You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
scrutiny/docs/INSTALL_MANUAL.md

4.8 KiB

Manual Install

While the easiest way to get started with Scrutiny is using Docker, it is possible to run it manually without much work. You can even mix and match, using Docker for one component and a manual installation for the other. There's also an installer which automates this manual installation procedure.

Scrutiny is made up of two components: a collector and a webapp/api. Here's how each component can be deployed manually.

Note: the /opt/scrutiny directory is not hardcoded, you can use any directory name/path.

Webapp/API

Dependencies

Since the webapp is packaged as a stand alone binary, there isn't really any software you need to install other than glibc which is included by most linux OS's already.

Directory Structure

Now let's create a directory structure to contain the Scrutiny files & binary.

mkdir -p /opt/scrutiny/config
mkdir -p /opt/scrutiny/web
mkdir -p /opt/scrutiny/bin

Config file

While it is possible to run the webapp/api without a config file, the defaults are designed for use in a container environment, and so will need to be overridden. So the first thing you'll need to do is create a config file that looks like the following:

# stored in /opt/scrutiny/config/scrutiny.yaml

version: 1

web:
  database:
    # The Scrutiny webapp will create a database for you, however the parent directory must exist.
    location: /opt/scrutiny/config/scrutiny.db
  src:
    frontend:
      # The path to the Scrutiny frontend files (js, css, images) must be specified.
      # We'll populate it with files in the next section
      path: /opt/scrutiny/web

Note: for a full list of available configuration options, please check the example.scrutiny.yaml file.

Download Files

Next, we'll download the Scrutiny API binary and frontend files from the latest Github release. The files you need to download are named:

  • scrutiny-web-linux-amd64 - save this file to /opt/scrutiny/bin
  • scrutiny-web-frontend.tar.gz - save this file to /opt/scrutiny/web

Prepare Scrutiny

Now that we have downloaded the required files, let's prepare the filesystem.

# Let's make sure the Scrutiny webapp is executable.
chmod +x /opt/scrutiny/bin/scrutiny-web-linux-amd64

# Next, lets extract the frontend files.
cd /opt/scrutiny/web
tar xvzf scrutiny-web-frontend.tar.gz --strip-components 1 -C .

# Cleanup
rm -rf scrutiny-web-frontend.tar.gz

Start Scrutiny Webapp

Finally, we start the Scrutiny webapp:

/opt/scrutiny/bin/scrutiny-web-linux-amd64 start --config /opt/scrutiny/config/scrutiny.yaml

The webapp listens for traffic on http://0.0.0.0:8080 by default.

Collector

Dependencies

Unlike the webapp, the collector does have some dependencies:

  • smartctl, v7+
  • cron (or an alternative process scheduler)

Unfortunately the version of smartmontools (which contains smartctl) available in some of the base OS repositories is ancient. So you'll need to install the v7+ version using one of the following commands:

  • Ubuntu: apt-get install -y smartmontools=7.0-0ubuntu1~ubuntu18.04.1
  • Centos8:
    • dnf install https://extras.getpagespeed.com/release-el8-latest.rpm
    • dnf install smartmontools
  • FreeBSD: pkg install smartmontools

Directory Structure

Now let's create a directory structure to contain the Scrutiny collector binary.

mkdir -p /opt/scrutiny/bin

Download Files

Next, we'll download the Scrutiny collector binary from the latest Github release. The file you need to download is named:

  • scrutiny-collector-metrics-linux-amd64 - save this file to /opt/scrutiny/bin

Prepare Scrutiny

Now that we have downloaded the required files, let's prepare the filesystem.

# Let's make sure the Scrutiny collector is executable.
chmod +x /opt/scrutiny/bin/scrutiny-collector-metrics-linux-amd64

Start Scrutiny Collector, Populate Webapp

Next, we will manually trigger the collector, to populate the Scrutiny dashboard:

NOTE: if you need to pass a config file to the scrutiny collector, you can provide it using the --config flag.

/opt/scrutiny/bin/scrutiny-collector-metrics-linux-amd64 run --api-endpoint "http://localhost:8080"

Schedule Collector with Cron

Finally you need to schedule the collector to run periodically. This may be different depending on your OS/environment, but it may look something like this:

# open crontab
crontab -e

# add a line for Scrutiny
*/15 * * * * . /etc/profile; /opt/scrutiny/bin/scrutiny-collector-metrics-linux-amd64 run --api-endpoint "http://localhost:8080"