Details
-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
4.12
-
None
-
No
-
CLOUD Sprint 251, CLOUD Sprint 252, CLOUD Sprint 253
-
3
-
False
-
Description
Description of problem:
on Azure, when kube-controller-manager verify whether a machine exists or not, if the machine was already deleted, the code may panic with sigsegv
I0320 12:02:55.806321 1 azure_backoff.go:91] GetVirtualMachineWithRetry(worker-e32ads-westeurope2-f72dr): backoff success
I0320 12:02:56.028287 1 azure_wrap.go:201] Virtual machine "worker-e16as-westeurope1-hpz2t" is under deleting
I0320 12:02:56.028328 1 azure_standard.go:752] GetPrimaryInterface(worker-e16as-westeurope1-hpz2t, ) abort backoff
E0320 12:02:56.028334 1 azure_standard.go:825] error: az.EnsureHostInPool(worker-e16as-westeurope1-hpz2t), az.VMSet.GetPrimaryInterface.Get(worker-e16as-westeurope1-hpz2t, ), err=instance not found
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x60 pc=0x33d21f6]goroutine 240642 [running]:
k8s.io/legacy-cloud-providers/azure.(*availabilitySet).EnsureHostInPool(0xc000016580, 0xc0262fb400, {0xc02d8a5080, 0x32}, {0xc021c1bc70, 0xc4}, {0x0, 0x0}, 0xa8?)
vendor/k8s.io/legacy-cloud-providers/azure/azure_standard.go:831 +0x4f6
k8s.io/legacy-cloud-providers/azure.(*availabilitySet).EnsureHostsInPool.func2()
vendor/k8s.io/legacy-cloud-providers/azure/azure_standard.go:928 +0x5f
k8s.io/apimachinery/pkg/util/errors.AggregateGoroutines.func1(0xc0159d0788?)
Version-Release number of selected component (if applicable):
4.12.48
(ships https://github.com/openshift/kubernetes/commit/6df21776c7879727ab53895df8a03e53fb725d74)
issue introduced by https://github.com/kubernetes/kubernetes/pull/111428/files#diff-0414c3aba906b2c0cdb2f09da32bd45c6bf1df71cbb2fc55950743c99a4a5fe4
How reproducible:
was unable to reproduce, happens occasionally
Steps to Reproduce:
1.
2.
3.
Actual results:
panic
Expected results:
no panic
Additional info:
internal case 03772590
