diff options
Diffstat (limited to 'roles/openshift_health_checker/openshift_checks/logging/logging.py')
| -rw-r--r-- | roles/openshift_health_checker/openshift_checks/logging/logging.py | 89 |
1 files changed, 51 insertions, 38 deletions
diff --git a/roles/openshift_health_checker/openshift_checks/logging/logging.py b/roles/openshift_health_checker/openshift_checks/logging/logging.py index 6e951e82c..ecd8adb64 100644 --- a/roles/openshift_health_checker/openshift_checks/logging/logging.py +++ b/roles/openshift_health_checker/openshift_checks/logging/logging.py @@ -5,49 +5,60 @@ Util functions for performing checks on an Elasticsearch, Fluentd, and Kibana st import json import os -from openshift_checks import OpenShiftCheck, OpenShiftCheckException, get_var +from openshift_checks import OpenShiftCheck, OpenShiftCheckException + + +class MissingComponentPods(OpenShiftCheckException): + """Raised when a component has no pods in the namespace.""" + pass + + +class CouldNotUseOc(OpenShiftCheckException): + """Raised when ocutil has a failure running oc.""" + pass class LoggingCheck(OpenShiftCheck): - """Base class for logging component checks""" + """Base class for OpenShift aggregated logging component checks""" + + # FIXME: this should not be listed as a check, since it is not meant to be + # run by itself. name = "logging" - @classmethod - def is_active(cls, task_vars): - return super(LoggingCheck, cls).is_active(task_vars) and cls.is_first_master(task_vars) + def is_active(self): + logging_deployed = self.get_var("openshift_hosted_logging_deploy", convert=bool, default=False) + return logging_deployed and super(LoggingCheck, self).is_active() and self.is_first_master() - @staticmethod - def is_first_master(task_vars): - """Run only on first master and only when logging is configured. Returns: bool""" - logging_deployed = get_var(task_vars, "openshift_hosted_logging_deploy", default=True) + def is_first_master(self): + """Determine if running on first master. Returns: bool""" # Note: It would be nice to use membership in oo_first_master group, however for now it # seems best to avoid requiring that setup and just check this is the first master. - hostname = get_var(task_vars, "ansible_ssh_host") or [None] - masters = get_var(task_vars, "groups", "masters", default=None) or [None] - return logging_deployed and masters[0] == hostname + hostname = self.get_var("ansible_ssh_host") or [None] + masters = self.get_var("groups", "masters", default=None) or [None] + return masters[0] == hostname - def run(self, tmp, task_vars): - pass + def run(self): + return {} - def get_pods_for_component(self, execute_module, namespace, logging_component, task_vars): - """Get all pods for a given component. Returns: list of pods for component, error string""" + def get_pods_for_component(self, logging_component): + """Get all pods for a given component. Returns: list of pods.""" pod_output = self.exec_oc( - execute_module, - namespace, "get pods -l component={} -o json".format(logging_component), [], - task_vars ) try: - pods = json.loads(pod_output) - if not pods or not pods.get('items'): + pods = json.loads(pod_output) # raises ValueError if deserialize fails + if not pods or not pods.get('items'): # also a broken response, treat the same raise ValueError() except ValueError: - # successful run but non-parsing data generally means there were no pods in the namespace - return None, 'There are no pods in the {} namespace. Is logging deployed?'.format(namespace) + # successful run but non-parsing data generally means there were no pods to be found + raise MissingComponentPods( + 'There are no "{}" component pods in the "{}" namespace.\n' + 'Is logging deployed?'.format(logging_component, self.logging_namespace()) + ) - return pods['items'], None + return pods['items'] @staticmethod def not_running_pods(pods): @@ -63,34 +74,36 @@ class LoggingCheck(OpenShiftCheck): ) ] - @staticmethod - def exec_oc(execute_module=None, namespace="logging", cmd_str="", extra_args=None, task_vars=None): + def logging_namespace(self): + """Returns the namespace in which logging is configured to deploy.""" + return self.get_var("openshift_logging_namespace", default="logging") + + def exec_oc(self, cmd_str="", extra_args=None): """ Execute an 'oc' command in the remote host. Returns: output of command and namespace, - or raises OpenShiftCheckException on error + or raises CouldNotUseOc on error """ - config_base = get_var(task_vars, "openshift", "common", "config_base") + config_base = self.get_var("openshift", "common", "config_base") args = { - "namespace": namespace, + "namespace": self.logging_namespace(), "config_file": os.path.join(config_base, "master", "admin.kubeconfig"), "cmd": cmd_str, "extra_args": list(extra_args) if extra_args else [], } - result = execute_module("ocutil", args, None, task_vars) + result = self.execute_module("ocutil", args) if result.get("failed"): - msg = ( - 'Unexpected error using `oc` to validate the logging stack components.\n' - 'Error executing `oc {cmd}`:\n' - '{error}' - ).format(cmd=args['cmd'], error=result['result']) - if result['result'] == '[Errno 2] No such file or directory': - msg = ( + raise CouldNotUseOc( "This host is supposed to be a master but does not have the `oc` command where expected.\n" "Has an installation been run on this host yet?" ) - raise OpenShiftCheckException(msg) + + raise CouldNotUseOc( + 'Unexpected error using `oc` to validate the logging stack components.\n' + 'Error executing `oc {cmd}`:\n' + '{error}'.format(cmd=args['cmd'], error=result['result']) + ) return result.get("result", "") |