Fix panic in acceptance tests

[grobinson-grafana]

This commit attempts to address a panic that occurs in acceptance tests if a server in the cluster fails to start:

panic: runtime error: invalid memory address or nil pointer dereference
	panic: runtime error: invalid memory address or nil pointer dereference [recovered]
	panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x2 addr=0xa0 pc=0x101785858]

goroutine 27 [running]:
testing.tRunner.func1.2({0x101cc73e0, 0x1023c09c0})
	/usr/local/go/src/testing/testing.go:1545 +0x1c8
testing.tRunner.func1()
	/usr/local/go/src/testing/testing.go:1548 +0x360
panic({0x101cc73e0?, 0x1023c09c0?})
	/usr/local/go/src/runtime/panic.go:914 +0x218
github.com/prometheus/alertmanager/test/with_api_v2.(*Alertmanager).Terminate(0x14000a50300)
	/Users/grobinson/go/src/github.com/prometheus/alertmanager/test/with_api_v2/acceptance.go:386 +0x38
github.com/prometheus/alertmanager/test/with_api_v2.(*AcceptanceTest).Run.func1(0x14000a50300)
	/Users/grobinson/go/src/github.com/prometheus/alertmanager/test/with_api_v2/acceptance.go:172 +0x28
panic({0x101cc73e0?, 0x1023c09c0?})
	/usr/local/go/src/runtime/panic.go:920 +0x26c
github.com/prometheus/alertmanager/test/with_api_v2.(*Alertmanager).Terminate(0x14000a505a0)
	/Users/grobinson/go/src/github.com/prometheus/alertmanager/test/with_api_v2/acceptance.go:386 +0x38
github.com/prometheus/alertmanager/test/with_api_v2.(*AcceptanceTest).Run.func1(0x14000a505a0)
	/Users/grobinson/go/src/github.com/prometheus/alertmanager/test/with_api_v2/acceptance.go:172 +0x28
runtime.Goexit()
	/usr/local/go/src/runtime/panic.go:541 +0x18c
testing.(*common).FailNow(0x14000103ba0)
	/usr/local/go/src/testing/testing.go:999 +0x48
testing.(*common).Fatal(0x14000103ba0, {0x1400092bdb8?, 0x0?, 0x14000a30330?})
	/usr/local/go/src/testing/testing.go:1076 +0x58
github.com/prometheus/alertmanager/test/with_api_v2.(*AcceptanceTest).Run(0x14000792240)
	/Users/grobinson/go/src/github.com/prometheus/alertmanager/test/with_api_v2/acceptance.go:181 +0x18c
github.com/prometheus/alertmanager/test/with_api_v2/acceptance.TestClusterDeduplication(0x14000103ba0)
	/Users/grobinson/go/src/github.com/prometheus/alertmanager/test/with_api_v2/acceptance/cluster_test.go:56 +0x680
testing.tRunner(0x14000103ba0, 0x101db24f8)
	/usr/local/go/src/testing/testing.go:1595 +0xe8
created by testing.(*T).Run in goroutine 1
	/usr/local/go/src/testing/testing.go:1648 +0x33c
FAIL	github.com/prometheus/alertmanager/test/with_api_v2/acceptance	0.454s
FAIL

It's not the perfect fix, but I'd like to see if it helps and how much. If it does work then we can look at improving on it further.

[grobinson-grafana]

The panic occurred because the test called Fatal which calls runtime.Goexit. This caused the stack to unwind and the defer function to be called. When this happened, am.cmd contained a nil Process field.