do not fail rollout when instance dies unexpectedly

[grosser]

[19:34:21] {"level"=>"error", "ts"=>1602099256.3341777, "logger"=>"controllers.RollingUpgrade", "msg"=>"Failed to runRestack", "rollingupgrade"=>"rollingupgrade-zorg-node-us-east-1c", "error"=>"Error updating instances, ErrorCount: 1, Errors: [InvalidInstanceID.NotFound: The instance ID 'i-05ec4ae16a265ef60' does not exist\n\tstatus code: 400, request id: 19b19626-88f9-41dc-8b86-b6f6377976ec]", "stacktrace"=>"github.com/go-logr/zapr.(*zapLogger).Error\n\t/go/pkg/mod/github.com/go-logr/[email protected]/zapr.go:128\ngithub.com/keikoproj/upgrade-manager/controllers.(*RollingUpgradeReconciler).error\n\t/workspace/controllers/rollingupgrade_controller.go:1202\ngithub.com/keikoproj/upgrade-manager/controllers.(*RollingUpgradeReconciler).Process\n\t/workspace/controllers/rollingupgrade_controller.go:706"}

not an urgent issue, but would be nice if instead of crashing it would retry ... simply restarted and it was fine

kubernetes: 1.19.2 keiko: 666fc9ff113b5b2d0990d91ae7c79fe68cfa895b 0.15.dev

[tekenstam]

I think this issue may be fixed by PR https://github.com/keikoproj/upgrade-manager/pull/138.

[uthark]

@grosser the issue should have been fixed by #138.