Merge lp:~gnuoy/charms/trusty/ceph/add-nrpe-checks into lp:~openstack-charmers-archive/charms/trusty/ceph/next

[Liam Young]

Add nrpe support. Based on branch from bradm with a few tweaks

[uosci-testing-bot (uosci-testing-bot)]

charm_unit_test #705 ceph-next for gnuoy mp246155
    UNIT OK: passed

Build: http://10.245.162.77:8080/job/charm_unit_test/705/

[uosci-testing-bot (uosci-testing-bot)]

charm_lint_check #676 ceph-next for gnuoy mp246155
    LINT OK: passed

Build: http://10.245.162.77:8080/job/charm_lint_check/676/

[uosci-testing-bot (uosci-testing-bot)]

charm_amulet_test #861 ceph-next for gnuoy mp246155
    AMULET OK: passed

Build: http://10.245.162.77:8080/job/charm_amulet_test/861/

[Liam Young (gnuoy)]

<jamespage> gnuoy, as they are re-syncs + tweaks to the nrpe stuff in the charms, I'm happy to give a conditional +1 across the board based on osci checking things out OK
<gnuoy> jamespage, I'll take that! thanks
...
<gnuoy> jamespage, osci is still working through. But on the subject of those mps, does your +1 stand for branches with no amulet tests?
<jamespage> gnuoy, yes

review: Approve

=== modified file 'charm-helpers-hooks.yaml'
--- charm-helpers-hooks.yaml	2014-12-11 16:46:45 +0000
+++ charm-helpers-hooks.yaml	2015-01-12 14:00:31 +0000
@@ -9,3 +9,4 @@
     - payload.execd
     - contrib.openstack.alternatives
     - contrib.network.ip
+    - contrib.charmsupport
 
=== modified file 'config.yaml'
--- config.yaml	2014-11-25 18:29:07 +0000
+++ config.yaml	2015-01-12 14:00:31 +0000
@@ -161,3 +161,12 @@
     description: |
       YAML-formatted associative array of sysctl key/value pairs to be set
       persistently e.g. '{ kernel.pid_max : 4194303 }'.
+  nagios_context:
+    default: "juju"
+    description: |
+      Used by the nrpe-external-master subordinate charm.
+      A string that will be prepended to instance name to set the host name
+      in nagios. So for instance the hostname would be something like:
+          juju-myservice-0
+      If you're running multiple environments with the same services in them
+      this allows you to differentiate between them.
 
=== added directory 'files/nagios'
=== added file 'files/nagios/check_ceph_status.py'
--- files/nagios/check_ceph_status.py	1970-01-01 00:00:00 +0000
+++ files/nagios/check_ceph_status.py	2015-01-12 14:00:31 +0000
@@ -0,0 +1,44 @@
+#!/usr/bin/env python
+
+# Copyright (C) 2014 Canonical
+# All Rights Reserved
+# Author: Jacek Nykis <[e-mail]>
+
+import re
+import argparse
+import subprocess
+import nagios_plugin
+
+
+def check_ceph_status(args):
+    if args.status_file:
+        nagios_plugin.check_file_freshness(args.status_file, 3600)
+        with open(args.status_file, "r") as f:
+            lines = f.readlines()
+            status_data = dict(l.strip().split(' ', 1) for l in lines if len(l) > 1)
+    else:
+        lines = subprocess.check_output(["ceph", "status"]).split('\n')
+        status_data = dict(l.strip().split(' ', 1) for l in lines if len(l) > 1)
+
+    if ('health' not in status_data
+            or 'monmap' not in status_data
+            or 'osdmap'not in status_data):
+        raise nagios_plugin.UnknownError('UNKNOWN: status data is incomplete')
+
+    if status_data['health'] != 'HEALTH_OK':
+        msg = 'CRITICAL: ceph health status: "{}"'.format(status_data['health'])
+        raise nagios_plugin.CriticalError(msg)
+    osds = re.search("^.*: (\d+) osds: (\d+) up, (\d+) in", status_data['osdmap'])
+    if osds.group(1) > osds.group(2):  # not all OSDs are "up"
+        msg = 'CRITICAL: Some OSDs are not up. Total: {}, up: {}'.format(
+            osds.group(1), osds.group(2))
+        raise nagios_plugin.CriticalError(msg)
+    print "All OK"
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description='Check ceph status')
+    parser.add_argument('-f', '--file', dest='status_file',
+                        default=False, help='Optional file with "ceph status" output')
+    args = parser.parse_args()
+    nagios_plugin.try_check(check_ceph_status, args)
 
=== added file 'files/nagios/collect_ceph_status.sh'
--- files/nagios/collect_ceph_status.sh	1970-01-01 00:00:00 +0000
+++ files/nagios/collect_ceph_status.sh	2015-01-12 14:00:31 +0000
@@ -0,0 +1,18 @@
+#!/bin/bash
+# Copyright (C) 2014 Canonical
+# All Rights Reserved
+# Author: Jacek Nykis <[e-mail]>
+
+LOCK=/var/lock/ceph-status.lock
+lockfile-create -r2 --lock-name $LOCK > /dev/null 2>&1
+if [ $? -ne 0 ]; then
+    exit 1
+fi
+trap "rm -f $LOCK > /dev/null 2>&1" exit
+
+DATA_DIR="/var/lib/nagios"
+if [ ! -d $DATA_DIR ]; then
+    mkdir -p $DATA_DIR
+fi
+
+ceph status >${DATA_DIR}/cat-ceph-status.txt
 
=== added directory 'hooks/charmhelpers/contrib/charmsupport'
=== added file 'hooks/charmhelpers/contrib/charmsupport/__init__.py'
=== added file 'hooks/charmhelpers/contrib/charmsupport/nrpe.py'
--- hooks/charmhelpers/contrib/charmsupport/nrpe.py	1970-01-01 00:00:00 +0000
+++ hooks/charmhelpers/contrib/charmsupport/nrpe.py	2015-01-12 14:00:31 +0000
@@ -0,0 +1,308 @@
+"""Compatibility with the nrpe-external-master charm"""
+# Copyright 2012 Canonical Ltd.
+#
+# Authors:
+#  Matthew Wedgwood <[e-mail]>
+
+import subprocess
+import pwd
+import grp
+import os
+import re
+import shlex
+import yaml
+
+from charmhelpers.core.hookenv import (
+    config,
+    local_unit,
+    log,
+    relation_ids,
+    relation_set,
+    relations_of_type,
+)
+
+from charmhelpers.core.host import service
+
+# This module adds compatibility with the nrpe-external-master and plain nrpe
+# subordinate charms. To use it in your charm:
+#
+# 1. Update metadata.yaml
+#
+#   provides:
+#     (...)
+#     nrpe-external-master:
+#       interface: nrpe-external-master
+#       scope: container
+#
+#   and/or
+#
+#   provides:
+#     (...)
+#     local-monitors:
+#       interface: local-monitors
+#       scope: container
+
+#
+# 2. Add the following to config.yaml
+#
+#    nagios_context:
+#      default: "juju"
+#      type: string
+#      description: |
+#        Used by the nrpe subordinate charms.
+#        A string that will be prepended to instance name to set the host name
+#        in nagios. So for instance the hostname would be something like:
+#            juju-myservice-0
+#        If you're running multiple environments with the same services in them
+#        this allows you to differentiate between them.
+#    nagios_servicegroups:
+#      default: ""
+#      type: string
+#      description: |
+#        A comma-separated list of nagios servicegroups.
+#        If left empty, the nagios_context will be used as the servicegroup
+#
+# 3. Add custom checks (Nagios plugins) to files/nrpe-external-master
+#
+# 4. Update your hooks.py with something like this:
+#
+#    from charmsupport.nrpe import NRPE
+#    (...)
+#    def update_nrpe_config():
+#        nrpe_compat = NRPE()
+#        nrpe_compat.add_check(
+#            shortname = "myservice",
+#            description = "Check MyService",
+#            check_cmd = "check_http -w 2 -c 10 http://localhost"
+#            )
+#        nrpe_compat.add_check(
+#            "myservice_other",
+#            "Check for widget failures",
+#            check_cmd = "/srv/myapp/scripts/widget_check"
+#            )
+#        nrpe_compat.write()
+#
+#    def config_changed():
+#        (...)
+#        update_nrpe_config()
+#
+#    def nrpe_external_master_relation_changed():
+#        update_nrpe_config()
+#
+#    def local_monitors_relation_changed():
+#        update_nrpe_config()
+#
+# 5. ln -s hooks.py nrpe-external-master-relation-changed
+#    ln -s hooks.py local-monitors-relation-changed
+
+
+class CheckException(Exception):
+    pass
+
+
+class Check(object):
+    shortname_re = '[A-Za-z0-9-_]+$'
+    service_template = ("""
+#---------------------------------------------------
+# This file is Juju managed
+#---------------------------------------------------
+define service {{
+    use                             active-service
+    host_name                       {nagios_hostname}
+    service_description             {nagios_hostname}[{shortname}] """
+                        """{description}
+    check_command                   check_nrpe!{command}
+    servicegroups                   {nagios_servicegroup}
+}}
+""")
+
+    def __init__(self, shortname, description, check_cmd):
+        super(Check, self).__init__()
+        # XXX: could be better to calculate this from the service name
+        if not re.match(self.shortname_re, shortname):
+            raise CheckException("shortname must match {}".format(
+                Check.shortname_re))
+        self.shortname = shortname
+        self.command = "check_{}".format(shortname)
+        # Note: a set of invalid characters is defined by the
+        # Nagios server config
+        # The default is: illegal_object_name_chars=`~!$%^&*"|'<>?,()=
+        self.description = description
+        self.check_cmd = self._locate_cmd(check_cmd)
+
+    def _locate_cmd(self, check_cmd):
+        search_path = (
+            '/usr/lib/nagios/plugins',
+            '/usr/local/lib/nagios/plugins',
+        )
+        parts = shlex.split(check_cmd)
+        for path in search_path:
+            if os.path.exists(os.path.join(path, parts[0])):
+                command = os.path.join(path, parts[0])
+                if len(parts) > 1:
+                    command += " " + " ".join(parts[1:])
+                return command
+        log('Check command not found: {}'.format(parts[0]))
+        return ''
+
+    def write(self, nagios_context, hostname, nagios_servicegroups=None):
+        nrpe_check_file = '/etc/nagios/nrpe.d/{}.cfg'.format(
+            self.command)
+        with open(nrpe_check_file, 'w') as nrpe_check_config:
+            nrpe_check_config.write("# check {}\n".format(self.shortname))
+            nrpe_check_config.write("command[{}]={}\n".format(
+                self.command, self.check_cmd))
+
+        if not os.path.exists(NRPE.nagios_exportdir):
+            log('Not writing service config as {} is not accessible'.format(
+                NRPE.nagios_exportdir))
+        else:
+            self.write_service_config(nagios_context, hostname,
+                                      nagios_servicegroups)
+
+    def write_service_config(self, nagios_context, hostname,
+                             nagios_servicegroups=None):
+        for f in os.listdir(NRPE.nagios_exportdir):
+            if re.search('.*{}.cfg'.format(self.command), f):
+                os.remove(os.path.join(NRPE.nagios_exportdir, f))
+
+        if not nagios_servicegroups:
+            nagios_servicegroups = nagios_context
+
+        templ_vars = {
+            'nagios_hostname': hostname,
+            'nagios_servicegroup': nagios_servicegroups,
+            'description': self.description,
+            'shortname': self.shortname,
+            'command': self.command,
+        }
+        nrpe_service_text = Check.service_template.format(**templ_vars)
+        nrpe_service_file = '{}/service__{}_{}.cfg'.format(
+            NRPE.nagios_exportdir, hostname, self.command)
+        with open(nrpe_service_file, 'w') as nrpe_service_config:
+            nrpe_service_config.write(str(nrpe_service_text))
+
+    def run(self):
+        subprocess.call(self.check_cmd)
+
+
+class NRPE(object):
+    nagios_logdir = '/var/log/nagios'
+    nagios_exportdir = '/var/lib/nagios/export'
+    nrpe_confdir = '/etc/nagios/nrpe.d'
+
+    def __init__(self, hostname=None):
+        super(NRPE, self).__init__()
+        self.config = config()
+        self.nagios_context = self.config['nagios_context']
+        if 'nagios_servicegroups' in self.config:
+            self.nagios_servicegroups = self.config['nagios_servicegroups']
+        else:
+            self.nagios_servicegroups = 'juju'
+        self.unit_name = local_unit().replace('/', '-')
+        if hostname:
+            self.hostname = hostname
+        else:
+            self.hostname = "{}-{}".format(self.nagios_context, self.unit_name)
+        self.checks = []
+
+    def add_check(self, *args, **kwargs):
+        self.checks.append(Check(*args, **kwargs))
+
+    def write(self):
+        try:
+            nagios_uid = pwd.getpwnam('nagios').pw_uid
+            nagios_gid = grp.getgrnam('nagios').gr_gid
+        except:
+            log("Nagios user not set up, nrpe checks not updated")
+            return
+
+        if not os.path.exists(NRPE.nagios_logdir):
+            os.mkdir(NRPE.nagios_logdir)
+            os.chown(NRPE.nagios_logdir, nagios_uid, nagios_gid)
+
+        nrpe_monitors = {}
+        monitors = {"monitors": {"remote": {"nrpe": nrpe_monitors}}}
+        for nrpecheck in self.checks:
+            nrpecheck.write(self.nagios_context, self.hostname,
+                            self.nagios_servicegroups)
+            nrpe_monitors[nrpecheck.shortname] = {
+                "command": nrpecheck.command,
+            }
+
+        service('restart', 'nagios-nrpe-server')
+
+        for rid in relation_ids("local-monitors"):
+            relation_set(relation_id=rid, monitors=yaml.dump(monitors))
+
+
+def get_nagios_hostcontext(relation_name='nrpe-external-master'):
+    """
+    Query relation with nrpe subordinate, return the nagios_host_context
+
+    :param str relation_name: Name of relation nrpe sub joined to
+    """
+    for rel in relations_of_type(relation_name):
+        if 'nagios_hostname' in rel:
+            return rel['nagios_host_context']
+
+
+def get_nagios_hostname(relation_name='nrpe-external-master'):
+    """
+    Query relation with nrpe subordinate, return the nagios_hostname
+
+    :param str relation_name: Name of relation nrpe sub joined to
+    """
+    for rel in relations_of_type(relation_name):
+        if 'nagios_hostname' in rel:
+            return rel['nagios_hostname']
+
+
+def get_nagios_unit_name(relation_name='nrpe-external-master'):
+    """
+    Return the nagios unit name prepended with host_context if needed
+
+    :param str relation_name: Name of relation nrpe sub joined to
+    """
+    host_context = get_nagios_hostcontext(relation_name)
+    if host_context:
+        unit = "%s:%s" % (host_context, local_unit())
+    else:
+        unit = local_unit()
+    return unit
+
+
+def add_init_service_checks(nrpe, services, unit_name):
+    """
+    Add checks for each service in list
+
+    :param NRPE nrpe: NRPE object to add check to
+    :param list services: List of services to check
+    :param str unit_name: Unit name to use in check description
+    """
+    for svc in services:
+        upstart_init = '/etc/init/%s.conf' % svc
+        sysv_init = '/etc/init.d/%s' % svc
+        if os.path.exists(upstart_init):
+            nrpe.add_check(
+                shortname=svc,
+                description='process check {%s}' % unit_name,
+                check_cmd='check_upstart_job %s' % svc
+            )
+        elif os.path.exists(sysv_init):
+            cronpath = '/etc/cron.d/nagios-service-check-%s' % svc
+            cron_file = ('*/5 * * * * root '
+                         '/usr/local/lib/nagios/plugins/check_exit_status.pl '
+                         '-s /etc/init.d/%s status > '
+                         '/var/lib/nagios/service-check-%s.txt\n' % (svc,
+                                                                     svc)
+                         )
+            f = open(cronpath, 'w')
+            f.write(cron_file)
+            f.close()
+            nrpe.add_check(
+                shortname=svc,
+                description='process check {%s}' % unit_name,
+                check_cmd='check_status_file.py -f '
+                          '/var/lib/nagios/service-check-%s.txt' % svc,
+            )
 
=== added file 'hooks/charmhelpers/contrib/charmsupport/volumes.py'
--- hooks/charmhelpers/contrib/charmsupport/volumes.py	1970-01-01 00:00:00 +0000
+++ hooks/charmhelpers/contrib/charmsupport/volumes.py	2015-01-12 14:00:31 +0000
@@ -0,0 +1,159 @@
+'''
+Functions for managing volumes in juju units. One volume is supported per unit.
+Subordinates may have their own storage, provided it is on its own partition.
+
+Configuration stanzas::
+
+  volume-ephemeral:
+    type: boolean
+    default: true
+    description: >
+      If false, a volume is mounted as sepecified in "volume-map"
+      If true, ephemeral storage will be used, meaning that log data
+         will only exist as long as the machine. YOU HAVE BEEN WARNED.
+  volume-map:
+    type: string
+    default: {}
+    description: >
+      YAML map of units to device names, e.g:
+        "{ rsyslog/0: /dev/vdb, rsyslog/1: /dev/vdb }"
+      Service units will raise a configure-error if volume-ephemeral
+      is 'true' and no volume-map value is set. Use 'juju set' to set a
+      value and 'juju resolved' to complete configuration.
+
+Usage::
+
+    from charmsupport.volumes import configure_volume, VolumeConfigurationError
+    from charmsupport.hookenv import log, ERROR
+    def post_mount_hook():
+        stop_service('myservice')
+    def post_mount_hook():
+        start_service('myservice')
+
+    if __name__ == '__main__':
+        try:
+            configure_volume(before_change=pre_mount_hook,
+                             after_change=post_mount_hook)
+        except VolumeConfigurationError:
+            log('Storage could not be configured', ERROR)
+
+'''
+
+# XXX: Known limitations
+# - fstab is neither consulted nor updated
+
+import os
+from charmhelpers.core import hookenv
+from charmhelpers.core import host
+import yaml
+
+
+MOUNT_BASE = '/srv/juju/volumes'
+
+
+class VolumeConfigurationError(Exception):
+    '''Volume configuration data is missing or invalid'''
+    pass
+
+
+def get_config():
+    '''Gather and sanity-check volume configuration data'''
+    volume_config = {}
+    config = hookenv.config()
+
+    errors = False
+
+    if config.get('volume-ephemeral') in (True, 'True', 'true', 'Yes', 'yes'):
+        volume_config['ephemeral'] = True
+    else:
+        volume_config['ephemeral'] = False
+
+    try:
+        volume_map = yaml.safe_load(config.get('volume-map', '{}'))
+    except yaml.YAMLError as e:
+        hookenv.log("Error parsing YAML volume-map: {}".format(e),
+                    hookenv.ERROR)
+        errors = True
+    if volume_map is None:
+        # probably an empty string
+        volume_map = {}
+    elif not isinstance(volume_map, dict):
+        hookenv.log("Volume-map should be a dictionary, not {}".format(
+            type(volume_map)))
+        errors = True
+
+    volume_config['device'] = volume_map.get(os.environ['JUJU_UNIT_NAME'])
+    if volume_config['device'] and volume_config['ephemeral']:
+        # asked for ephemeral storage but also defined a volume ID
+        hookenv.log('A volume is defined for this unit, but ephemeral '
+                    'storage was requested', hookenv.ERROR)
+        errors = True
+    elif not volume_config['device'] and not volume_config['ephemeral']:
+        # asked for permanent storage but did not define volume ID
+        hookenv.log('Ephemeral storage was requested, but there is no volume '
+                    'defined for this unit.', hookenv.ERROR)
+        errors = True
+
+    unit_mount_name = hookenv.local_unit().replace('/', '-')
+    volume_config['mountpoint'] = os.path.join(MOUNT_BASE, unit_mount_name)
+
+    if errors:
+        return None
+    return volume_config
+
+
+def mount_volume(config):
+    if os.path.exists(config['mountpoint']):
+        if not os.path.isdir(config['mountpoint']):
+            hookenv.log('Not a directory: {}'.format(config['mountpoint']))
+            raise VolumeConfigurationError()
+    else:
+        host.mkdir(config['mountpoint'])
+    if os.path.ismount(config['mountpoint']):
+        unmount_volume(config)
+    if not host.mount(config['device'], config['mountpoint'], persist=True):
+        raise VolumeConfigurationError()
+
+
+def unmount_volume(config):
+    if os.path.ismount(config['mountpoint']):
+        if not host.umount(config['mountpoint'], persist=True):
+            raise VolumeConfigurationError()
+
+
+def managed_mounts():
+    '''List of all mounted managed volumes'''
+    return filter(lambda mount: mount[0].startswith(MOUNT_BASE), host.mounts())
+
+
+def configure_volume(before_change=lambda: None, after_change=lambda: None):
+    '''Set up storage (or don't) according to the charm's volume configuration.
+       Returns the mount point or "ephemeral". before_change and after_change
+       are optional functions to be called if the volume configuration changes.
+    '''
+
+    config = get_config()
+    if not config:
+        hookenv.log('Failed to read volume configuration', hookenv.CRITICAL)
+        raise VolumeConfigurationError()
+
+    if config['ephemeral']:
+        if os.path.ismount(config['mountpoint']):
+            before_change()
+            unmount_volume(config)
+            after_change()
+        return 'ephemeral'
+    else:
+        # persistent storage
+        if os.path.ismount(config['mountpoint']):
+            mounts = dict(managed_mounts())
+            if mounts.get(config['mountpoint']) != config['device']:
+                before_change()
+                unmount_volume(config)
+                mount_volume(config)
+                after_change()
+        else:
+            before_change()
+            mount_volume(config)
+            after_change()
+        return config['mountpoint']
 
=== modified file 'hooks/charmhelpers/contrib/storage/linux/ceph.py'
--- hooks/charmhelpers/contrib/storage/linux/ceph.py	2014-11-26 09:07:27 +0000
+++ hooks/charmhelpers/contrib/storage/linux/ceph.py	2015-01-12 14:00:31 +0000
@@ -372,3 +372,46 @@
             return None
     else:
         return None
+
+
+class CephBrokerRq(object):
+    """Ceph broker request.
+
+    Multiple operations can be added to a request and sent to the Ceph broker
+    to be executed.
+
+    Request is json-encoded for sending over the wire.
+
+    The API is versioned and defaults to version 1.
+    """
+    def __init__(self, api_version=1):
+        self.api_version = api_version
+        self.ops = []
+
+    def add_op_create_pool(self, name, replica_count=3):
+        self.ops.append({'op': 'create-pool', 'name': name,
+                         'replicas': replica_count})
+
+    @property
+    def request(self):
+        return json.dumps({'api-version': self.api_version, 'ops': self.ops})
+
+
+class CephBrokerRsp(object):
+    """Ceph broker response.
+
+    Response is json-decoded and contents provided as methods/properties.
+
+    The API is versioned and defaults to version 1.
+    """
+    def __init__(self, encoded_rsp):
+        self.api_version = None
+        self.rsp = json.loads(encoded_rsp)
+
+    @property
+    def exit_code(self):
+        return self.rsp.get('exit-code')
+
+    @property
+    def exit_msg(self):
+        return self.rsp.get('stderr')
 
=== added file 'hooks/charmhelpers/core/decorators.py'
--- hooks/charmhelpers/core/decorators.py	1970-01-01 00:00:00 +0000
+++ hooks/charmhelpers/core/decorators.py	2015-01-12 14:00:31 +0000
@@ -0,0 +1,41 @@
+#
+# Copyright 2014 Canonical Ltd.
+#
+# Authors:
+#  Edward Hope-Morley <[e-mail]>
+#
+
+import time
+
+from charmhelpers.core.hookenv import (
+    log,
+    INFO,
+)
+
+
+def retry_on_exception(num_retries, base_delay=0, exc_type=Exception):
+    """If the decorated function raises exception exc_type, allow num_retries
+    retry attempts before raise the exception.
+    """
+    def _retry_on_exception_inner_1(f):
+        def _retry_on_exception_inner_2(*args, **kwargs):
+            retries = num_retries
+            multiplier = 1
+            while True:
+                try:
+                    return f(*args, **kwargs)
+                except exc_type:
+                    if not retries:
+                        raise
+
+                delay = base_delay * multiplier
+                multiplier += 1
+                log("Retrying '%s' %d more times (delay=%s)" %
+                    (f.__name__, retries, delay), level=INFO)
+                retries -= 1
+                if delay:
+                    time.sleep(delay)
+
+        return _retry_on_exception_inner_2
+
+    return _retry_on_exception_inner_1
 
=== modified file 'hooks/charmhelpers/core/host.py'
--- hooks/charmhelpers/core/host.py	2014-12-11 16:47:24 +0000
+++ hooks/charmhelpers/core/host.py	2015-01-12 14:00:31 +0000
@@ -162,13 +162,16 @@
     uid = pwd.getpwnam(owner).pw_uid
     gid = grp.getgrnam(group).gr_gid
     realpath = os.path.abspath(path)
-    if os.path.exists(realpath):
-        if force and not os.path.isdir(realpath):
+    path_exists = os.path.exists(realpath)
+    if path_exists and force:
+        if not os.path.isdir(realpath):
             log("Removing non-directory file {} prior to mkdir()".format(path))
             os.unlink(realpath)
-    else:
+            os.makedirs(realpath, perms)
+        os.chown(realpath, uid, gid)
+    elif not path_exists:
         os.makedirs(realpath, perms)
-    os.chown(realpath, uid, gid)
+        os.chown(realpath, uid, gid)
 
 
 def write_file(path, content, owner='root', group='root', perms=0o444):
 
=== modified file 'hooks/charmhelpers/fetch/__init__.py'
--- hooks/charmhelpers/fetch/__init__.py	2014-11-25 17:07:46 +0000
+++ hooks/charmhelpers/fetch/__init__.py	2015-01-12 14:00:31 +0000
@@ -64,9 +64,16 @@
     'trusty-juno/updates': 'trusty-updates/juno',
     'trusty-updates/juno': 'trusty-updates/juno',
     'juno/proposed': 'trusty-proposed/juno',
-    'juno/proposed': 'trusty-proposed/juno',
     'trusty-juno/proposed': 'trusty-proposed/juno',
     'trusty-proposed/juno': 'trusty-proposed/juno',
+    # Kilo
+    'kilo': 'trusty-updates/kilo',
+    'trusty-kilo': 'trusty-updates/kilo',
+    'trusty-kilo/updates': 'trusty-updates/kilo',
+    'trusty-updates/kilo': 'trusty-updates/kilo',
+    'kilo/proposed': 'trusty-proposed/kilo',
+    'trusty-kilo/proposed': 'trusty-proposed/kilo',
+    'trusty-proposed/kilo': 'trusty-proposed/kilo',
 }
 
 # The order of this list is very important. Handlers should be listed in from
 
=== modified file 'hooks/hooks.py'
--- hooks/hooks.py	2014-11-25 18:29:07 +0000
+++ hooks/hooks.py	2015-01-12 14:00:31 +0000
@@ -25,12 +25,15 @@
     relation_set,
     remote_unit,
     Hooks, UnregisteredHookError,
-    service_name
+    service_name,
+    relations_of_type
 )
 from charmhelpers.core.host import (
     service_restart,
     umount,
     mkdir,
+    write_file,
+    rsync,
     cmp_pkgrevno
 )
 from charmhelpers.fetch import (
@@ -56,8 +59,15 @@
     process_requests
 )
 
+from charmhelpers.contrib.charmsupport import nrpe
+
 hooks = Hooks()
 
+NAGIOS_PLUGINS = '/usr/local/lib/nagios/plugins'
+SCRIPTS_DIR = '/usr/local/bin'
+STATUS_FILE = '/var/lib/nagios/cat-ceph-status.txt'
+STATUS_CRONFILE = '/etc/cron.d/cat-ceph-health'
+
 
 def install_upstart_scripts():
     # Only install upstart configurations for older versions
@@ -152,6 +162,9 @@
                         reformat_osd(), config('ignore-device-errors'))
         ceph.start_osds(get_devices())
 
+    if relations_of_type('nrpe-external-master'):
+        update_nrpe_config()
+
 
 def get_mon_hosts():
     hosts = []
@@ -334,6 +347,36 @@
         ceph.start_osds(get_devices())
 
 
[e-mail]('nrpe-external-master-relation-joined')
[e-mail]('nrpe-external-master-relation-changed')
+def update_nrpe_config():
+    # python-dbus is used by check_upstart_job
+    apt_install('python-dbus')
+    log('Refreshing nagios checks')
+    if os.path.isdir(NAGIOS_PLUGINS):
+        rsync(os.path.join(os.getenv('CHARM_DIR'), 'files', 'nagios',
+                           'check_ceph_status.py'),
+              os.path.join(NAGIOS_PLUGINS, 'check_ceph_status.py'))
+
+    script = os.path.join(SCRIPTS_DIR, 'collect_ceph_status.sh')
+    rsync(os.path.join(os.getenv('CHARM_DIR'), 'files',
+                       'nagios', 'collect_ceph_status.sh'),
+          script)
+    cronjob = "{} root {}\n".format('*/5 * * * *', script)
+    write_file(STATUS_CRONFILE, cronjob)
+
+    # Find out if nrpe set nagios_hostname
+    hostname = nrpe.get_nagios_hostname()
+    current_unit = nrpe.get_nagios_unit_name()
+    nrpe_setup = nrpe.NRPE(hostname=hostname)
+    nrpe_setup.add_check(
+        shortname="ceph",
+        description='Check Ceph health {%s}' % current_unit,
+        check_cmd='check_ceph_status.py -f {}'.format(STATUS_FILE)
+    )
+    nrpe_setup.write()
+
+
 if __name__ == '__main__':
     try:
         hooks.execute(sys.argv)
 
=== added symlink 'hooks/nrpe-external-master-relation-changed'
=== target is u'hooks.py'
=== added symlink 'hooks/nrpe-external-master-relation-joined'
=== target is u'hooks.py'
=== modified file 'metadata.yaml'
--- metadata.yaml	2013-07-14 19:46:24 +0000
+++ metadata.yaml	2015-01-12 14:00:31 +0000
@@ -10,9 +10,16 @@
   mon:
     interface: ceph
 provides:
+  nrpe-external-master:
+    interface: nrpe-external-master
+    scope: container
   client:
     interface: ceph-client
   osd:
     interface: ceph-osd
   radosgw:
     interface: ceph-radosgw
+  nrpe-external-master:
+    interface: nrpe-external-master
+    scope: container
+    gets: [nagios_hostname, nagios_host_context]