1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
|
try:
# py2/3
import simplejson as json
except ImportError:
# py2
import json
from pkg_resources import parse_version
import ddt
from glusto.core import Glusto as g
import pytest
from openshiftstoragelibs.baseclass import GlusterBlockBaseClass
from openshiftstoragelibs import command
from openshiftstoragelibs import exceptions
from openshiftstoragelibs import heketi_ops
from openshiftstoragelibs import openshift_ops
@ddt.ddt
class TestPrometheusAndGlusterRegistryValidation(GlusterBlockBaseClass):
def setUp(self):
"""Initialize all the variables which are necessary for test cases"""
super(TestPrometheusAndGlusterRegistryValidation, self).setUp()
try:
prometheus_config = g.config['openshift']['prometheus']
self._prometheus_project_name = prometheus_config[
'prometheus_project_name']
self._prometheus_resources_selector = prometheus_config[
'prometheus_resources_selector']
self._alertmanager_resources_selector = prometheus_config[
'alertmanager_resources_selector']
self._registry_heketi_server_url = (
g.config['openshift']['registry_heketi_config'][
'heketi_server_url'])
self._registry_project_name = (
g.config['openshift']['registry_project_name'])
except KeyError as err:
self.skipTest("Config file doesn't have key {}".format(err))
# Skip the test if iscsi-initiator-utils version is not the expected
cmd = ("rpm -q iscsi-initiator-utils "
"--queryformat '%{version}-%{release}\n'"
"| cut -d '.' -f 1,2,3,4")
e_pkg_version = "6.2.0.874-17"
for g_server in self.gluster_servers:
out = self.cmd_run(cmd, g_server)
if parse_version(out) < parse_version(e_pkg_version):
self.skipTest(
"Skip the test as iscsi-initiator-utils package version {}"
"is less than version {} found on the node {}, for more "
"info refer to BZ-1624670".format(
out, e_pkg_version, g_server))
self._master = self.ocp_master_node[0]
# Switch to namespace conatining prometheus pods
cmd = "oc project --short=true"
current_project = command.cmd_run(cmd, self._master)
openshift_ops.switch_oc_project(
self._master, self._prometheus_project_name)
self.addCleanup(
openshift_ops.switch_oc_project, self._master, current_project)
def _fetch_metric_from_promtheus_pod(self, metric):
"""Fetch metric from prometheus pod using api call"""
prometheus_pods = list(openshift_ops.oc_get_pods(
self._master, selector=self._prometheus_resources_selector).keys())
fetch_metric_cmd = ("curl 'http://localhost:9090/api/v1/query"
"?query={}'".format(metric))
ret, metric_data, _ = openshift_ops.oc_rsh(
self._master, prometheus_pods[0], fetch_metric_cmd)
metric_result = json.loads(metric_data)["data"]["result"]
if (not metric_result) or ret:
raise exceptions.ExecutionError(
"Failed to fecth data for metric {}, output {}".format(
metric, metric_result))
return metric_result
def _get_pod_names_and_pvc_names(self):
# Get pod names and PVC names
pod_custom = ".:metadata.name"
pvc_custom = ":.spec.volumes[*].persistentVolumeClaim.claimName"
pvc_names, pod_names = [], []
for selector in (self._prometheus_resources_selector,
self._alertmanager_resources_selector):
pods = openshift_ops.oc_get_custom_resource(
self._master, "pod", pod_custom, selector=selector)
pod_names.extend(pods)
for pod_name in pods:
pvc_name = openshift_ops.oc_get_custom_resource(
self._master, "pod", pvc_custom, pod_name[0])[0]
pvc_names.append(pvc_name)
return pod_names, pvc_names
@pytest.mark.tier2
def test_promethoues_pods_and_pvcs(self):
"""Validate prometheus pods and PVC"""
# Wait for PVCs to be bound
pod_names, pvc_names = self._get_pod_names_and_pvc_names()
openshift_ops.wait_for_pvcs_be_bound(self._master, pvc_names)
# Validate that there should be no or zero pods in non-running state
field_selector, pod_count = "status.phase!=Running", 0
openshift_ops.wait_for_pods_be_ready(
self._master, pod_count, field_selector=field_selector)
# Validate iscsi and multipath
for (pvc_name, pod_name) in zip(pvc_names, pod_names):
self.verify_iscsi_sessions_and_multipath(
pvc_name, pod_name[0], rtype='pod',
heketi_server_url=self._registry_heketi_server_url,
is_registry_gluster=True)
# Try to fetch metric from prometheus pod
self._fetch_metric_from_promtheus_pod(metric='kube_node_info')
@ddt.data('delete', 'drain')
@pytest.mark.tier2
def test_respin_prometheus_pod(self, motive="delete"):
"""Validate respin of prometheus pod"""
if motive == 'drain':
# Get the number of infra nodes
infra_node_count_cmd = (
'oc get nodes '
'--no-headers -l node-role.kubernetes.io/infra=true|wc -l')
infra_node_count = command.cmd_run(
infra_node_count_cmd, self._master)
# Skip test case if number infra nodes are less than #2
if int(infra_node_count) < 2:
self.skipTest('Available number of infra nodes "{}", it should'
' be more than 1'.format(infra_node_count))
# Get PVC names and pod names
pod_names, pvc_names = self._get_pod_names_and_pvc_names()
# Validate iscsi and multipath
for (pvc_name, pod_name) in zip(pvc_names, pod_names):
_, _, node = self.verify_iscsi_sessions_and_multipath(
pvc_name, pod_name[0], rtype='pod',
heketi_server_url=self._registry_heketi_server_url,
is_registry_gluster=True)
# Delete the prometheus pods
if motive == 'delete':
for pod_name in pod_names:
openshift_ops.oc_delete(self._master, 'pod', pod_name[0])
# Drain the node
elif motive == 'drain':
drain_cmd = ('oc adm drain {} --force=true --ignore-daemonsets '
'--delete-local-data'.format(node))
command.cmd_run(drain_cmd, hostname=self._master)
# Cleanup to make node schedulable
cmd_schedule = (
'oc adm manage-node {} --schedulable=true'.format(node))
self.addCleanup(
command.cmd_run, cmd_schedule, hostname=self._master)
# Validate that there should be no or zero pods in non-running state
field_selector, pod_count = "status.phase!=Running", 0
openshift_ops.wait_for_pods_be_ready(
self._master, pod_count, field_selector=field_selector)
# Validate iscsi and multipath
for (pvc_name, pod_name) in zip(pvc_names, pod_names):
self.verify_iscsi_sessions_and_multipath(
pvc_name, pod_name[0], rtype='pod',
heketi_server_url=self._registry_heketi_server_url,
is_registry_gluster=True)
# Try to fetch metric from prometheus pod
self._fetch_metric_from_promtheus_pod(metric='kube_node_info')
@pytest.mark.tier2
def test_heketi_and_prometheus_device_count(self):
"""Check if device count is same in heketi and promtheus"""
cluster_ids_metrics, cluster_ids_promtheus = [], []
hostnames_metrics, hostnames_promtheus = [], []
total_value_metrics, total_value_promtheus = 0, 0
metrics = heketi_ops.get_heketi_metrics(
self.heketi_client_node, self.heketi_server_url)
heketi_device_count_metric = metrics.get('heketi_device_count')
for result in heketi_device_count_metric:
cluster_ids_metrics.append(result.get('cluster'))
hostnames_metrics.append(result.get('hostname'))
total_value_metrics += int(result.get('value'))
metric_result = self._fetch_metric_from_promtheus_pod(
metric='heketi_device_count')
for result in metric_result:
total_value_promtheus += int(result.get('value')[1])
cluster_ids_promtheus.append(result.get('metric')['cluster'])
hostnames_promtheus.append(result.get('metric')['hostname'])
self.assertEqual(cluster_ids_metrics, cluster_ids_promtheus,
"Cluster ID's are not same")
self.assertEqual(hostnames_metrics, hostnames_promtheus,
"Hostnames are not same")
self.assertEqual(total_value_metrics, total_value_promtheus,
"Total device counts are not same")
|