Skip to content

Latest commit

 

History

History
91 lines (64 loc) · 3.2 KB

README.md

File metadata and controls

91 lines (64 loc) · 3.2 KB

NVD River for ElasticSearch

Welcome to the NVD River Plugin for Elasticsearch. The purpose of this plugin is to read the CVE vulnerability XML feeds of the National Vulnerability Database and push it into ElasticSearch. Another purpose was to learn ElasticSearch and get some hands on experience.

Versions

The plugin code is considered highly experimental and was only tested with Java 7.

Installing

  1. ~/code$ git clone https://github.com/frosenberg/elasticsearch-nvd-river
  2. ~/code$ cd elasticsearch-nvd-river
  3. ~/code/elasticsearch-nvd-river$ mvn clean package (will take a while b/c of some long running tests)
  4. Copy the package from ~/code/elasticsearch-nvd-river/target/release/nvdriver-1.0.0-SNAPSHOT.zip to your ElasticSearch instance
  5. bin/plugin --url file:///path/to/nvdriver-1.0.0-SNAPSHOT.zip --install nvdriver

Usage

You just need to execute a few simple cURL command to feed the data from the NIST NVD database.

Create an index

curl -XPUT 'http://localhost:9200/nvd/' -d '{}'

Create a mapping

Mappings are not needed and currently not supported. I did not see the need in my use case. ElasticSearch has pretty good dynamic mapping generation that was sufficient. If you feel the need to customize the mapping please open an issue or contribute.

Create the river

This will create a river and will push in the following XML feed. You can of course add more if needed, or remove some. Keep in mind this may take some time until all of the are processed.

curl -XPUT 'localhost:9200/_river/nvd/_meta' -d '{
  "type": "nvd",
  "nvd": {
    "streams" : [
      {
        "name": "cve-modified",
        "url": "http://static.nvd.nist.gov/feeds/xml/cve/nvdcve-2.0-modified.xml"
      },
      {
        "name": "cve-2014",
        "url": "http://static.nvd.nist.gov/feeds/xml/cve/nvdcve-2.0-2014.xml"
      },
      {
        "name": "cve-2013",
        "url": "http://static.nvd.nist.gov/feeds/xml/cve/nvdcve-2.0-2013.xml"
      },
      {
        "name": "cve-2012",
        "url": "http://static.nvd.nist.gov/feeds/xml/cve/nvdcve-2.0-2012.xml"
      },
      {
        "name": "cve-2011",
        "url": "http://static.nvd.nist.gov/feeds/xml/cve/nvdcve-2.0-2011.xml"
      },
      {
        "name": "cve-2010",
        "url": "http://static.nvd.nist.gov/feeds/xml/cve/nvdcve-2.0-2010.xml"
      }
    ]
  }
}'

Test the river plugin

curl -XGET 'http://localhost:9200/nvd/_search?q=oracle'

Developer Notes

The code contains a lot of generated classes that were generated from the CVE XML schema 2.0 located at [http://nvd.nist.gov/schema/nvd-cve-feed_2.0.xsd]

Compiling the NVD XML schema files:

The following command will compile the schema into Java JAXB version 2.1 data bindings.

~/code/elasticsearch-nvd-river$ xjc -d src/main/java/ -target 2.1 http://nvd.nist.gov/schema/nvd-cve-feed_2.0.xsd

We do not use a specific namespace to Java package customization because this will only complicate matters.