E2E: investigate failure of configuration update test about cloudnative-pg HOT 6 CLOSED

Add some trace in branch to investigating

from cloudnative-pg.

A failure https://github.com/cloudnative-pg/cloudnative-pg/runs/6384630653?check_suite_focus=true#step:8:1814
the pg sever kept in pending restart status for 300 seconds, then case failed. investigating...

from cloudnative-pg.

The failure happens waiting for all pods in a status without pending restart, during that time, from the log I find he primary needs to be restarted, we'll trigger a switchover to do that, this means the instance is going to have a swtich over. probally this is the required switch over in prevous test case, so we need give privous test case a well good ending.

from cloudnative-pg.

The failure is caused by a bug. while do the configuration change by decreasing the sensitive param. expected process is

1. primary decrease by a restart.
2. standby decrease. there is no switchover. but switchover happens after apply the decreasing configuration. opened another tiket to solve the issue, and use this one to refact the e2e

from cloudnative-pg.

The fix for failure is covered by issue #168

for this issue, refactor the configuration test suite, move each configuration change into a seperate spec, make it well orginazed and easy to diagnose and expand in future.

from cloudnative-pg.

LGTM.

from cloudnative-pg.