mgr/dashboard: fix ESOCKETTIMEDOUT E2E failure

[avanthakkar]

Increasing responseTimeout may help resolve the e2e failures with cy.visit()/cy.request() failed to load error, as it's considered the most accepted solution here cypress-io/cypress#7062 too, which happens only in CI jobs. The top failure jobs in last 3 months have the same error of timeout/failed to load.
Example error message : cy.request() timed out waiting 30000ms for a response from your server.
Fixes: https://tracker.ceph.com/issues/49828
Signed-off-by: Avan Thakkar athakkar@redhat.com

Checklist

• References tracker ticket
• Updates documentation if necessary
• Includes tests for new functionality or reproducer for bug

---

Show available Jenkins commands

• jenkins retest this please
• jenkins test classic perf
• jenkins test crimson perf
• jenkins test signed
• jenkins test make check
• jenkins test make check arm64
• jenkins test submodules
• jenkins test dashboard
• jenkins test api
• jenkins test docs
• jenkins render docs
• jenkins test ceph-volume all
• jenkins test ceph-volume tox

[avanthakkar]

I think we don't need to make changes in script in order to change the responseTimeout. As when we run npx cypress run... it looks for cypress.json file and takes the environment var values from there unless we have changed it in script itself like for baseURL. Also no need to change defaultCommandTimeout as it's taken as 120s only while running e2e jenkins job. @alfonsomthd

[avanthakkar]

I don't understand why we do this in afterEach hook ? Shouldn't we instead call something like cy.logout(), as for all components testing we login in beforeEach() hook and I think it also keeps up with best practices. Thoughts @ceph/dashboard ?

[alfonsomthd]

I don't understand why we do this in afterEach hook ? Shouldn't we instead call something like cy.logout(), as for all components testing we login in beforeEach() hook and I think it also keeps up with best practices. Thoughts @ceph/dashboard ?

Try to run the whole suite without the afterEach hook to check if it's not needed anymore.

[avanthakkar]

I don't understand why we do this in afterEach hook ? Shouldn't we instead call something like cy.logout(), as for all components testing we login in beforeEach() hook and I think it also keeps up with best practices. Thoughts @ceph/dashboard ?

Try to run the whole suite without the afterEach hook to check if it's not needed anymore.

Yes..but I'm sure abt why we redirect to 403 page..I understand atleast why we should redirect (I think it's following best practices in order to clean up the state generated by the current test(s)), but not sure about 403 page in specific..can't we do instead logout?

But in any case the fix should remain the same i.e. to increase responseTimeout. It sounds a bit strange here as usually we should increase it if cy.request() is taking long enough than expected, but looking at this thread I think it's the same issue here, also the workaround suggested looks the same.

[avanthakkar]

jenkins test dashboard

[avanthakkar]

jenkins test dashboard

[epuertat]

Are we sure we need to increase the response time from 30 secs to 45 secs? Are we talking about HTTP requests? If that's the case I feel that the issue will be somewhere else :/

[avanthakkar]

Are we sure we need to increase the response time from 30 secs to 45 secs? Are we talking about HTTP requests? If that's the case I feel that the issue will be somewhere else :/

I followed this cypress-io/cypress#7062 (comment) and I think it's same issue for us. Somehow cy.request() is taking some more time in cypress browser env so cy.visit() after it is failing to load(also I tried removing that afterEach hook(which isn't consider best practice) but it's somehow broke some tests).