:14
    at [object Object].webdriver.promise.ControlFlow.runInNewFrame_ (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1654:20)
    at [object Object].webdriver.promise.ControlFlow.runEventLoop_ (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1518:8)
    at [object Object].wrapper [as _onTimeout] (timers.js:261:14)
==== async task ====
Asynchronous test function: beforeEach()
    at [object Object].<anonymous> (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/jasminewd/index.js:93:33)
    at [object Object].<anonymous> (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/minijasminenode/lib/async-callback.js:45:37)
Error
    at [object Object].<anonymous> (/data/src/github.com/openshift/origin/assets/test/e2e/test.js:207:5)
    at [object Object].<anonymous> (/data/src/github.com/openshift/origin/assets/test/e2e/test.js:206:3)
    at Object.<anonymous> (/data/src/github.com/openshift/origin/assets/test/e2e/test.js:3:1)

[jwforres]

Next time someone sees this flake, please see if there are any screenshots available in the artifacts on the jenkins job.

[jwforres]

Example with a screenshot #6120 (comment)

[spadgett]

See also #6407 (comment)

[spadgett]

Also https://ci.openshift.redhat.com/jenkins/job/merge_pull_requests_origin/4506/console

1)  unauthenticated user should be able to log in
  - Error: Error while waiting for Protractor to sync with the page: "[ng:test] no injector found for element argument to getTestability\nhttp://errors.angularjs.org/1.3.8/ng/test"
      at Error (<anonymous>)
    at /data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/promise.js:1526:15
...
==== async task ====
Protractor.waitForAngular()
    at [object Object].webdriver.WebDriver.schedule (/data/src/github.com/openshift/origin/assets/node_modules/protractor/node_modules/selenium-webdriver/lib/webdriver/webdriver.js:345:15)
...
Error
    at [object Object].<anonymous> (/data/src/github.com/openshift/origin/assets/test/integration/e2e.js:21:5)
    at [object Object].<anonymous> (/data/src/github.com/openshift/origin/assets/test/integration/e2e.js:12:3)
    at Object.<anonymous> (/data/src/github.com/openshift/origin/assets/test/integration/e2e.js:4:1)

Screenshot:

[spadgett]

I wonder if adding the ng-cloak around the user dropdown made this worse. Looks like we're waiting for the presence of the .username element, which won't appear now until Angular is loaded and the template is compiled.

  driver.wait(function() {
    return driver.isElementPresent(by.css(".navbar-utility .username"));
  }, 3000);

I'm not sure 3s is long enough.

[spadgett]

I realize there are different errors here, but the most common seems to be login.

[spadgett]

Hm, maybe isElementPresent will still return true with ng-cloak because the element is there, just not visible.

[smarterclayton]

Haven't seen since.

[smarterclayton]

Nm, caps was new.

[soltysh]

Seen here: https://ci.openshift.redhat.com/jenkins/job/merge_pull_requests_origin/5023/console
And this screenshot appears relevant:

[spadgett]

UI tests are looking for a deployment triggered by an image change, but it's not there.

https://github.com/openshift/origin/blob/master/assets/test/integration/e2e.js#L67

[jwforres]

@ironcladlou what would cause a deployment to say it got triggered by a config change when it was really triggered by an image change?? see the screenshot #5881 (comment) the frontend deployment was the result of a build pushing an image.

[ironcladlou]

@jwforres

@ironcladlou what would cause a deployment to say it got triggered by a config change when it was really triggered by an image change?? see the screenshot #5881 (comment) the frontend deployment was the result of a build pushing an image.

I think we'll need to see the deploymentConfig itself to verify assumptions first (i.e. is the image trigger cause really missing from deploymentConfig.status). Possible to get?

[jwforres]

@ironcladlou should be in the etcd dump from test-end-to-end https://ci.openshift.redhat.com/jenkins/job/merge_pull_requests_origin/5023/artifact/origin/artifacts/test-end-to-end-docker/artifacts/etcd_dump.json

[ironcladlou]

@jwforres
@ironcladlou should be in the etcd dump from test-end-to-end https://ci.openshift.redhat.com/jenkins/job/merge_pull_requests_origin/5023/artifact/origin/artifacts/test-end-to-end-docker/artifacts/etcd_dump.json

Thanks. I made #7464 to track it.

[danmcp]

I don't think this is valid here anymore.