Monitoring with Prometheus

Monitoring With Prometheus
Richard Langlois P. Eng. and Gervais Naoussi, Sept. 2018

Agenda
2
• Monitoring Concepts
• Micrometer
• Prometheus
• Grafana
• Demo

Monitoring
3
Monitoring is the tools and processes by which you measure your technology systems.
A monitoring system has two customers:
• Technology (Engineering, Operations, DevOps)
• The business (measure the value that technology delivers to business)
If you’re building a specification or user stories for your application:
include metrics and monitoring for each component of your application.
Don’t wait until the end of a project or just before deployment.

Monitoring
Approach to Monitoring
4
A good approach to your monitoring is to design a top-down monitoring plan based on
value.
Identify the parts of the application that deliver value and monitor those first, working
your way down the stack.
Monitoring for the correctness of a service first
 e.g. monitor the content or rates of a business transaction rather than the uptime
of the web server it runs on.

Monitoring
Monitoring Approaches
5
2 major approaches:
• Probing monitoring probes the outside of an application (black-box monitoring).
e.g. Nagios
• Introspection monitoring looks at what’s inside the application (white-box monitoring)
application is instrumented and returns measurements of its state

Monitoring
Pull vs Push
6
Two approaches to how monitoring checks are executed:
• Pull-based: systems scrape or check a remote application—for example, an endpoint
containing metrics.
• Push-based: applications emit events that are received by the monitoring system.
Prometheus is primarily a pull-based system, but it also supports receiving events pushed
into a gateway.

Monitoring
Metric
7
Metrics are measures of properties of components of software or hardware.
To make a metric useful we keep track of its state, generally recording data points over time (called
observations).
An observation consists of:
value,
a timestamp,
and sometimes a series of properties that describe the observation such as a source or tags.
A collection of observations is called a time series.
Time series data is a chronologically ordered list of these observations.
Time series metrics are often visualized as a two-dimensional plot with data values on the y-axis and
time on the x-axis.

Monitoring
Types of monitoring data
8
Monitoring tools can collect 2 types of data:
• Metrics are stored as time series data that record the state of measures of your
applications.
• Logs are (usually textual) events emitted from an application.
Prometheus is primarily focused on collecting time series data.

Monitoring
Type of Metrics
9
Variety of different types of metrics:
Gauges: are numbers that are expected to go up or down. A snapshot of a specific measurement.
 e.g. Disk usage, number of customers present on a site.
Counters: are numbers that increase over time and never decrease.
 e.g. system uptime, number of sales in a month
Histograms: is a metric that samples observations. Each observation is counted and placed into
buckets.
Metric Summaries: mathematical transformations applied to metrics
• Average
• Median
• Standard Deviation
• Percentile

Agenda
10
• Micrometer
• Prometheus
• Grafana
• Demo

Micrometer
Intro
11
Micrometer is a metrics instrumentation library allowing to instrument JVM-based
application code without vendor lock-in.
It provides a simple façade over the instrumentation clients for the most popular monitoring
systems.
Think SLF4J, but for application metrics.
As of Spring Boot 2.0.0.M5, Micrometer is the instrumentation library used by Spring.
Some supported monitoring systems:
• Datadog
• Graphite
• Influx
• JMX
• New Relic
• Prometheus
• SignalFX
• StatsD

Micrometer
API
12
Meter is the interface for collecting a set of measurements (called metrics).
MeterRegistry: eters are created from and held in a Meter Registry
Each supported monitoring system has an implementation of MeterRegistry.
SimpleMeterRegistry: Automatically autowired in Spring-based apps.
MeterRegistry registry = new SimpleMeterRegistry.
Set of meter primitives:
Timer, Counter, Gauge, DistributionSummary, LongTaskTimer, FunctionCounter, FunctionTimer
and TimeGauge.
Dimensions allow a particular named metric to be sliced to drill down.
E.g. Registry.counter(“http.server.requests”, “uri”, “/api/users”)
Fluent builder:
Counter counter = Counter .builder("counter")
.baseUnit(“ms")
.description("a description of what this counter does")
.tags("region", "test")
.register(registry);

Agenda
13
• Micrometer
• Prometheus
• Grafana
• Demo

Prometheus
Intro
14
Prometheus is a simple, effective open-source monitoring system.
Promoted from incubation to graduation (in August 2018), in Cloud Native Computing Foundation (CNCF)
Prometheus works by scraping (pulling) time series data exposed from applications.
The time series data is exposed by the applications themselves often via client libraries or via proxies
called exporters, as HTTP endpoints.

Prometheus
Concepts
15
Prometheus calls the source of metrics it can scrape endpoints.
An endpoint usually corresponds to a single process, host, service, or application.
The resulting time series data is collected
and stored locally on the Prometheus server (15 days retention)
and can be sent from the server to external storage or to another time series database.
Prometheus can also define Rules for alerting.

Prometheus
PromQL – inbuilt querying language
16
The Prometheus server also comes with an inbuilt querying language, PromQL, allowing to
query and aggregate metrics.
Use this query language in the query input box in the Expression Browser.
e.g: Query all metrics with a label of quantile=“0.5”:

Prometheus
Expression Browser
17
http://localhost:9090/graph

Prometheus
Scalability
18
Designed to scale to millions of time series from many thousands of hosts.
Its data storage format is designed to keep disk use down and provide fast retrieval of time series
during queries and aggregations.
SSD disks are recommended for Prometheus servers, for speed and reliability.
Redundant Prometheus Architecture:

Prometheus
Data Model
19
Prometheus collects time series data.
Format:
<time series name>{<label name>=<label value>, ...}
Each time series is uniquely identified by the combination of names and key/value pairs
called labels (provide the dimensions).
Name usually describes the general nature of the time series data being collected
 e.g. total_website_visits as the total number of website visits.
Labels enable the Prometheus dimensional data model, they add context to a specific
time series.
 e.g. the name of the website, IP of the requester

Prometheus
Time Series Notation
20
Example
total_website_visits{site=“alithya.com", location="NJ", instance="webserver“, job="web"}
All time series generally have
• an instance label, which identifies the source host or application
• a job label, which contains the name of the job that scraped the specific time series.
Actual value of the time series is called a sample.
Consists of:
• A float64 value.
• A millisecond-precision timestamp.

Prometheus configuration
prometheus.yml
21
Prometheus is configured via YAML configuration files.
Default Configuration file has the following 4 YAML blocks defined:
Global: contains global settings for controlling the Prometheus server’s behavior.
Alerting: configures Prometheus’ alerting.
rule_files: specifies a list of files that can contain recording or alerting rules.
scrape_configs: specifies all of the targets that Prometheus will scrape.

Prometheus and Spring Boot
22
Spring Boot auto-configures a composite MeterRegistry and adds a registry to the composite for each of the supported
implementations that it finds on the classpath.
pom.xml:
<dependency>
<groupId>io.prometheus</groupId>
<artifactId>simpleclient_spring_boot</artifactId>
<version>0.1.0</version>
</dependency>
<dependency>
<groupId>io.prometheus</groupId>
<artifactId>simpleclient_hotspot</artifactId>
<version>0.1.0</version>
</dependency
The simpleclient_spring_boot dependency provides the @EnablePrometheusEndpoint annotation.
Adding it to a @configuration class, creates a HTTP endpoint accessible via /actuator/prometheus that exposes all registered
(actuator) metrics in a Prometheus data format.

Prometheus configuration
Scrape Config for Spring Boot application
23
Prometheus scrapes the following 2 endpoints
• /prometheus endpoint: contains Spring boot metrics
• /metrics endpoint: Prometheus own metrics
scrape_configs:
# The job name is added as a label `job=<job_name>` to any time series scraped from this config.
- job_name: 'prometheus'
# metrics_path defaults to '/metrics'
# scheme defaults to 'http'.
static_configs:
- targets: ['PROM_IP:9090']
- job_name: 'spring-boot'
metrics_path: '/prometheus'
scrape_interval: 5s
static_configs:
- targets: ['App_IP:8080']

Prometheus
Dashboard
24
Example of metrics on Prometheus dashboard:

Prometheus
Alerting
25
Alerting is provided by a tool called Alertmanager
Alerting rules are defined on the Prometheus server.
When the threshold or criteria is met, an alert will be generated and pushed to Alertmanager.
The alerts are received on an HTTP endpoint on the Alertmanager.
Alertmanager handles deduplicating, grouping, and routing alerts to receivers (e.g. email, SMS, PagerDuty)

Prometheus
Alerting Configuration
26
A simple alertmanager.yml configuration file, sending alerts by email:

Prometheus
Alerting Web Interface
27
Web interface to:
• view current alerts
• manage maintenance window alert suppression (silences)

Prometheus
Pushgateway
28
Metrics can be pushed to Pushgateway when there isn’t a target from which to scrape metrics because:
• can’t reach the target resources because of security
• target resource has too short a lifespan (e.g. container starting, executing, and stopping).
• target resource doesn’t have an endpoint, (e.g. batch job).
Pushgateway sits between an application sending metrics and the Prometheus server.
Pushgateway is scraped as a target to deliver the metrics to the Prometheus server.

Agenda
29
• Micrometer
• Prometheus
• Grafana
• Demo

Grafana
30
Prometheus UI is not really nice.
Alternative: Grafana is open source metrics Dashboard platform.
It supports multiple backend time-series databases including:
Prometheus , InfluxDB, Elasticsearch, Cloudwatch …
Example of Grafana dashboard:

Grafana
Prometheus as Data source
31

Grafana
Prometheus as Datasource
32
Before you create your first dashboard you need to add your data source.

Grafana
Prometheus as Datasource
33
Name: your choice
Default: Check to tell Grafana to search for data in this source by default
Type: Prometheus
URL: URL of the Prometheus server to query.

Agenda
34
• Micrometer
• Prometheus
• Grafana
• Demo

References
35
Micrometer:
• https://micrometer.io
Prometheus
• Book: Monitoring with Prometheus, James Turnbull, 2018.
• https://prometheus.io
Graphana
• https://grafana.com/

Monitoring with Prometheus

Recommended

More Related Content

What's hot (20)

Similar to Monitoring with Prometheus (20)

More from Richard Langlois P. Eng. (7)

Recently uploaded (20)

Monitoring with Prometheus