Adding upstream version 1.37.1.upstream/1.37.1upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

4 files changed, 553 insertions, 0 deletions

diff --git a/collectors/python.d.plugin/hpssa/Makefile.inc b/collectors/python.d.plugin/hpssa/Makefile.inc
new file mode 100644
index 0000000..1c04aa4
--- /dev/null
+++ b/collectors/python.d.plugin/hpssa/Makefile.inc
@@ -0,0 +1,13 @@
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+# THIS IS NOT A COMPLETE Makefile
+# IT IS INCLUDED BY ITS PARENT'S Makefile.am
+# IT IS REQUIRED TO REFERENCE ALL FILES RELATIVE TO THE PARENT
+
+# install these files
+dist_python_DATA       += hpssa/hpssa.chart.py
+dist_pythonconfig_DATA += hpssa/hpssa.conf
+
+# do not install these files, but include them in the distribution
+dist_noinst_DATA       += hpssa/README.md hpssa/Makefile.inc
+
diff --git a/collectors/python.d.plugin/hpssa/README.md b/collectors/python.d.plugin/hpssa/README.md
new file mode 100644
index 0000000..c1d2182
--- /dev/null
+++ b/collectors/python.d.plugin/hpssa/README.md
@@ -0,0 +1,83 @@
+<!--
+title: "HP Smart Storage Arrays monitoring with Netdata"
+custom_edit_url: https://github.com/netdata/netdata/edit/master/collectors/python.d.plugin/hpssa/README.md
+sidebar_label: "HP Smart Storage Arrays"
+-->
+
+# HP Smart Storage Arrays monitoring with Netdata
+
+Monitors controller, cache module, logical and physical drive state and temperature using `ssacli` tool.
+
+Executed commands:
+
+- `sudo -n ssacli ctrl all show config detail`
+
+## Requirements:
+
+This module uses `ssacli`, which can only be executed by root. It uses
+`sudo` and assumes that it is configured such that the `netdata` user can execute `ssacli` as root without a password.
+
+- Add to your `/etc/sudoers` file:
+
+`which ssacli` shows the full path to the binary.
+
+```bash
+netdata ALL=(root)       NOPASSWD: /path/to/ssacli
+```
+
+- Reset Netdata's systemd
+  unit [CapabilityBoundingSet](https://www.freedesktop.org/software/systemd/man/systemd.exec.html#Capabilities) (Linux
+  distributions with systemd)
+
+The default CapabilityBoundingSet doesn't allow using `sudo`, and is quite strict in general. Resetting is not optimal, but a next-best solution given the inability to execute `ssacli` using `sudo`.
+
+As the `root` user, do the following:
+
+```cmd
+mkdir /etc/systemd/system/netdata.service.d
+echo -e '[Service]\nCapabilityBoundingSet=~' | tee /etc/systemd/system/netdata.service.d/unset-capability-bounding-set.conf
+systemctl daemon-reload
+systemctl restart netdata.service
+```
+
+## Charts
+
+- Controller status
+- Controller temperature
+- Logical drive status
+- Physical drive status
+- Physical drive temperature
+
+## Enable the collector
+
+The `hpssa` collector is disabled by default. To enable it, use `edit-config` from the
+Netdata [config directory](/docs/configure/nodes.md), which is typically at `/etc/netdata`, to edit the `python.d.conf`
+file.
+
+```bash
+cd /etc/netdata   # Replace this path with your Netdata config directory, if different
+sudo ./edit-config python.d.conf
+```
+
+Change the value of the `hpssa` setting to `yes`. Save the file and restart the Netdata Agent with `sudo systemctl
+restart netdata`, or the [appropriate method](/docs/configure/start-stop-restart.md) for your system.
+
+## Configuration
+
+Edit the `python.d/hpssa.conf` configuration file using `edit-config` from the
+Netdata [config directory](/docs/configure/nodes.md), which is typically at `/etc/netdata`.
+
+```bash
+cd /etc/netdata   # Replace this path with your Netdata config directory, if different
+sudo ./edit-config python.d/hpssa.conf
+```
+
+If `ssacli` cannot be found in the `PATH`, configure it in `hpssa.conf`.
+
+```yaml
+ssacli_path: /usr/sbin/ssacli
+```
+
+Save the file and restart the Netdata Agent with `sudo systemctl restart netdata`, or the [appropriate
+method](/docs/configure/start-stop-restart.md) for your system.
+
diff --git a/collectors/python.d.plugin/hpssa/hpssa.chart.py b/collectors/python.d.plugin/hpssa/hpssa.chart.py
new file mode 100644
index 0000000..4da73dc
--- /dev/null
+++ b/collectors/python.d.plugin/hpssa/hpssa.chart.py
@@ -0,0 +1,396 @@
+# -*- coding: utf-8 -*-
+# Description: hpssa netdata python.d module
+# Author: Peter Gnodde (gnoddep)
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+import os
+import re
+from copy import deepcopy
+
+from bases.FrameworkServices.ExecutableService import ExecutableService
+from bases.collection import find_binary
+
+disabled_by_default = True
+update_every = 5
+
+ORDER = [
+    'ctrl_status',
+    'ctrl_temperature',
+    'ld_status',
+    'pd_status',
+    'pd_temperature',
+]
+
+CHARTS = {
+    'ctrl_status': {
+        'options': [
+            None,
+            'Status 1 is OK, Status 0 is not OK',
+            'Status',
+            'Controller',
+            'hpssa.ctrl_status',
+            'line'
+        ],
+        'lines': []
+    },
+    'ctrl_temperature': {
+        'options': [
+            None,
+            'Temperature',
+            'Celsius',
+            'Controller',
+            'hpssa.ctrl_temperature',
+            'line'
+        ],
+        'lines': []
+    },
+    'ld_status': {
+        'options': [
+            None,
+            'Status 1 is OK, Status 0 is not OK',
+            'Status',
+            'Logical drives',
+            'hpssa.ld_status',
+            'line'
+        ],
+        'lines': []
+    },
+    'pd_status': {
+        'options': [
+            None,
+            'Status 1 is OK, Status 0 is not OK',
+            'Status',
+            'Physical drives',
+            'hpssa.pd_status',
+            'line'
+        ],
+        'lines': []
+    },
+    'pd_temperature': {
+        'options': [
+            None,
+            'Temperature',
+            'Celsius',
+            'Physical drives',
+            'hpssa.pd_temperature',
+            'line'
+        ],
+        'lines': []
+    }
+}
+
+adapter_regex = re.compile(r'^(?P<adapter_type>.+) in Slot (?P<slot>\d+)')
+ignored_sections_regex = re.compile(
+    r'''
+    ^
+        Physical[ ]Drives
+        | None[ ]attached
+        | (?:Expander|Enclosure|SEP|Port[ ]Name:)[ ].+
+        | .+[ ]at[ ]Port[ ]\S+,[ ]Box[ ]\d+,[ ].+
+        | Mirror[ ]Group[ ]\d+:
+    $
+    ''',
+    re.X
+)
+mirror_group_regex = re.compile(r'^Mirror Group \d+:$')
+disk_partition_regex = re.compile(r'^Disk Partition Information$')
+array_regex = re.compile(r'^Array: (?P<id>[A-Z]+)$')
+drive_regex = re.compile(
+    r'''
+    ^
+        Logical[ ]Drive:[ ](?P<logical_drive_id>\d+)
+        | physicaldrive[ ](?P<fqn>[^:]+:\d+:\d+)
+    $
+    ''',
+    re.X
+)
+key_value_regex = re.compile(r'^(?P<key>[^:]+): ?(?P<value>.*)$')
+ld_status_regex = re.compile(r'^Status: (?P<status>[^,]+)(?:, (?P<percentage>[0-9.]+)% complete)?$')
+error_match = re.compile(r'Error:')
+
+
+class HPSSAException(Exception):
+    pass
+
+
+class HPSSA(object):
+    def __init__(self, lines):
+        self.lines = [line.strip() for line in lines if line.strip()]
+        self.current_line = 0
+        self.adapters = []
+        self.parse()
+
+    def __iter__(self):
+        return self
+
+    def __next__(self):
+        if self.current_line == len(self.lines):
+            raise StopIteration
+
+        line = self.lines[self.current_line]
+        self.current_line += 1
+
+        return line
+
+    def next(self):
+        """
+        This is for Python 2.7 compatibility
+        """
+        return self.__next__()
+
+    def rewind(self):
+        self.current_line = max(self.current_line - 1, 0)
+
+    @staticmethod
+    def match_any(line, *regexes):
+        return any([regex.match(line) for regex in regexes])
+
+    def parse(self):
+        for line in self:
+            match = adapter_regex.match(line)
+            if match:
+                self.adapters.append(self.parse_adapter(**match.groupdict()))
+
+    def parse_adapter(self, slot, adapter_type):
+        adapter = {
+            'slot': int(slot),
+            'type': adapter_type,
+
+            'controller': {
+                'status': None,
+                'temperature': None,
+            },
+            'cache': {
+                'present': False,
+                'status': None,
+                'temperature': None,
+            },
+            'battery': {
+                'status': None,
+                'count': 0,
+            },
+
+            'logical_drives': [],
+            'physical_drives': [],
+        }
+
+        for line in self:
+            if error_match.match(line):
+                raise HPSSAException('Error: {}'.format(line))
+            elif adapter_regex.match(line):
+                self.rewind()
+                break
+            elif array_regex.match(line):
+                self.parse_array(adapter)
+            elif line == 'Unassigned' or line == 'HBA Drives':
+                self.parse_physical_drives(adapter)
+            elif ignored_sections_regex.match(line):
+                self.parse_ignored_section()
+            else:
+                match = key_value_regex.match(line)
+                if match:
+                    key, value = match.group('key', 'value')
+                    if key == 'Controller Status':
+                        adapter['controller']['status'] = value == 'OK'
+                    elif key == 'Controller Temperature (C)':
+                        adapter['controller']['temperature'] = int(value)
+                    elif key == 'Cache Board Present':
+                        adapter['cache']['present'] = value == 'True'
+                    elif key == 'Cache Status':
+                        adapter['cache']['status'] = value == 'OK'
+                    elif key == 'Cache Module Temperature (C)':
+                        adapter['cache']['temperature'] = int(value)
+                    elif key == 'Battery/Capacitor Count':
+                        adapter['battery']['count'] = int(value)
+                    elif key == 'Battery/Capacitor Status':
+                        adapter['battery']['status'] = value == 'OK'
+                else:
+                    raise HPSSAException('Cannot parse line: {}'.format(line))
+
+        return adapter
+
+    def parse_array(self, adapter):
+        for line in self:
+            if HPSSA.match_any(line, adapter_regex, array_regex, ignored_sections_regex):
+                self.rewind()
+                break
+
+            match = drive_regex.match(line)
+            if match:
+                data = match.groupdict()
+                if data['logical_drive_id']:
+                    self.parse_logical_drive(adapter, int(data['logical_drive_id']))
+                else:
+                    self.parse_physical_drive(adapter, data['fqn'])
+            elif not key_value_regex.match(line):
+                self.rewind()
+                break
+
+    def parse_physical_drives(self, adapter):
+        for line in self:
+            match = drive_regex.match(line)
+            if match:
+                self.parse_physical_drive(adapter, match.group('fqn'))
+            else:
+                self.rewind()
+                break
+
+    def parse_logical_drive(self, adapter, logical_drive_id):
+        ld = {
+            'id': logical_drive_id,
+            'status': None,
+            'status_complete': None,
+        }
+
+        for line in self:
+            if HPSSA.match_any(line, mirror_group_regex, disk_partition_regex):
+                self.parse_ignored_section()
+                continue
+
+            match = ld_status_regex.match(line)
+            if match:
+                ld['status'] = match.group('status') == 'OK'
+
+                if match.group('percentage'):
+                    ld['status_complete'] = float(match.group('percentage')) / 100
+            elif HPSSA.match_any(line, adapter_regex, array_regex, drive_regex, ignored_sections_regex) \
+                    or not key_value_regex.match(line):
+                self.rewind()
+                break
+
+        adapter['logical_drives'].append(ld)
+
+    def parse_physical_drive(self, adapter, fqn):
+        pd = {
+            'fqn': fqn,
+            'status': None,
+            'temperature': None,
+        }
+
+        for line in self:
+            if HPSSA.match_any(line, adapter_regex, array_regex, drive_regex, ignored_sections_regex):
+                self.rewind()
+                break
+
+            match = key_value_regex.match(line)
+            if match:
+                key, value = match.group('key', 'value')
+                if key == 'Status':
+                    pd['status'] = value == 'OK'
+                elif key == 'Current Temperature (C)':
+                    pd['temperature'] = int(value)
+            else:
+                self.rewind()
+                break
+
+        adapter['physical_drives'].append(pd)
+
+    def parse_ignored_section(self):
+        for line in self:
+            if HPSSA.match_any(line, adapter_regex, array_regex, drive_regex, ignored_sections_regex) \
+                    or not key_value_regex.match(line):
+                self.rewind()
+                break
+
+
+class Service(ExecutableService):
+    def __init__(self, configuration=None, name=None):
+        super(Service, self).__init__(configuration=configuration, name=name)
+        self.order = ORDER
+        self.definitions = deepcopy(CHARTS)
+        self.ssacli_path = self.configuration.get('ssacli_path', 'ssacli')
+        self.use_sudo = self.configuration.get('use_sudo', True)
+        self.cmd = []
+
+    def get_adapters(self):
+        try:
+            adapters = HPSSA(self._get_raw_data(command=self.cmd)).adapters
+            if not adapters:
+                # If no adapters are returned, run the command again but capture stderr
+                err = self._get_raw_data(command=self.cmd, stderr=True)
+                if err:
+                    raise HPSSAException('Error executing cmd {}: {}'.format(' '.join(self.cmd), '\n'.join(err)))
+            return adapters
+        except HPSSAException as ex:
+            self.error(ex)
+            return []
+
+    def check(self):
+        if not os.path.isfile(self.ssacli_path):
+            ssacli_path = find_binary(self.ssacli_path)
+            if ssacli_path:
+                self.ssacli_path = ssacli_path
+            else:
+                self.error('Cannot locate "{}" binary'.format(self.ssacli_path))
+                return False
+
+        if self.use_sudo:
+            sudo = find_binary('sudo')
+            if not sudo:
+                self.error('Cannot locate "{}" binary'.format('sudo'))
+                return False
+
+            allowed = self._get_raw_data(command=[sudo, '-n', '-l', self.ssacli_path])
+            if not allowed or allowed[0].strip() != os.path.realpath(self.ssacli_path):
+                self.error('Not allowed to run sudo for command {}'.format(self.ssacli_path))
+                return False
+
+            self.cmd = [sudo, '-n']
+
+        self.cmd.extend([self.ssacli_path, 'ctrl', 'all', 'show', 'config', 'detail'])
+        self.info('Command: {}'.format(self.cmd))
+
+        adapters = self.get_adapters()
+
+        self.info('Discovered adapters: {}'.format([adapter['type'] for adapter in adapters]))
+        if not adapters:
+            self.error('No adapters discovered')
+            return False
+
+        return True
+
+    def get_data(self):
+        netdata = {}
+
+        for adapter in self.get_adapters():
+            status_key = '{}_status'.format(adapter['slot'])
+            temperature_key = '{}_temperature'.format(adapter['slot'])
+            ld_key = 'ld_{}_'.format(adapter['slot'])
+
+            data = {
+                'ctrl_status': {
+                    'ctrl_' + status_key: adapter['controller']['status'],
+                    'cache_' + status_key: adapter['cache']['present'] and adapter['cache']['status'],
+                    'battery_' + status_key:
+                        adapter['battery']['status'] if adapter['battery']['count'] > 0 else None
+                },
+
+                'ctrl_temperature': {
+                    'ctrl_' + temperature_key: adapter['controller']['temperature'],
+                    'cache_' + temperature_key: adapter['cache']['temperature'],
+                },
+
+                'ld_status': {
+                    ld_key + '{}_status'.format(ld['id']): ld['status'] for ld in adapter['logical_drives']
+                },
+
+                'pd_status': {},
+                'pd_temperature': {},
+            }
+
+            for pd in adapter['physical_drives']:
+                pd_key = 'pd_{}_{}'.format(adapter['slot'], pd['fqn'])
+                data['pd_status'][pd_key + '_status'] = pd['status']
+                data['pd_temperature'][pd_key + '_temperature'] = pd['temperature']
+
+            for chart, dimension_data in data.items():
+                for dimension_id, value in dimension_data.items():
+                    if value is None:
+                        continue
+
+                    if dimension_id not in self.charts[chart]:
+                        self.charts[chart].add_dimension([dimension_id])
+
+                    netdata[dimension_id] = value
+
+        return netdata
diff --git a/collectors/python.d.plugin/hpssa/hpssa.conf b/collectors/python.d.plugin/hpssa/hpssa.conf
new file mode 100644
index 0000000..cc50c98
--- /dev/null
+++ b/collectors/python.d.plugin/hpssa/hpssa.conf
@@ -0,0 +1,61 @@
+# netdata python.d.plugin configuration for hpssa
+#
+# This file is in YaML format. Generally the format is:
+#
+# name: value
+#
+
+# ----------------------------------------------------------------------
+# Global Variables
+# These variables set the defaults for all JOBs, however each JOB
+# may define its own, overriding the defaults.
+
+# update_every sets the default data collection frequency.
+# If unset, the python.d.plugin default is used.
+# update_every: 5
+
+# priority controls the order of charts at the netdata dashboard.
+# Lower numbers move the charts towards the top of the page.
+# If unset, the default for python.d.plugin is used.
+# priority: 60000
+
+# penalty indicates whether to apply penalty to update_every in case of failures.
+# Penalty will increase every 5 failed updates in a row. Maximum penalty is 10 minutes.
+# penalty: yes
+
+# autodetection_retry sets the job re-check interval in seconds.
+# The job is not deleted if check fails.
+# Attempts to start the job are made once every autodetection_retry.
+# This feature is disabled by default.
+# autodetection_retry: 0
+
+# ----------------------------------------------------------------------
+# JOBS (data collection sources)
+#
+# The default JOBS share the same *name*. JOBS with the same name
+# are mutually exclusive. Only one of them will be allowed running at
+# any time. This allows autodetection to try several alternatives and
+# pick the one that works.
+#
+# Any number of jobs is supported.
+#
+# All python.d.plugin JOBS (for all its modules) support a set of
+# predefined parameters. These are:
+#
+# job_name:
+#     name: myname            # the JOB's name as it will appear at the
+#                             # dashboard (by default is the job_name)
+#                             # JOBs sharing a name are mutually exclusive
+#     update_every: 5         # the JOB's data collection frequency
+#     priority: 60000         # the JOB's order on the dashboard
+#     penalty: yes            # the JOB's penalty
+#     autodetection_retry: 0  # the JOB's re-check interval in seconds
+#
+# Additionally to the above, hpssa also supports the following:
+#
+#    ssacli_path: /usr/sbin/ssacli  # The path to the ssacli executable
+#    use_sudo: True                 # Whether to use sudo or not
+# ----------------------------------------------------------------------
+
+# ssacli_path: /usr/sbin/ssacli
+# use_sudo: True