Skip to content

Graphoscope is a solution to access multiple independent data sources from a common UI and show data relations as a graph

Notifications You must be signed in to change notification settings

hashlookup/graphoscope

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

75 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Graphoscope

An interactive solution to access multiple independent data sources from a common UI and show data relations as a graph:

Contains a list of by default available data sources plugins. 3rd party plugins are also possible. Example workflow if all needed data sources are connected:

  1. IP address only is know at the beginning, use it as initial query
  2. From Passive DNS find related domain names
  3. From second data source find institution that address belongs to
  4. From clients database find institution contact person's email
  5. From leaked passwords database find where that email was used and secret passwords

At the end of IP address -> Institution -> Contact person's email -> Password process there could be a possible explanation why something bad has happened to the institution's server.

Features

  • SQL query syntax with features like BETWEEN, IN, LIMIT
  • Usage examples to skip searching the correct SQL syntax
  • Use ... OR ... or field IN (...) queries even if data source doesn't support them
  • Common fields (like ip, domain, timestamp) that work across all the data sources
  • Fields autocomplete with a Tab key
  • Can query both Web GUI and API
  • Every single node & edge type can have its own style
  • No need to remember all the connected data source names - select it from a dropdown
  • Search for more node neighbors
  • Expand all selected nodes or just one
  • Combine the same node attributes from different sources
  • Select multiple nodes with a mouse right button
  • Ability to add notes to nodes and edges, visible to everybody
  • Center graph when it was scrolled/moved far outside of the visible area
  • Delete selected nodes with Del
  • Green inclusion filters, which work similar to Kibana's filters
  • Red exclusion filters are applied to all nodes (data source independent)
  • Get node & edge details by clicking on them
  • Pie-chart statistics when the amount of requested data exceeds the limit
  • Save & restore dashboards, private or shared with a team
  • Export/import all graph visible data
  • Display the amount of visible nodes grouped by type
  • Group node neighbors of specific types into the clusters
  • Cache results for a faster re-query
  • Save canvas as an image
  • Notification about new features
  • Format comma or space separated values into a valid search query
  • Upload indicators list and wait for the report
  • If enabled, query debug info is returned to the user to see what happens in a background
  • Background processing of the data collectors responses for graph enrichment, etc.
  • Personal graph settings
  • Global graph settings for administrators
  • User management
  • Built-in documentation
  • Real-time communication between the browser and server through the Websockets

Architecture

Documentation

Installation instructions can be found in INSTALL.md file.

Additional information is available in a Web GUI built-in documentation (also available in docs/):

  • UI elements in section UI
  • Search features in Search
  • Administration and connecting your own data sources in Administration

Plugins for the data sources

Available plugins are in plugins/src:

  • Elasticsearch
  • CSV file
  • HTTP API
  • MongoDB
  • MySQL
  • Pastelyzer
  • PostgreSQL
  • Redis
  • SQLite
  • AbuseIPDB

3rd party compiled *.so plugins should be placed in plugins/sources directory.

Plugins for the background processing of the data sources response

Available plugins are in plugins/src:

  • Taxonomy
  • Modify

3rd party compiled *.so plugins should be placed in plugins/processors directory.

Plugins development

Check a built-in documentation, section Administration.

Usage from Web GUI

With a connected demo data source (files/demo.csv) select it from a sources dropdown and request all people with an age over 30:

age > 30

Then extend the graph by searching for more of John's neighbors - right click on John and choose Search Demo to search for more data in a Demo data source. We find that Jennifer and Kate also are his friends:

ui-demo

Querying the global special source retrieves data from all the data sources connected to this namespace concurrently:

FROM global WHERE age > 30

API usage demo

API can be queried by the external tools, for example with curl:

# SELECT * FROM demo WHERE age > 30
curl -XGET 'https://localhost:443/api?uuid=09e545f2-3986-493c-983a-e39d310f695a&sql=FROM+demo+WHERE+age>30'
# SELECT * FROM global WHERE datetime BETWEEN '2019-01-20T07:27:54+02:00' AND '2019-01-20T07:27:54+02:00'
curl -XGET 'https://localhost:443/api?uuid=09e545f2-3986-493c-983a-e39d310f695a&sql=FROM+global+WHERE+datetime+BETWEEN+%272019-01-20T07:27:54%2B02:00%27+AND+%272019-01-20T07:27:54%2B02:00%27'
# SELECT * FROM intelmq WHERE feed.provider='ShadowServer' AND source.ip='10.10.10.1'
curl -XGET 'https://localhost:443/api?uuid=09e545f2-3986-493c-983a-e39d310f695a&sql=FROM+intelmq+WHERE+feed.provider=%27ShadowServer%27+AND+source.ip=%2710.10.10.1%27'

Response example for the first query:

{
    "relations": [
        {
            "from": {
                "id": "Monica",
                "attributes": {
                    "age": 35
                },
                "search": "name",
                "group": "name"
            },
            "edge": {
                "label": "lives in"
            },
            "to": {
                "id": "Canada",
                "search": "country",
                "group": "country"
            },
            "source": "demo"
        },{
            "from": {
                "id": "Chin",
                "search": "name",
                "group": "name"
            },
            "to": {
                "id": "Ben",
                "search": "name",
                "group": "name"
            },
            "source": "demo"
        }
    ]
}

... where main fields can be:

  • relations - list of relations, which consist of:
    • from - describes From node of a single relation
    • to - describes To node of the same relation
    • edge - describes a single connection between From and To nodes
    • source - data source name
  • stats - statistics for the processed data when the amount of entries exceeds the limit
  • error - possible error message

Fields to use in queries

sources/*.yaml data sources definitions allow to create common query fields, like ip, domain, datetime, etc. Other fields come from a related data source.

TODO & ideas

  • Generate DEB and RPM packages
  • In graph.js remove custom zoom limiting when visjs/vis-network#629 or similar is merged & new version released
  • Edges groups styling. TODO from search.js. Implement visjs/vis-network#1229
  • Generate PDF documentation from the existing *.md files
  • Generate a direct link to the saved dashboard
  • Data sources access based on user groups permissions
  • Implement other SQL features, like NOT BETWEEN
  • Data source plugins:
    • RTIR
    • MS SQL
    • Oracle SQL
    • Apache Cassandra
    • Hashlookup
    • Genji
    • Presto
    • MISP
    • VirusTotal
    • Shodan
    • Poland malware Sandbox / mwdb
    • General TCP
    • ipinfo.io
    • phishtank.org
  • Processor plugins:
    • Filter
  • Output plugins:
    • RTIR
    • JIRA
    • MISP
    • RITA
    • REST API

Useful info

License

This project is released under the GNU Affero General Public License v3 or later.

See LICENSE to see the full text.

About

Graphoscope is a solution to access multiple independent data sources from a common UI and show data relations as a graph

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Go 61.0%
  • JavaScript 23.3%
  • HTML 9.0%
  • CSS 6.7%