Skip to content
This repository has been archived by the owner on Sep 27, 2020. It is now read-only.

Latest commit

 

History

History
217 lines (168 loc) · 5.84 KB

Workshop Script.md

File metadata and controls

217 lines (168 loc) · 5.84 KB
Raw

We will follow this script

Configure InfluxDB

  • Download Influx DB

    docker pull influxdb:1.2.3
docker tag influxdb:1.2.3 influxdb:latest

  • Get and update the configuration file

    docker run --rm influxdb influxd config > influxdb.conf
vi influxdb.conf

  • turn on the admin interface

    [admin]
  enabled = true
  bind-address = ":8083"
  https-enabled = false
  https-certificate = "/etc/ssl/influxdb.pem"

  • Bring up InfluxDB

    cat docker-compose.yml
docker-compose up -d influxdb
docker-compose logs influxdb

  • Go to administrative interface.

    show measurements

Configure Telegraf

  • Download Telegraf

    docker pull telegraf:1.2.1
docker tag telegraf:1.2.1 telegraf:latest

  • Export sample config for Telegraf

    docker run --rm telegraf -sample-config > telegraf.conf

  • Configure Telegraf to talk to InfluxDB

    [[outputs.influxdb]]
## The full HTTP or UDP endpoint URL for your InfluxDB instance.
## Multiple urls can be specified as part of the same cluster,
## this means that only ONE of the urls will be written to each interval.
# urls = ["udp://localhost:8089"] # UDP endpoint example
urls = ["http://influxdb:8086"] # required
## The target database for metrics (telegraf will create it if not exists).
database = "telegraf" # required

  • Configure Telegraf to listen for StatsD Metrics

    # # Statsd Server
[[inputs.statsd]]
#   ## Address and port to host UDP listener on
service_address = ":8125"

  • Start StatsD and check the logs

    cat docker-compose.yml
docker-compose up -d telegraf
docker-compose logs telegraf

  • Validate the metrics made it to InfluxDB, from the administrative interface

    show measurements
select * from cpu

Learn about the different types of counters

How does Telegraf summarize data for InfluxDB? Default 10s, summarizes and writes to the database.

Counters

Counters are added up written to the DB every 10s. Demonstrate how counters work.

cat docker-compose.yml
docker-compose up test_counter
select * from test_counter
  • leave one counter up and running docker-compose up -d test_counter

Gauges

Only the last gauge measurement will be written to the database. Change sleep interval to demonstrate the behavior of multiple samples during a 10s intervale (last gauge is winner)

    cat docker-compose.yml
    cat gauge.sh
    docker-compose run -e SLEEP=10 --rm --name=test_gauge test_gauge
    docker-compose run -e SLEEP=1 --rm --name=test_gauge test_gauge
    docker-compose run -e SLEEP=5 --rm --name=test_gauge test_gauge
  • leave one gauge up and running docker-compose up -d test_gauge

Sets Sets can be used to count unique occurences. In the above example, the unique.users metric will be incremented by 1, then will not be incremented no matter how many times the value 100 is sent.

    cat docker-compose.yml
    cat sets.sh
    docker-compose run -e SLEEP=1 --rm --name=test_sets test_sets
    docker-compose run -e SLEEP=5 --rm --name=test_sets test_sets
    docker-compose run -e SLEEP=10 --rm --name=test_sets test_sets
  • leave one set up and running docker-compose up -d test_sets

Timings and Histograms

Timings are special, and have additional statics captured and calculated for the 10s interval. Each interval will count how many measurements were received, the minimum, maximum, stddev and mean are tracked. Run the below commands to demonstrate 10 second intervals, show metrics, repeat with 1 second interval

    cat docker-compose.yml
    cat timing.sh
    docker-compose run -e SLEEP=1 --rm --name=test_timing test_timing
    docker-compose run -e SLEEP=3 --rm --name=test_timing test_timing
    docker-compose run -e SLEEP=5 --rm --name=test_timing test_timing

Tagging your metrics

Tags allow you to put additional details on each metric for filtering or grouping.

    cat timing.yml
  • leave one timer up and running docker-compose up -d test_timing

Configure Grafana

We will now fire up Grafana.

cat docker-compose.yml
docker-compose up -d grafana

  • Log into the grafana console

  • Username/Pass = admin/admin (class should change this)

  • add datasource

Add Dashboard

  • Graph - just as it sounds, it's a graph
  • SingleStat - can show a current value of a metric with a guage and other effects
  • Table - it's a table...
  • Text - It's text
  • Alert List

Gauge metric

  • show mean and count in select. *Show what happens with the units
  • Discuss granularity of time interval >10s

Counter metric Sets metric

Timing metric

  • show field(count) available values
  • scale the counter
  • introduce tags to the statsd line

Configure Telegraf Simulated Clients

Create a dashboard with lots of stats, including single stats counters and tables

We will simulate a number of machines running telegraf by simply defining a new service in our docker-compose.yml file. Each node will only show the stats from the docker container, but this will serve our purposes for the demonstration. In the real world, telegraf would be on each node we want to monitor for system stats.

```
docker-compose up -d test_client
docker-compose scale test_client=10
```

Single Metric Counters

  • simulate response time with timers counters
    • Min response
    • Avg response
    • Worst Response

Create a Parameterized Dashboard

Create templates

SHOW TAG VALUES FROM "cpu" WITH KEY = "host"