infoblox-discovery 0.4.0

A Prometheus file and http discovery for infoblox

infoblox-discovery

Overview

The infoblox-discovery service collect different data from a Infloblox master server about different objects that can be monitored by Prometheus using different exporters as the blackbox-exporter and infoblox-exporter. The current supporter objects are:

• members
• nodes
• dns_servers
• dhcp_ranges
• zones
• web_endpoints

Labels naming (since 0.2.0)

All labels are returned prefixed as __meta_infoblox_

Design

The infoblox-discovery can run in 2 modes, as file discovery or as http discovery. In http discovery mode the infoblox-discovery will run as a http service that Prometheus can access using http_discovery. In file discovery the infoblox-discovery will just create Prometheus file discovery files.

When running in http discovery mode the collection of data is done on an interval, default every 3600 sec, and the data is cached. The main reason is to limit a high number of calls to the infoblox master server.

Zones

The query is based on object 'zone_auth' with the query where 'view' is 'External'. The logic detect reverse and fqdn based zones.

Web endpoints

These fqdn "hosts" that are based on networks, e.g. 192.91.218.0/24. The result is based on two queries.

1. Get all ipv4address from the network where type is HOST
2. For all the above get all dns_aliases from record:host and check if External is in the _ref string

The networks that are subject to be scraped is based on the networks defined in the configuration file, see below.

Configuration

See the example_config.yml file.

Inclusion and exclusion filters

To manage what objects to include or exclude from discovery you can use labels defined in extattrs in infoblox. Inclusion and exclusion labels work as a filter to only include objects or exclude objects with the defined labels. Inclusion labels take precedence over exclusion labels. If inclusion labels are defined for an object type no exclusion labels are applied for that object type. Please see the example configuration file, example_config.yml for details.

Run

The infoblox-discovery is available on PyPI and can be installed using pip:

pip install infoblox-discovery

Environment variables

• INFOBLOX_DISCOVERY_CONFIG - the configuration file, default to config.yml
• INFOBLOX_DISCOVERY_PROMETHEUS_SD_FILE_DIRECTORY - the directory where file discovery based will be created, no default only used when run for file discovery
• INFOBLOX_DISCOVERY_HOST - the host to run the discovery service, default 0.0.0.0
• INFOBLOX_DISCOVERY_PORT - the port to run the discovery service, default 9694
• INFOBLOX_DISCOVERY_BASIC_AUTH_USERNAME - the basic auth username to the discovery service, no default.
• INFOBLOX_DISCOVERY_BASIC_AUTH_PASSWORD - the basic auth username to the discovery service, no default.
• INFOBLOX_DISCOVERY_LOG_FILE - log file, default stdout
• INFOBLOX_DISCOVERY_LOG_LEVEL - the log level, default INFO
• INFOBLOX_DISCOVERY_CACHE_TTL - the discovered data ttl in seconds, must be higher than INFOBLOX_DISCOVERY_FETCH_INTERVAL, default 7200
• INFOBLOX_DISCOVERY_FETCH_INTERVAL - the interval to collect discover data, default 3600

INFOBLOX_DISCOVERY_BASIC_AUTH_USERNAME and INFOBLOX_DISCOVERY_BASIC_AUTH_PASSWORD must be set - the discovery can not run without basic authentication.

File discovery mode

python -m infoblox_discovery

Http discovery mode

python -m infoblox_discovery --server

Test

curl -s 'localhost:9694/prometheus-sd-targets?master=infoblox.foo.com&type=members'

The master must match the master entry in the configuration file. The type can be:

• members
• nodes
• dns_servers
• dhcp_ranges
• zones
• web_endpoints