def get_logging_namespace(self):
""" Determine which logging namespace is in use """
# Assume the correct namespace is 'openshift-logging' and fall back to 'logging'
# if that assumption ends up being wrong.
oc_client = OCUtil(namespace='openshift-logging',
config_file='/tmp/admin.kubeconfig',
verbose=self.args.verbose)
logger.info("Determining which namespace is in use...")
try:
oc_client.get_dc('logging-kibana')
# If the previous call didn't throw an exception, logging is deployed in this namespace.
logger.info("Using namespace: openshift-logging")
return 'openshift-logging'
except subprocess.CalledProcessError:
logger.info("Using namespace: logging")
return 'logging'
def main():
''' Gather and send details on all visible S3 buckets '''
logger.info("start")
discovery_key = "disc.aws"
discovery_macro = "#S3_BUCKET"
prototype_s3_size = "disc.aws.size"
prototype_s3_count = "disc.aws.objects"
args = parse_args()
if args.verbose:
logger.setLevel(logging.DEBUG)
logger.debug("verbose flag set")
ocutil = OCUtil()
dc_yaml = ocutil.get_dc('docker-registry')
registry_config_secret = get_registry_config_secret(dc_yaml)
oc_yaml = ocutil.get_secrets(registry_config_secret)
aws_access, aws_secret = get_aws_creds(oc_yaml)
awsutil = AWSUtil(aws_access, aws_secret, args.debug)
bucket_list = awsutil.get_bucket_list(args.debug)
bucket_stats = {}
for bucket in bucket_list:
s3_size, s3_objects = awsutil.get_bucket_info(bucket, args.debug)
bucket_stats[bucket] = {"size": s3_size, "objects": s3_objects}
if args.debug:
print "Bucket stats: " + str(bucket_stats)
if args.test:
print "Test-only. Received results: " + str(bucket_stats)
else:
zgs = ZaggSender(verbose=args.debug)
zgs.add_zabbix_dynamic_item(discovery_key, discovery_macro,
bucket_list)
for bucket in bucket_stats.keys():
zab_key = "{}[{}]".format(prototype_s3_size, bucket)
zgs.add_zabbix_keys(
{zab_key: int(round(bucket_stats[bucket]["size"]))})
zab_key = "{}[{}]".format(prototype_s3_count, bucket)
zgs.add_zabbix_keys({zab_key: bucket_stats[bucket]["objects"]})
zgs.send_metrics()
def main():
''' Gather and send details on all visible S3 buckets '''
discovery_key = "disc.gcp"
discovery_macro = "#GCS_BUCKET"
prototype_bucket_size = "disc.gcp.size"
prototype_bucket_count = "disc.gcp.objects"
args = parse_args()
ocutil = OCUtil()
dc_yaml = ocutil.get_dc('docker-registry')
registry_config_secret = get_registry_config_secret(dc_yaml)
oc_yaml = ocutil.get_secrets(registry_config_secret)
bucket = get_gcp_info(oc_yaml)
gsutil = GcloudUtil(verbose=args.debug)
bucket_list = gsutil.get_bucket_list()
bucket_stats = {}
for bucket in bucket_list:
size, objects = gsutil.get_bucket_info(bucket)
bucket_stats[bucket] = {"size": size, "objects": objects}
if args.debug:
print "Bucket stats: " + str(bucket_stats)
if args.test:
print "Test-only. Received results: " + str(bucket_stats)
else:
zgs = ZaggSender(verbose=args.debug)
zgs.add_zabbix_dynamic_item(discovery_key, discovery_macro, bucket_list)
for bucket in bucket_stats.keys():
zab_key = "{}[{}]".format(prototype_bucket_size, bucket)
zgs.add_zabbix_keys({zab_key: int(round(bucket_stats[bucket]["size"]))})
zab_key = "{}[{}]".format(prototype_bucket_count, bucket)
zgs.add_zabbix_keys({zab_key: bucket_stats[bucket]["objects"]})
zgs.send_metrics()
def main():
''' Gather and send details on all visible S3 buckets '''
#get the region
with open('/container_setup/monitoring-config.yml', 'r') as f:
doc = yaml.load(f)
bucket_region = doc['oso_region']
args = parse_args()
ocutil = OCUtil()
dc_yaml = ocutil.get_dc('docker-registry')
registry_config_secret = get_registry_config_secret(dc_yaml)
oc_yaml = ocutil.get_secrets(registry_config_secret)
aws_access, aws_secret = get_aws_creds(oc_yaml)
awsutil = AWSUtil(aws_access, aws_secret, args.debug)
bucket_list = awsutil.get_bucket_list(verbose=args.debug,
BucketRegion=bucket_region)
bucket_stats = {}
for bucket in bucket_list:
#print bucket
s3_size, s3_objects = awsutil.get_bucket_info(
bucket, verbose=args.debug, BucketRegion=bucket_region)
bucket_stats[bucket] = {"size": s3_size, "objects": s3_objects}
if args.debug:
print "Bucket stats: " + str(bucket_stats)
if args.test:
print "Test-only. Received results: " + str(bucket_stats)
else:
send_zagg_data(bucket_list, bucket_stats, args)
class InfraNodePodStatus(object):
'''
This is a check for making sure the internal pods like
router and registry running and located on different infra nodes
'''
def __init__(
self,
args=None,
):
'''initial for the InfraNodePodStatus'''
self.args = args
self.kubeconfig = '/tmp/admin.kubeconfig'
self.oc = OCUtil(namespace=self.args.namespace,
config_file=self.kubeconfig)
self.all_pods = self.get_all_pods()
def get_all_pods(self):
''' get all the pod information '''
pods = self.oc.get_pods()
pod_report = {}
for pod in pods['items']:
pod_name = pod['metadata']['name']
pod_report[pod_name] = {}
pod_report[pod_name]['hostIP'] = pod['status']['hostIP']
pod_report[pod_name]['status'] = pod['status']['phase']
return pod_report
def get_expected_replicas(self, deploymentconfig):
''' get expected replica count from deploymentconfig '''
defined_replicas = self.oc.get_dc(deploymentconfig)['spec']['replicas']
return defined_replicas
def get_pods_by_name(self, podname):
"""get_pods_by_name"""
return [
self.all_pods[i] for i in self.all_pods.keys()
if i.startswith(podname + '-')
]
def check_pods(
self,
podname,
keybase="",
pod_optional=False,
):
''' to compare the pod host ip and check the pod status '''
logging.getLogger().info("Finding pods for: %s", podname)
result_code = 1
pods = self.get_pods_by_name(podname)
logging.getLogger().info("Pods Found: %s", len(pods))
expected_replicas = 0
try:
expected_replicas = self.get_expected_replicas(podname)
except Exception:
logging.getLogger().warn("dc not found for pod %s", podname)
if pod_optional:
logging.getLogger().warn(
"Some clusters don't have pod %s, please confirm before trying to fix this",
podname)
return # nothing we should do, so quit early, don't do more checks
logging.getLogger().info("Expected Replicas: %s", expected_replicas)
if len(pods) != expected_replicas:
result_code = 0
logging.getLogger().critical("Count Pods and Replicas don't match")
count_pods_running = len([i for i in pods if i['status'] == "Running"])
logging.getLogger().info("Pods Running: %s", count_pods_running)
if len(pods) != count_pods_running:
result_code = 0
logging.getLogger().critical("Some pods are not in running state")
host_ips = set([x['hostIP'] for x in pods])
logging.getLogger().info("Hosts found: %d", len(host_ips))
if len(host_ips) < 2 or len(pods) < 2:
result_code = 0
logging.getLogger().critical(
"%s has %d pods on %d hosts, not distributed", podname,
len(pods), len(host_ips))
if result_code == 0:
logging.getLogger().critical("Please check pods are in running "
"state, and on unique hosts")
logging.getLogger().critical("oc get pods -n %s -o wide",
self.args.namespace)
# result_code 1 means the pods are on different nodes
# count_pods_running means the running pod number
self.send_metrics(keybase=keybase,
location=result_code,
status=count_pods_running)
def send_metrics(self, keybase="", location="", status=""):
"""send_metrics"""
ms = MetricSender(verbose=self.args.verbose)
ms.add_metric({keybase + '.location': location})
ms.add_metric({keybase + '.status': status})
ms.send_metrics()
