def check(warn, crit, project):
if crit < warn:
msg = "critical threshold cannot be lower than warning threshold: %d < %d"
raise ValueError(msg % (crit, warn))
if not project:
project = openshift.get_project()
results = []
errors = []
pcs = openshift.get_running_pod_containers(project)
for pod_name, container_name, container_data in pcs:
memory_limit = container_data.get('resources',
{}).get('limits', {}).get('memory')
try:
memory_usage = openshift.exec_in_pod_container(
project, pod_name, container_name, check_memory_usage_cmd)
if memory_limit:
memory_limit = openshift.exec_in_pod_container(
project, pod_name, container_name, check_memory_limit_cmd)
results.extend(
analize(pod_name, container_name, memory_limit,
memory_usage, warn, crit))
except Exception as e:
errors.append((pod_name, container_name, e))
return report(results, errors)
def check(service, containers):
project = openshift.get_project()
selectors = openshift.get_service_selectors(project, service)
pcs = openshift.get_running_pod_containers(
project, selector=selectors, container_names=containers.split(','))
# since core only has single mysql with replication, we only check one even if there are multiple results
pod_name, container_name, container_data = pcs[0]
env = openshift.get_container_env(project, pod_name, container_name)
args = (check_mysql_path, "-H", service, "-u", env["MYSQL_USER"], "-p",
env["MYSQL_PASSWORD"])
try:
output = check_output(args)
nag_status = nagios.OK
except CalledProcessError as cpe:
if cpe.returncode in (nagios.OK, nagios.WARN, nagios.CRIT,
nagios.UNKNOWN):
output = cpe.output
nag_status = cpe.returncode
else:
output = str(cpe)
nag_status = nagios.UNKNOWN
return report(nag_status, output)
def check(warn, crit, minimum, project):
if crit < warn:
msg = "critical threshold cannot be lower than warning threshold: %d < %d"
raise ValueError(msg % (crit, warn))
if not project:
project = openshift.get_project()
results = []
errors = []
pcs = openshift.get_running_pod_containers(project)
for pod_name, container_name, container_data in pcs:
if "mongo" in pod_name:
try:
result = openshift.exec_in_pod_container(
project, pod_name, container_name, check_disk_cmd)
results.extend(
analize(pod_name, container_name, parse_df_lines(result),
warn, crit))
except Exception as e:
errors.append((pod_name, container_name, e))
return report(results, errors, minimum)
def check(service, containers):
project = openshift.get_project()
selectors = openshift.get_service_selectors(project, service)
pcs = openshift.get_running_pod_containers(project, selector=selectors, container_names=containers.split(','))
# since core only has single mysql with replication, we only check one even if there are multiple results
pod_name, container_name, container_data = pcs[0]
env = openshift.get_container_env(project, pod_name, container_name)
args = (check_mysql_path, "-H", service, "-u", env["MYSQL_USER"], "-p", env["MYSQL_PASSWORD"])
try:
output = check_output(args)
nag_status = nagios.OK
except CalledProcessError as cpe:
if cpe.returncode in (nagios.OK, nagios.WARN, nagios.CRIT, nagios.UNKNOWN):
output = cpe.output
nag_status = cpe.returncode
else:
output = str(cpe)
nag_status = nagios.UNKNOWN
return report(nag_status, output)
def check(containers):
project = openshift.get_project()
pods = openshift.get_running_pod_names(project, container_names=containers.split(','))
rs_statuses = map(parse_mongo_result, openshift.exec_in_pods(project, pods, check_mongodb_cmd))
nag_statuses = map(analize, rs_statuses)
return report(pods, rs_statuses, nag_statuses)
def check(containers):
project = openshift.get_project()
pods = openshift.get_running_pod_names(
project, container_names=containers.split(','))
collections = map(parse_mongo_result,
openshift.exec_in_pods(project, pods, check_mongodb_cmd))
result = topic_existance(topics, collections)
return report(pods, project, result)
def check(containers):
project = openshift.get_project()
pods = openshift.get_running_pod_names(
project, container_names=containers.split(','))
rs_statuses = map(parse_mongo_result,
openshift.exec_in_pods(project, pods, check_mongodb_cmd))
nag_statuses = map(analize, rs_statuses)
return report(pods, rs_statuses, nag_statuses)
def check(warn, crit):
if crit < warn:
msg = "critical threshold cannot be lower than warning threshold: %d < %d"
raise ValueError(msg % (crit, warn))
project = openshift.get_project()
results = []
pods = openshift.get_running_pod_names(project)
execs = openshift.exec_in_pods(project, pods, check_disk_cmd)
for pod, lines in zip(pods, execs):
results.extend(analize(pod, parse_df_lines(lines), warn, crit))
return report(results)
def check():
issues = []
project = openshift.get_project()
deploymentConfigs = openshift.get_deploymentconfigs(project)
for deploymentConfig in deploymentConfigs["items"]:
componentName = deploymentConfig["metadata"]["name"]
pods = openshift.get_running_pod_names(
project, container_names=componentName)
nodes = openshift.get_nodes_from_names(pods)
if len(pods) > 1:
for node in set(nodes):
nodeCount = nodes.count(node)
if nodeCount > 1:
issues.append("WARN: %s has %s pods running on the same node: %s" % (
componentName, nodeCount, node))
return report(issues)
def check(project):
if not project:
project = openshift.get_project()
results = []
errors = []
pcs = openshift.get_running_pod_containers(project)
for pod_name, container_name, container_data in pcs:
resource_limits = container_data.get('resources', {}).get('limits', {})
try:
results.extend(analize(pod_name, container_name, resource_limits))
except Exception as e:
errors.append((pod_name, container_name, e))
return report(results, errors)
def check(project):
if not project:
project = openshift.get_project()
results = []
errors = []
pcs = openshift.get_running_pod_containers(project)
for pod_name, container_name, container_data in pcs:
resource_limits = container_data.get('resources', {}).get('limits', {})
try:
results.extend(analize(pod_name, container_name, resource_limits))
except Exception as e:
errors.append((pod_name, container_name, e))
return report(results, errors)
def check():
project = openshift.get_project()
pods = openshift.get_running_pod_names(project, container_names="mongodb")
if not pods:
output = "Unable to locate any mongodb containers"
return nagios.UNKNOWN
nodes = openshift.get_nodes_from_names(pods)
nodes_pods = dict(zip(pods, nodes))
if len(nodes) < 3:
output = nodes_pods
return nagios.CRIT
if nodes[0] == nodes[1] or nodes[0] == nodes[2] or nodes[1] == nodes[2]:
output = nodes_pods
nag_status = nagios.WARN
else:
output = nodes_pods
nag_status = nagios.OK
return report(nag_status, output)
def check(warn, crit, minimum, project):
if crit < warn:
msg = "critical threshold cannot be lower than warning threshold: %d < %d"
raise ValueError(msg % (crit, warn))
if not project:
project = openshift.get_project()
results = []
errors = []
pcs = openshift.get_running_pod_containers(project)
for pod_name, container_name, container_data in pcs:
try:
result = openshift.exec_in_pod_container(project, pod_name, container_name, check_disk_cmd)
results.extend(analize(pod_name, container_name, parse_df_lines(result), warn, crit))
except Exception as e:
errors.append((pod_name, container_name, e))
return report(results, errors, minimum)
def check(warn, crit, project):
if crit < warn:
msg = "critical threshold cannot be lower than warning threshold: %d < %d"
raise ValueError(msg % (crit, warn))
if not project:
project = openshift.get_project()
results = []
errors = []
prev_cpu_usage = read_cpu_usage()
curr_cpu_usage = {}
pcs = openshift.get_running_pod_containers(project)
for pod_name, container_name, container_data in pcs:
cpu_limit = container_data.get('resources', {}).get('limits',
{}).get('cpu')
try:
if cpu_limit:
curr_usage, curr_uptime, limit = get_container_cpu_usage(
project, pod_name, container_name)
curr_cpu_usage[pod_name] = curr_cpu_usage.get(pod_name, {})
curr_cpu_usage[pod_name][container_name] = [
curr_usage, curr_uptime
]
prev_usage, prev_uptime = prev_cpu_usage.get(pod_name, {}).get(
container_name, [None, None])
results.extend(
analize(pod_name, container_name, prev_usage, curr_usage,
prev_uptime, curr_uptime, limit, warn, crit))
except Exception as e:
errors.append((pod_name, container_name, e))
write_cpu_usage(curr_cpu_usage)
return report(results, errors)
def check(project):
if not project:
project = openshift.get_project()
results, error = do_request(project)
return report(results, error)
def check(project):
if not project:
project = openshift.get_project()
data = do_request(project)
results, errors = parse_response(data)
return report(results, errors)