Rosette Server is a Java-based service that offers Rosette as a locally-deployable package providing access to Rosette's functions as RESTful web service endpoints.
Rosette Server can be installed on Windows, Linux, or macOS. It can also be run as a Docker application.
Note
Rosette Server was previously named Rosette Enterprise.
Minimum system requirements
Important
Many installations will require more than 32 GB of disk space. A complete installation may require up to 90GB to install Rosette Server. The exact amount needed will depend on the endpoints and the languages installed.
Any installation including the /entities, /sentiment, /topics, or /relationships endpoints will require additional space.
x86_64 CPU with 4 or more physical cores
-
Basic Memory Requirements:
64-bit macOS, Linux, or Windows
64-bit JDK 11 or 17 installed (tested with OpenJDK)
-
The following commands must be installed on Linux
curl
netstat
ps
gettext
bash
-
Native OS Libraries, needed for some endpoints
Notice
Endpoints requiring native OS libraries.
/name-similarity, /name-translation, and name-deduplication when the language is Chinese, Japanese, Korean, Russian, or Arabic.
/sentiment when the option for using DNN model is specified ("options": {"modelType": "dnn"})
-
/morphology/*, /sentences, /tokens, when the endpoint is using a neural model. This includes:
When the language of the data is Hebrew and the option for using the DNN models is specified "options": {"disambiguatorType": "dnn"}.
When the language of the data is Korean and the option for using the DNN models is specified "options": {"modelType": "dnn"}.
/morphology when the language of the data is Indonesian, Standard Malay, or Tagalog and the morphoFeature is complete or parts-of-speech.
/entities when the language of the data is English, Arabic, or Korean and the option for using DNN model is specified ("options": {"modelType": "dnn"}).
/relationships.
Rosette’s memory consumption includes the JVM heap and memory mapping of on-disk files. The size of these vary depending on the endpoint(s) enabled in the instance.
Rosette Server’s data files are loaded into virtual memory. Some endpoints, such as /entities, involve a large amount of data. In order for Rosette to operate at its peak performance, we recommend that you reserve enough free memory to allow memory mapping of all our data files so that page misses are minimized at runtime.
To estimate the size for memory mapping, you can sum up the files in the unpacked roots folder in your installation.
Many of Rosette’s endpoints organize their data by language. So you can further refine your estimates if you know exactly which languages your input documents are in. Just look for the sub-folders and files under roots/<component-name>-<version> that carry a 3-character ISO 693 code and exclude those not applicable to you.
You should have sufficient amount of free space to unpack the application and data from your shipment. This could range from 1GB to 90GB. The package size will grow as product updates are released. The amount required is the unpacked directories and files from all *.tar.gz in the shipment package.
Rosette Server also requires space to hold logs and other temporary files. Logs can grow depending on the log level and the number of calls. If you have a log rotation mechanism in place, a couple of GBs should be sufficient. Otherwise, experiment with your call patterns and plan for growth accordingly.
You will receive an email containing all the files needed to install Rosette Server, for multiple operating systems and also for using Docker. The files you download depends on your operating system and type of install.
The Rosette license key file, rosette-license.xml, may be sent in the same email or a separate email.
Tip
Each endpoint has a supported-languages method which returns the languages supported by the endpoint in addition to your license status for the language. The method returns a boolean for the field licensed where true indicates that you are licensed for the particular language.
The release notes (rosette-server-release-notes-<version>.pdf) and this user guide (rosette-server-user-guide-<version>.pdf) are included with every shipment. Before downloading any other files, you can review the new features and bug fixes included in the release.
If installing using a Docker container, download only the Docker compose file (docker_compose.yml) and the license file (rosette-license.xml). Docker will download the remaining files. For offline installation, you will need to download all the files while connected. Those files will be in the last section of the shipment email.
You must run the installer while connected to the internet to download the files and create the volumes. Once the files are downloaded, you can run docker-compose up without be connected to the internet.
Docker - offline installation
If you are only connected to the internet for the file download, and will be performing the remaining install while not connected to the internet, you will download the Docker yaml file and the all Docker images.
Download all the component tar files and import_docker_images.sh to create the docker volumes locally.
When installing on macOS or Linux, only download the installer (install_rosette.sh) and the license file (rosette-license.xml). The installer will download the remaining files.
Note
You must run the installer while connected to the internet to download the files. Once the files are downloaded, re-run the installer locally, without a connection, to resume the install.
The Windows install requires manually downloading install files, product files, as well as the component (root) files.
Custom Endpoint Installation
All shipments contain a file for implementing custom endpoints. It contains an installation script and the files for the reverse proxy and application server. The installation script is currently for Linux and macOS only.
The component files (roots) contain the endpoint specific models and data. Each shipment contains one or more root files, based on your license. You may receive roots that you did not order because of Root dependencies.
Table 1. Root Component Files
File Name |
Endpoints |
ascent-root-<version>.tar.gz
|
Sentiment Analysis |
nlp4j-root-<version>.tar.gz
|
Relationship Extraction and Syntactical Dependencies |
rbl-root-<version>.tar.gz
|
Morphological Analysis, Tokenization, and Sentence Tagging |
rct-root-<version>.tar.gz
|
Transliteration |
relax-root-<version>.tar.gz
|
Relationship Extraction and Syntactical Dependencies |
|
rex-root-<version>.tar.gz
rex-root-<version>-<lang>.tar.gz
|
Entity Extraction |
rli-root-<version>.tar.gz
|
Language Identification |
rni-rnt-root-<version>.tar.gz
|
Name Similarity, Translation, and Deduplication |
tcat-root-<version>.tar.gz
|
Categorization |
topics-root-<version>.tar.gz
|
Topics |
tvec-root-<version>-<lang>.tar.gz
|
Semantic Vectors and Similar Terms |
To minimize the size of your Rosette Server installation, the entity extraction (rex-root) and semantic similarity (tvec-root) components are shipped by language. The name of the language specific files contain the three letter ISO-639 language code, indicating which language is supported by the file.
If you receive any roots that are not part of your licensed endpoints, it's most likely because that root is a dependency for one of your licensed endpoints. For example, if you license entity extraction, you will also have the root for morphological analysis. The language identification root (rli-root) is shipped with many endpoints to determine the language of the request.
Table 2. Package Contents by Endpoint
Endpoint |
Package |
categories |
rbl-root-<version>.tar.gz
tcat-root-<version>.tar.gz
|
entities |
rbl-root-<version>.tar.gz
rex-root-<version>.tar.gz
|
language |
rbl-root-<version>.tar.gz
rli-root-<version>.tar.gz
|
morphology |
rbl-root-<version>.tar.gz
|
name-deduplication |
rni-rnt-root-<version>.tar.gz
|
name-similarity |
rni-rnt-root-<version>.tar.gz
|
name-translation |
rni-rnt-root-<version>.tar.gz
|
relationships |
nlp4j-root-<version>.tar.gz
rbl-root-<version>.tar.gz
relax-root-<version>.tar.gz
rex-root-<version>.tar.gz
|
sentences |
rbl-root-<version>.tar.gz
|
sentiment |
ascent-root-<version>.tar.gz
rbl-root-<version>.tar.gz
rex-root-<version>.tar.gz
|
syntax/dependencies |
nlp4j-root-<version>.tar.gz
|
semantics |
rbl-root-<version>.tar.gz
tvec-root-<version>.tar.gz
|
tokens |
rbl-root-<version>.tar.gz
|
topics |
rbl-root-<version>.tar.gz
rex-root-<version>.tar.gz
topics-root-<version>.tar.gz
|
transliteration |
rct-root-<version>.tar.gz
|