Question 1

Uninstall the monitoring agent

Accepted Answer

To remove this server from Nixstats first remove the monitoring agent. About 30 minutes after the monitoring agent stops sending data a remove server button should appear on the servers page.

For version 1.0 and up:

If your system uses systemd

```systemctl stop nixstatsagent systemctl disable nixstatsagent rm -f /etc/systemd/system/nixstatsagent

For init

service nixstatsagent stop update-rc.d -f nixstatsagent remove rm -f /etc/init.d/nixstatsagent```

Finally run:

rm -f /etc/nixstats.ini
rm -f /etc/nixstats-token.ini
pip uninstall nixstatsagent
userdel nixstats```

For version 0.963:

systemctl stop nixstats systemctl disable nixstats rm -f /etc/systemd/system/nixstats rm -rf /opt/nixstats/```

Question 2

Monitoring NGiNX

Accepted Answer

Metrics

The following metrics are submitted to Nixstats.

Requests per second
Accepts per second
Handled per second
Total requests
Active connections
Writing
Waiting
Reading

Dependencies

urllib2
nginx compiled with --with-httpstubstatus_module parameter

Stub status page

To enable the nginx stub status page add the following code to your nginx site-enabled configuration file.

server { listen 127.0.0.1:8080; server_name localhost; location /status_page { stub_status on; allow 127.0.0.1; deny all; } }

The status page will only be available from the machine running NGiNX. If you have something else running on port 8080 feel free to change this to an available port.

You can test if the status page is working by running wget -qO- http://127.0.0.1:8080/status_page

Configure the plugin

To enable the plugin edit the nixstats configuration file. This file is usually located at /etc/nixstats.ini.

Add the following settings to the end of the file.

[nginx] enabled = yes status_page_url = http://127.0.0.1:8080/status_page Test to see if the plugin is working by running nixstatsagent --test nginx. It should show the following metrics:

`` nginx: { "accepts": 505768, "acceptspersecond": 0.0, "active_connections": 198, "handled": 505768,

Question 3

Monitoring PHP-FPM

Accepted Answer

Metrics

The following metrics are submitted to Nixstats.

Accepted connections (per second, total)
Active processes
Total processes
Idle processes
Max active processes
Listen queue
Listen queue length
Max listen queue
Max children reached
Start time
Start since
Pool
Process manager
Slow requests

Dependencies

urllib2
json

PHP-FPM status page

To enable the PHP-FPM status page add the following code to your PHP-FPM pool configuration file.

pm.status_path = /status_phpfpm Now make the following changes to your NGiNX configuration file.

server { listen 127.0.0.1:8080; server_name localhost; location /status_phpfpm { fastcgi_pass 127.0.0.1:9000; include fastcgi_params; fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; allow 127.0.0.1; deny all; } }

Or for Apache edit etc/apache2/mods-available/fastcgi.conf (or similar):

``` # Define a named handler AddHandler php7-fcgi .php # Generate an alias pointing to /usr/lib/cgi-bin/php[VersionNumber]-fcgi Alias /php7-fcgi /usr/lib/cgi-bin/php7-fcgi # Configure an external server handling your upcoming requests (note where the alias is pointing towards) FastCgiExternalServer /usr/lib/cgi-bin/php7-fcgi -soc

Question 4

Creating a custom plugin

Accepted Answer

The nixstats monitoring agent comes with a set of plugins to monitor your servers CPU, Memory, Disk, Network, Processes and a set of plugins for your services such as MongoDB, PHP-FPM, NGiNX, Apache2/httpd.

Besides these plugins it's possible to monitor other metrics on your server by building your own plugin. Plugins are written in Python 2 or Python 3. To find out the version of python your agent is running on run nixstatsagent info the path to the plugin directory should reveal the python version.

The base of a plugin is as follows.

``#!/usr/bin/env python import plugins

class Plugin(plugins.BasePlugin): name = 'fridge'

def run(self, *unused):
    results = {'milk': 10, 'eggs': 4}
    return results

if name == 'main': Plugin().execute() ``

This is a simple example plugin. You have to save the plugin to the plugins directory. You can find out your plugin directory by running the following command.

root@ns:~# nixstatsagent info Version: 1.1.18 Plugins enabled: memory, diskusage, process, loadavg, network, cpu, ping, iostat, swap, system Plugins directory: /usr/local/lib/python2.7/dist-packages/nixstatsagent/plugins Server: 53bad1a222c9eea0378a4567

Save the plugin title under the same value as the __name__ for example in this examples case we would save the plugin to /usr/local/lib/python2.7/dist-packages/nixstatsagent/plugins/fridge.py

Now we can test the plugin by running:

`` root@ns:~# sudo -u nixstats nixstatsagent

Question 5

Monitoring MySQL

Accepted Answer

Metrics

The following metrics are submitted to Nixstats.

maxusedconnections
open_files
open_tables
qcachefreeblocks
qcachefreememory
qcachetotalblocks
slaveopentemp_tables
threads_cached
threads_connected
threads_running
uptime
aborted_clients
aborted_connects
binlogcachedisk_use
binlogcacheuse
bytes_received
bytes_sent
com_delete
comdeletemulti
com_insert
cominsertselect
com_load
com_replace
comreplaceselect
com_select
com_update
comupdatemulti
connections
createdtmpdisk_tables
createdtmpfiles
createdtmptables
key_reads
keyreadrequests
key_writes
keywriterequests
maxusedconnections
open_files
open_tables
opened_tables
qcachefreeblocks
qcachefreememory
qcache_hits
qcache_inserts
qcachelowmemprunes
qcachenotcached
qcachequeriesin_cache
qcachetotalblocks
questions
selectfulljoin
selectfullrange_join
select_range
selectrangecheck
select_scan
slaveopentemp_tables
slaveretriedtransactions
slowlaunchthreads
slow_queries
sort_range
sort_rows
sort_scan
tablelocksimmediate
tablelockswaited
threads_cached
threads_connected
threads_created
threads_running

Dependencies

MySQL-python

Run apt-get install python-mysqldb on debian based distro's or `yum in

Question 6

Monitoring HDD/SSD/NVMe Disk health

Accepted Answer

Using the diskstatus plugin you will be able to monitor the status of your hdd, ssd and/or nvme drives. This plugin requires the agent to be run under the root user.

This plugin requires nixstatsagent version 1.1.50 or higher. To upgrade run pip install nixstatsagent --upgrade.

Run as root

The smartctl command needs to be run as the root user, therefor it's required to run the agent as the root user.

Running the monitoring agent as root user.

Install dependencies

For NVMe drives you need nvme-cli, instructions are available here.

For all other drives you need smartctl from smartmontools. On debian and ubuntu run apt-get install smartmontools or under Cent OS run yum install smartmontools.

Test the plugin (smartmontools)

Run nixstatsagent test diskstatus to check if the plugin is working.

Test the plugin (nvme-cli)

Run nixstatsagent test diskstatus-nvme to check if the plugin is working.

Enabling the plugin

Edit /etc/nixstats.ini and add the following to enable the plugin.

[diskstatus]
enabled = yes

Enabling the NVME plugin

Edit /etc/nixstats.ini and add the following to enable the plugin.

[diskstatus-nvme]
enabled = yes

Restart the agent

Run `service nixsta

Frequently Read Articles

Browse All Categories