logstash/docs/static/netflow-module.asciidoc
2019-09-16 18:49:22 +00:00

195 lines
6.5 KiB
Text

[[netflow-module]]
=== Logstash Netflow Module
++++
<titleabbrev>Netflow Module (deprecated)</titleabbrev>
++++
deprecated[7.4.0, Replaced by the {filebeat-ref}/filebeat-module-netflow.html[{Filebeat} Netflow Module] which is compliant with the {ecs-ref}/index.html[Elastic Common Schema (ECS)]]
The Logstash Netflow module simplifies the collection, normalization, and
visualization of network flow data. With a single command, the module parses
network flow data, indexes the events into Elasticsearch, and installs a suite
of Kibana dashboards to get you exploring your data immediately.
Logstash modules support Netflow Version 5 and 9.
==== What is Flow Data?
Netflow is a type of data record streamed from capable network devices. It
contains information about connections traversing the device, and includes
source IP addresses and ports, destination IP addresses and ports, types of
service, VLANs, and other information that can be encoded into frame and
protocol headers. With Netflow data, network operators can go beyond monitoring
simply the volume of data crossing their networks. They can understand where the
traffic originated, where it is going, and what services or applications it is
part of.
[[netflow-requirements]]
===== Requirements
These instructions assume you have already installed Elastic Stack
(Logstash, Elasticsearch, and Kibana) version 5.6 or higher. The products you
need are https://www.elastic.co/downloads[available to download] and easy to
install.
[[netflow-getting-started]]
==== Getting Started
NOTE: The {ls} Netflow Module has been deprecated and replaced by the
{filebeat-ref}/filebeat-module-netflow.html[{Filebeat} Netflow Module], which is
compliant with the {ecs-ref}/index.html[Elastic Common Schema (ECS)].
. Start the Logstash Netflow module by running the following command in the
Logstash installation directory:
+
[source,shell]
-----
bin/logstash --modules netflow --setup -M netflow.var.input.udp.port=NNNN
-----
+
Where `NNNN` is the UDP port on which Logstash will listen for network traffic
data. If you don't specify a port, Logstash listens on port 2055 by default.
+
The `--modules netflow` option spins up a Netflow-aware Logstash pipeline
for ingestion.
+
The `--setup` option creates a `netflow-*` index pattern in Elasticsearch and
imports Kibana dashboards and visualizations. Running `--setup` is a one-time
setup step. Omit this option for subsequent runs of the module to avoid
overwriting existing Kibana dashboards.
+
The command shown here assumes that you're running Elasticsearch and Kibana on
your localhost. If you're not, you need to specify additional connection
options. See <<configuring-netflow>>.
. Explore your data in Kibana:
.. Open your browser and navigate to
http://localhost:5601[http://localhost:5601]. If security is enabled, you'll
need to specify the Kibana username and password that you used when you set up
security.
.. Open *Netflow: Network Overview Dashboard*.
.. See <<exploring-data-netflow>> for additional details on data exploration.
[[exploring-data-netflow]]
==== Exploring Your Data
Once the Logstash Netflow module starts processing events, you can immediately
begin using the packaged Kibana dashboards to explore and visualize your
network flow data.
You can use the dashboards as-is, or tailor them to work better with existing
use cases and business requirements.
[[network-dashboards-netflow]]
===== Example Dashboards
On the *Overview* dashboard, you can see a summary of basic traffic data and set
up filters before you drill down to gain deeper insight into the data.
[role="screenshot"]
image::static/images/netflow-overview.png[Netflow overview dashboard]
For example, on the *Conversation Partners* dashboard, you can see the source
and destination addresses of the client and server in any conversation.
[role="screenshot"]
image::static/images/netflow-conversation-partners.png[Netflow conversation partners dashboard]
On the *Traffic Analysis* dashboard, you can identify high volume conversations
by viewing the traffic volume in bytes.
[role="screenshot"]
image::static/images/netflow-traffic-analysis.png[Netflow traffic analysis dashboard]
Then you can go to the *Geo Location* dashboard where you can visualize the
location of destinations and sources on a heat map.
[role="screenshot"]
image::static/images/netflow-geo-location.png[Netflow geo location dashboard]
[[configuring-netflow]]
==== Configuring the Module
You can further refine the behavior of the Logstash Netflow module by specifying
settings in the `logstash.yml` settings file, or overriding settings at the
command line.
For example, the following configuration in the `logstash.yml` file sets
Logstash to listen on port 9996 for network traffic data:
[source,yaml]
-----
modules:
- name: netflow
var.input.udp.port: 9996
-----
To specify the same settings at the command line, you use:
[source,shell]
-----
bin/logstash --modules netflow -M netflow.var.input.udp.port=9996
-----
For more information about configuring modules, see
<<logstash-modules>>.
[[netflow-module-config]]
===== Configuration Options
The Netflow module provides the following settings for configuring the behavior
of the module. These settings include Netflow-specific options plus common
options that are supported by all Logstash modules.
When you override a setting at the command line, remember to prefix the setting
with the module name, for example, `netflow.var.input.udp.port` instead of
`var.input.udp.port`.
If you don't specify configuration settings, Logstash uses the defaults.
*Netflow Options*
*`var.input.udp.port:`*::
+
--
* Value type is <<number,number>>
* Default value is 2055.
--
+
Sets the UDP port on which Logstash listens for network traffic data. Although
2055 is the default for this setting, some devices use ports in the range of
9995 through 9998, with 9996 being the most commonly used alternative.
*`var.input.udp.workers:`*::
+
--
* Value type is <<number,number>>
* Default value is 2.
--
+
Number of threads processing packets.
*`var.input.udp.receive_buffer_bytes:`*::
+
--
* Value type is <<number,number>>
* Default value is 212992.
--
+
The socket receive buffer size in bytes.
The operating system will use the max allowed value if receive_buffer_bytes is larger than allowed.
Consult your operating system documentation if you need to increase this max allowed value.
*`var.input.udp.queue_size:`*::
+
--
* Value type is <<number,number>>
* Default value is 2000.
--
+
This is the number of unprocessed UDP packets you can hold in memory before
packets will start dropping.
include::shared-module-options.asciidoc[]