NUCLEUS Data Server

The NUCLEUS Data Server is a REST server that is used with NUCLEUS to provide datastore adapters, run datastore queries, transform query results, and perform data processing. The Neon Dashboard is a UI that interacts with the NUCLEUS Data Server.

Table of Content

Initial Setup Instructions

Install Dependencies

Java 9+ (Tested on OpenJDK versions 11 and 12)
- In addition to installing Java9+, you may also need to remove earlier versions (MacOS)
Docker
A supported datastore

Download Source Code

git clone https://github.com/NextCenturyCorporation/nucleus-data-server.git; cd nucleus-data-server

Load Data

If you were given a sample data bundle by the Neon / NUCLEUS Development Team, please download it and follow the specific instructions in its README file.

If you want to use your own data, please see the Datastore Configuration for more information.

Customize Build (Optional)

Update the runtime properties of your NUCLEUS Data Server installation by editing the server/src/main/resources/application.properties file. See the Spring Boot Configuration Documentation for details. Please note that the server.servlet.context-path application property should always end with /services due to current assumptions made in NUCLEUS.

Datastore Authentication

Basic Auth

In the server/src/main/resources/application.properties file, add the following property, replacing my_datastore_type with your datastore type (elasticsearch, mysql, postgresql):

my_datastore_type.auth={'hostname':'username:password'}

Replace the hostname, username, and password as needed. The hostname can be an IP address or a CNAME and can optionally have a port. If you need multiple authentication entries, separate them with commas:

my_datastore_type.auth={'hostname1':'username1:password1','hostname2':'username2:password2'}

Local Development Instructions

Build and Run Tests

./gradlew build

Run Locally

./runLocal.sh

This will run bootRun from the Spring Boot Gradle plugin. To pass custom arguments into bootRun from the command line, use --args='<arguments>'. For example, to run on a specific port: ./runLocal.sh --args='--server.port=1234'`

Production Deployment Instructions

The NUCLEUS Data Server is deployed as an docker container independent from other applications (like the Neon Dashboard).

Deploy as Docker Container

1. Perform All Initial Setup

Follow the Initial Setup Instructions above.

2. (Optional) Update the NUCLEUS Data Server's Port

By default, the NUCLEUS Data Server runs on port 8090. If you want to use a different port:

In <nucleus-data-server>/server/src/main/resources/application.properties, change the line server.port=8090 to use your port
In Dockerfile, change the line EXPOSE 8090 to use your port

3. Build the Docker Image

./gradlew clean docker

4. Verify the Docker Image

Run docker images to verify that you have created a docker image with the repository com.ncc.neon/server and tag latest.

5. (Optional) Run the Docker Container Locally

docker run -it --network=host --rm com.ncc.neon/server:latest

Datastore Support

The NUCLEUS Data Server supports the following datastores:

Elasticsearch support provided by the official Java High Level REST Client. SQL support provided by R2DBC drivers.

Want us to support another datastore? Please let us know!

Datastore Configuration

Elasticsearch 6 and 7

We recommend installing elasticdump to load bulk data: npm install -g elasticdump

If you have previously installed elasticdump, we recommend that you rerun the command to update it to its latest version.

Elasticsearch Data Format

If your data is spread across multiple indexes, we recommend that you denormalize any fields that you want to use as filters.

Elasticsearch Mapping Files

It's usually very important to load a mapping file associated with your data index into Elasticsearch BEFORE loading any data into that index.

If you HAVE loaded data before loading your mapping file, you'll either need to reindex your data index or delete the index and start over again.

More information about mapping files can be found on the Elasticsearch website.

Elasticsearch Date Fields

Date fields should have the format "yyyy-MM-dd||dateOptionalTime||E MMM d HH:mm:ss zzz yyyy". For example:

"timestamp": {
    "type": "date",
    "format": "yyyy-MM-dd||dateOptionalTime||E MMM d HH:mm:ss zzz yyyy"
}

Note that you may need to add an additional date format to the end of the format string (separated by two pipe characters ||). For example, if the dates in your data look like 12/25/2018 01:23:45, you would use the following format string:

"date_field": {
    "type": "date",
    "format": "yyyy-MM-dd||dateOptionalTime||E MMM d HH:mm:ss zzz yyyy||MM/dd/yyyy HH:mm:ss"
}

For more information on date format mappings, please see the Elasticsearch documentation.

Elasticsearch Keyword Fields

We recommend that any string field not containing document text (including news articles, social media posts, or any multi-sentence text field) should have the type keyword. For example, fields of names, links, categories, and alphanumeric IDs should all have the type keyword.

"name": {
    "type": "keyword"
}

Elasticsearch Text Fields

Text fields should have the fielddata property set to true. For example:

"social_media_post": {
    "type": "text",
    "fielddata": true
}

Elasticsearch Data Ingest Tips

Data file format: define individual JSON objects on separate lines in the file. Example:

{ "_index": "index_name", "_type": "index_mapping_type", "_source": { "whatever_field": "whatever values" }}
{ "_index": "index_name", "_type": "index_mapping_type", "_source": { "whatever_field": "more values" }}

CURL Mapping file format: start with the "properties". Example:

{
  "properties": {
    "whatever_field": {
      "type": "whatever_type"
    }
  }
}

CURL [ES6] Create an index:

curl -XPUT hostname:port/index_name
curl -XPUT hostname:port/index_name/_mapping/index_mapping_type -H "Content-Type: application/json" -d @mapping_file.json

CURL [ES7] Create an index:

curl -XPUT hostname:port/index_name
curl -XPUT hostname:port/index_name/_mapping -H "Content-Type: application/json" -d @mapping_file.json

CURL Delete an index:

curl -XDELETE hostname:port/index_name

ELASTICDUMP Mapping file format: start with the index name. Example:

{
  "index_name": {
    "mappings": {
      "index_mapping_type": {
        "properties": {
          "whatever_field": {
            "type": "whatever_type"
          }
        }
      }
    }
  }
}

ELASTICDUMP Create an index:

elasticdump --type=mapping --input=mapping_file.json --output=hostname:port/index_name

ELASTICDUMP Ingest data into an index:

elasticdump --type=data --limit=10000 --input=data_file.json --output=hostname:port/index_name

Technical Stack

The NUCLEUS Data Server is a Spring Boot WebFlux Java application built using the Gradle plugins.

Modules

The application is built using multiple custom modules:

The server module contains the core Data Server code.
The sqladapter module contains the SQL JDBC datastore adapter that currently supports MySQL and PostgreSQL.
The esadapter module contains the Elasticsearch REST datastore adapter.
The common module contains the common adapter and model classes.

Change the modules included in your build (to add or remove adapter dependencies) by editing the gradle.properties and settings.gradle files.

Spring Documentation

Architecture Documentation

Field Types

NUCLEUS returns the following field types from its queryservice/fields/types endpoint:

boolean
date
decimal
geo
id
integer
keyword
object
text

Query Objects

NUCLEUS saves datastore-agnostic query objects in a SQL-like JSON syntax that can be (de)serialized to/from Java classes.

Notes on NUCLEUS query JSON/Java objects:

Fields are always represented as JSON objects containing their corresponding "database", "table", and "field" names.
The "selectClause" is always required.

More info coming soon!

Query Example:

{
  "selectClause": {
    "database": "testDatabase",
    "table": "testTable",
    "fieldClauses": []
  },
  "whereClause": {
    "type": "where",
    "lhs": { "database": "testDatabase", "table": "testTable", "field": "testWhereField" },
    "operator": "!=",
    "rhs": null
  },
  "clusterClause": {
    "count": 0,
    "type": "testType",
    "clusters": [[0, 25], [26, 50]],
    "aggregationName": "testAggregationName",
    "fieldType": "testFieldType",
    "fieldNames": ["testFieldName1", "testFieldName2"]
  },
  "aggregateClauses": [
    { "type": "field", "operation": "count", "label": "testAggregateLabel", "fieldClause": { "database": "testDatabase", "table": "testTable", "field": "testAggregateField" } }
  ],
  "groupByClauses": [
    { "type": "field", "fieldClause": { "database": "testDatabase", "table": "testTable", "field": "testGroupField" } }
  ],
  "orderByClauses": [
    { "type": "field", "fieldClause": { "database": "testDatabase", "table": "testTable", "field": "testOrderField" }, "order": -1 }
  ],
  "limitClause": { "limit": 12 },
  "offsetClause":{ "offset": 34 },
  "joinClauses": [],
  "isDistinct": false
}

Apache 2 Open Source License

NUCLEUS is made available by Next Century under the Apache 2 Open Source License. You may freely download, use, and modify, in whole or in part, the source code or release packages. Any restrictions or attribution requirements are spelled out in the license file. NUCLEUS attribution information can be found in the LICENSE file. For more information about the Apache license, please visit the The Apache Software Foundation’s License FAQ.

Contact Us

Email: neon-and-nucleus-support@caci.com

Name		Name	Last commit message	Last commit date
Latest commit History 299 Commits
.vscode		.vscode
common		common
esadapter		esadapter
gradle/wrapper		gradle/wrapper
server		server
sqladapter		sqladapter
.gitignore		.gitignore
CODEOWNERS		CODEOWNERS
Dockerfile		Dockerfile
Jenkinsfile		Jenkinsfile
LICENSE		LICENSE
README.md		README.md
build.gradle		build.gradle
gradle.properties		gradle.properties
gradlew		gradlew
gradlew.bat		gradlew.bat
runLocal.sh		runLocal.sh
settings.gradle		settings.gradle

License

NextCenturyCorporation/nucleus-data-server

Folders and files

Latest commit

History

Repository files navigation