aboutsummaryrefslogtreecommitdiffstats
path: root/functest_kubernetes
diff options
context:
space:
mode:
authorCédric Ollivier <cedric.ollivier@orange.com>2020-09-12 13:52:47 +0200
committerCédric Ollivier <cedric.ollivier@orange.com>2020-09-12 13:53:42 +0200
commit4f197110710e9f148eae4533792e8e7e2d72f053 (patch)
tree9d7002040a55c3f6df1853b0dab760ea3450ba10 /functest_kubernetes
parent3afa7d700eebf2bfcaabab7bdd3aa09198ba4fe0 (diff)
Enhance kube-hunter result postprocessing
It fills self.details and checks if the test case passes according to criteria (severity = high by default) Change-Id: Ib20779b4b5dca078c65b546c8703bc99856c6f41 Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
Diffstat (limited to 'functest_kubernetes')
-rw-r--r--functest_kubernetes/security/kube-hunter.yaml2
-rw-r--r--functest_kubernetes/security/security.py66
2 files changed, 64 insertions, 4 deletions
diff --git a/functest_kubernetes/security/kube-hunter.yaml b/functest_kubernetes/security/kube-hunter.yaml
index e5a81ed3..921005bc 100644
--- a/functest_kubernetes/security/kube-hunter.yaml
+++ b/functest_kubernetes/security/kube-hunter.yaml
@@ -9,6 +9,6 @@ spec:
- name: kube-hunter
image: aquasec/kube-hunter:0.3.1
command: ["python", "kube-hunter.py"]
- args: ["--pod"]
+ args: ["--pod", "--report", "json", "--statistics", '--active']
restartPolicy: Never
backoffLimit: 4
diff --git a/functest_kubernetes/security/security.py b/functest_kubernetes/security/security.py
index 2051f956..d06ffb28 100644
--- a/functest_kubernetes/security/security.py
+++ b/functest_kubernetes/security/security.py
@@ -13,14 +13,17 @@ Define the parent for Kubernetes testing.
from __future__ import division
+import json
import logging
import time
+import textwrap
import yaml
from kubernetes import client
from kubernetes import config
from kubernetes import watch
import pkg_resources
+import prettytable
from xtesting.core import testcase
@@ -37,6 +40,7 @@ class SecurityTesting(testcase.TestCase):
self.corev1 = client.CoreV1Api()
self.batchv1 = client.BatchV1Api()
self.pod = None
+ self.pod_log = ""
self.job_name = None
self.output_log_name = 'functest-kubernetes.log'
self.output_debug_log_name = 'functest-kubernetes.debug.log'
@@ -75,10 +79,9 @@ class SecurityTesting(testcase.TestCase):
pods = self.corev1.list_namespaced_pod(
self.namespace, label_selector='job-name={}'.format(self.job_name))
self.pod = pods.items[0].metadata.name
- api_response = self.corev1.read_namespaced_pod_log(
+ self.pod_log = self.corev1.read_namespaced_pod_log(
name=self.pod, namespace=self.namespace)
- self.__logger.warning("\n\n%s", api_response)
- self.result = 100
+ self.__logger.info("\n\n%s", self.pod_log)
def run(self, **kwargs):
assert self.job_name
@@ -119,10 +122,63 @@ class KubeHunter(SecurityTesting):
See https://github.com/aquasecurity/kube-hunter for more details
"""
+ __logger = logging.getLogger(__name__)
+
def __init__(self, **kwargs):
super(KubeHunter, self).__init__(**kwargs)
self.job_name = "kube-hunter"
+ def process_results(self, **kwargs):
+ """Process kube-hunter details"""
+ self.details = json.loads(self.pod_log.splitlines()[-1])
+ if self.details["vulnerabilities"]:
+ self.result = 100
+ msg = prettytable.PrettyTable(
+ header_style='upper', padding_width=5,
+ field_names=['category', 'vulnerability', 'severity'])
+ severity = kwargs.get("severity", "high")
+ if severity == "low":
+ allowed_severity = []
+ elif severity == "medium":
+ allowed_severity = ["low"]
+ elif severity == "high":
+ allowed_severity = ["low", "medium"]
+ else:
+ self.__logger.warning(
+ "Selecting high as default severity (%s is incorrect)",
+ kwargs.get("severity", "high"))
+ severity = "high"
+ allowed_severity = ["low", "medium"]
+ for vulnerability in self.details["vulnerabilities"]:
+ if vulnerability["severity"] in allowed_severity:
+ self.__logger.warning(
+ "Skipping %s (severity is configured as %s)",
+ vulnerability["vulnerability"], severity)
+ else:
+ self.result = 0
+ msg.add_row(
+ [vulnerability["category"], vulnerability["vulnerability"],
+ vulnerability["severity"]])
+ self.__logger.warning("\n\n%s\n", msg.get_string())
+ if self.details["hunter_statistics"]:
+ msg = prettytable.PrettyTable(
+ header_style='upper', padding_width=5,
+ field_names=['name', 'description', 'vulnerabilities'])
+ for statistics in self.details["hunter_statistics"]:
+ msg.add_row(
+ [statistics["name"],
+ textwrap.fill(statistics["description"], width=50),
+ statistics["vulnerabilities"]])
+ self.__logger.info("\n\n%s\n", msg.get_string())
+
+ def run(self, **kwargs):
+ super(KubeHunter, self).run(**kwargs)
+ try:
+ self.process_results(**kwargs)
+ except Exception: # pylint: disable=broad-except
+ self.__logger.exception("Cannot process results")
+ self.result = 0
+
class KubeBench(SecurityTesting):
"""kube-bench checks whether Kubernetes is deployed securelyself.
@@ -135,3 +191,7 @@ class KubeBench(SecurityTesting):
def __init__(self, **kwargs):
super(KubeBench, self).__init__(**kwargs)
self.job_name = "kube-bench"
+
+ def run(self, **kwargs):
+ super(KubeBench, self).run(**kwargs)
+ self.result = 100