Search for sporadic failures 2 (group_management, execution_environments_use_in_controller.) #1707
Conversation
MilanPospisil
changed the title
| @@ -625,6 +625,9 @@ Cypress.Commands.add('syncRemoteContainer', {}, (name) => { | |||
| '.pf-c-alert__title', | |||
| `Sync started for execution environment "${name}".`, | |||
| ); | |||
| // wait for finish | |||
| cy.contains('a', 'detail page').click(); | |||
| cy.contains('.title-box h1', 'Completed', { timeout: 20000 }); | |||
There was a problem hiding this comment.
👀 .. but I agree this fixes the problem for now 👍
Definitely a FIXME for later though, seems like now that we know sync sometimes takes longer than the usual timeouts, we can fix it properly too...
we can wait for the alert, click that "detail page" link, land on a task detail...
and that page auto-updates to say Complete when done.
EDIT: oh, it already does that, and still needs the extra timeout .. oh well, nothing we can do for now, we'll probably need to sync from a local registry or something like that
|
Restarting tests just to make sure :) |
|
Merging, this fixes the group_management failure that's now happening everywhere 👍 But, see the failure here, execution_environment_use_in_controller is still failing sporadically. |
Issue: AAH-1410
Focus mostly on execution_enviroments_use_in_controller.js. It tries to repair it - it now waits for the delete task to be completed (delete user command).
And it repairs also group management - there is new error due to alert PR. New alerts now contains name of the user and the test had very bad selector (no selector) that searches the user name at the whole page (for not contains). It worked before, but now the alert also contains the user name and it wrongly finds it there and fail.