[bradm] Initial nrpe checks

charm-helpers-hooks.yaml

@@ -8,3 +8,4 @@ include:
     - payload.execd
     - contrib.openstack.alternatives
     - contrib.network.ip
+    - contrib.charmsupport

config.yaml

@@ -155,3 +155,14 @@ options:
       order for this charm to function correctly, the privacy extension must be
       disabled and a non-temporary address must be configured/available on
       your network interface.
+  nagios_context:
+    default: "juju"
+    type: string
+    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.
+

files/nagios/check_ceph_status.py

@@ -0,0 +1,44 @@
+#!/usr/bin/env python
+
+# Copyright (C) 2014 Canonical
+# All Rights Reserved
+# Author: Jacek Nykis <jacek.nykis@canonical.com>
+
+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)

files/nagios/collect_ceph_status.sh

@@ -0,0 +1,18 @@
+#!/bin/bash
+# Copyright (C) 2014 Canonical
+# All Rights Reserved
+# Author: Jacek Nykis <jacek.nykis@canonical.com>
+
+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

hooks/charmhelpers/contrib/charmsupport/nrpe.py

@@ -0,0 +1,219 @@
+"""Compatibility with the nrpe-external-master charm"""
+# Copyright 2012 Canonical Ltd.
+#
+# Authors:
+#  Matthew Wedgwood <matthew.wedgwood@canonical.com>
+
+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,
+)
+
+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.
+#
+# 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):
+        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)
+
+    def write_service_config(self, nagios_context, hostname):
+        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))
+
+        templ_vars = {
+            'nagios_hostname': hostname,
+            'nagios_servicegroup': nagios_context,
+            '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']
+        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)
+            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))

hooks/charmhelpers/contrib/charmsupport/volumes.py

@@ -0,0 +1,156 @@
+'''
+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']

hooks/hooks.py

@@ -23,13 +23,16 @@ from charmhelpers.core.hookenv import (
     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 (
@@ -51,8 +54,15 @@ from utils import (
     assert_charm_supports_ipv6
 )
 
+from charmhelpers.contrib.charmsupport.nrpe 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
@@ -143,6 +153,9 @@ def config_changed():
                         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 = []
@@ -307,6 +320,37 @@ def start():
         ceph.start_osds(get_devices())
 
 
+@hooks.hook('nrpe-external-master-relation-joined')
+@hooks.hook('nrpe-external-master-relation-changed')
+def update_nrpe_config():
+    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 = None
+    for rel in relations_of_type('nrpe-external-master'):
+        if 'nagios_hostname' in rel:
+            hostname = rel['nagios_hostname']
+            break
+    nrpe = NRPE(hostname=hostname)
+    nrpe.add_check(
+        shortname="ceph",
+        description='Check Ceph health',
+        check_cmd='check_ceph_status.py -f {}'.format(STATUS_FILE)
+    )
+    nrpe.write()
+
+
 if __name__ == '__main__':
     try:
         hooks.execute(sys.argv)

hooks/nrpe-external-master-relation-changed

@@ -0,0 +1 @@
+hooks.py

hooks/nrpe-external-master-relation-joined

@@ -0,0 +1 @@
+hooks.py

metadata.yaml

@@ -19,3 +19,7 @@ provides:
     interface: ceph-osd
   radosgw:
     interface: ceph-radosgw
+  nrpe-external-master:
+    interface: nrpe-external-master
+    scope: container
+    gets: [nagios_hostname, nagios_host_context]