Description of problem:

- Customer updated the internal master DNS, after the change was propagated they restarted atomic-openshift-node in the env one by one.
- This node is in SIGSEGV crash loop after that change.
- The service appears to have stabilized after it was evacuated. It appears that it was crashing for each pod that was removed during `oc adm drain` but has been up since then.
- There are a lot coredumps generated under /var/lib/origin
- The other nodes doesn't show any malfunction


========= stack =======================
reconciler.go:376] Could not construct volume information: impossible to reconstruct glusterfs volume spec from volume mountpath
reconciler.go:376] Could not construct volume information: impossible to reconstruct glusterfs volume spec from volume mountpath
reconciler.go:376] Could not construct volume information: impossible to reconstruct glusterfs volume spec from volume mountpath
reconciler.go:376] Could not construct volume information: impossible to reconstruct glusterfs volume spec from volume mountpath
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x8 pc=0x43740cd]

### goroutine raising panic leads to volumemanager.(*volumeManager).Run

goroutine 447 [running]: 
panic(0x4d294e0, 0xf162930)
/usr/lib/golang/src/runtime/panic.go:540 +0x45e fp=0xc424609b48 sp=0xc424609aa0 pc=0x42e76eruntime.panicmem()
/usr/lib/golang/src/runtime/panic.go:63 +0x5e fp=0xc424609b68 sp=0xc424609b48 pc=0x42d49eruntime.sigpanic()
/usr/lib/golang/src/runtime/signal_unix.go:367 +0x17c fp=0xc424609bb8 sp=0xc424609b68 pc=0x445d5c
github.com/openshift/origin/vendor/k8s.io/kubernetes/pkg/kubelet/volumemanager/reconciler.(*reconciler).syncStates(0xc4223e6180, 0xc4225aa030, 0x2c)

[...]

created by github.com/openshift/origin/vendor/k8s.io/kubernetes/pkg/kubelet/volumemanager.(*volumeManager).Run
/builddir/build/BUILD/atomic-openshift-git-0.6bc473e/_output/local/go/src/github.com/openshift/origin/vendor/k8s.io/kubernetes/pkg/kubelet/volumemanager/volume_manager.go:249 +0x153
=======================================



===== analysis of a coredump =====
(dlv) thread 1076
Switched from 1369 to 1076

(dlv) bt
0  0x0000000000461354 in runtime.memmove at /usr/lib/golang/src/runtime/memmove_amd64.s:277
1  0x0000000000445eeb in runtime.sigtrampgo at /usr/lib/golang/src/runtime/signal_unix.go:774
2  0x000000c424609aa0 in ??? at ?:-1
3  0x000000000042f2c2 in runtime.startpanic_m at /usr/lib/golang/src/runtime/panic.go:658
4  0x01007fce58325dc0 in ??? at ?:-1
5  0x000000000045c2bc in runtime.newdefer.func1 at /usr/lib/golang/src/runtime/panic.go:208

(dlv) up 0
> runtime.memmove() /usr/lib/golang/src/runtime/memmove_amd64.s:277 (PC: 0x461354)
Frame 0: /usr/lib/golang/src/runtime/memmove_amd64.s:277 (PC: 461354)
   272:		MOVOU	X9, 144(DI)
   273:		MOVOU	X10, 160(DI)
   274:		MOVOU	X11, 176(DI)
   275:		MOVOU	X12, 192(DI)
   276:		MOVOU	X13, 208(DI)
=> 277:		MOVOU	X14, 224(DI)
   278:		MOVOU	X15, 240(DI)
   279:		CMPQ	BX, $256
   280:		LEAQ	256(SI), SI
   281:		LEAQ	256(DI), DI
   282:		JGE	move_256through2048

(dlv) up 1
> runtime.memmove() /usr/lib/golang/src/runtime/memmove_amd64.s:277 (PC: 0x461354)
Frame 1: /usr/lib/golang/src/runtime/signal_unix.go:774 (PC: 445eeb)
   769:		}
   770:		stsp := uintptr(unsafe.Pointer(st.ss_sp))
   771:		g.m.gsignal.stack.lo = stsp
   772:		g.m.gsignal.stack.hi = stsp + st.ss_size
   773:		g.m.gsignal.stackguard0 = stsp + _StackGuard
=> 774:		g.m.gsignal.stackguard1 = stsp + _StackGuard
   775:	}
   776:	
   777:	// restoreGsignalStack restores the gsignal stack to the value it had
   778:	// before entering the signal handler.
   779:	//go:nosplit

(dlv) up 3
> runtime.memmove() /usr/lib/golang/src/runtime/memmove_amd64.s:277 (PC: 0x461354)
Frame 4: ?:-1 (PC: 1007fce58325dc0)
======================================

Version-Release number of selected component (if applicable):
atomic-openshift-3.9.25-1.git.0.6bc473e.el7.x86_64

How reproducible:
Not reproducible