+++ This bug was initially created as a clone of Bug #1796716 +++
+++ This bug was initially created as a clone of Bug #1796715 +++
The upgrade test was running a service load balancer verification that ensured upgrades preserved reachability, but it provided insufficient alerting and debug to narrow down upgrade failures and assign probable cause. In addition, the test failed to verify reuse of connections vs newly created connections. Finally, we don't explicitly measure API reachability during upgrade.
Add an e2e test for upgrade that verifies all of these conditions and backport it to 4.2 to ensure we are sufficiently certain about the guarantees we provide to end users.
I’m adding UpcomingSprint, because I was occupied by fixing bugs with
higher priority/severity, developing new features with higher
priority, or developing new features to improve stability at a macro
level. I will revisit this bug next sprint.