To deploy LEMA to a Docker system, use the deploy.py
tool, which requires Docker Compose.
Required software versions:
- Python 3, version 3.6 or later
- Docker, version 1.13.0 or later
- Docker Compose, version 1.10.0 or later
Log in to gain access to the Micro Focus IDOL containers on Docker Hub:
docker login -u microfocusidolreadonly
Configure the location of your IDOL License Server in config/base.env
, and grant the admin
role
in your License Server configuration to the host you will deploy the analysis
component to.
Add TLS certificates in config/https/
(see the Encryption
section).
Run the deploy.py
tool using Python. (Much like when running docker
or docker-compose
, you
may have to run it as a different user with sufficient permissions to manage Docker containers.)
python3 deploy.py --init auth entity filestore analysis audit api ui
With the default configuration, the LEMA UI will be available at https://localhost:8070
once the
system has started.
After the system has started, log in with a user that has the admin
role. The LEMA UI will then perform a one off initialization.
To show options and other usage information, run:
python3 deploy.py --help
All configuration files are in the config
directory. base.env
contains settings relevant to
multiple components, and, for example, api.env
contains settings relevant only to the api
component. Lines starting with #
are ignored, and these are used to explain the meaning of each
of the settings.
By default, the user-facing servers (authentication server, API, and UI) only accept encrypted
connections. For this to work, you must obtain TLS certificates and copy them into the config
directory. The required files are:
config/https/api/tls.key
: Private key for the API.config/https/api/tls.crt
: Server certificate for the API.config/https/auth/tls.key
: Private key for the authentication server.config/https/auth/tls.crt
: Server certificate for the authentication server.config/ui/auth/nginx.key
: Private key for the UI.config/ui/auth/nginx.crt
: Server certificate for the UI.
To use HTTP instead of HTTPS, for testing purposes only, run:
python3 deploy.py --disable-encryption --init auth entity filestore analysis audit api ui
note: changes to the encryption state of a deployed system require manual deletion of the realm in Keycloak before running
deploy.py
with the new state.
To resume a stopped LEMA system, or to apply changes made to configuration files, or to change which
components are deployed: run the normal command to deploy, but without the --init
argument:
python3 deploy.py auth entity filestore analysis audit api ui
You can deploy components on different hosts, or deploy some components separately using a
compatible implementation (read the deploy.py
tool help text for a list of components). For
example, to use an existing object storage server and deploy the audit database on a separate host,
configure hosts and ports in the files in config/
, and then run on separate hosts (note that
--init
need only be run once):
python3 deploy.py audit
python3 deploy.py --init auth entity analysis api ui
By default, the following ports are forwarded ('public' ports listen on all interfaces (0.0.0.0), while others listen on 127.0.0.1 only):
Component | Port | Public | Purpose |
---|---|---|---|
auth | 8000 | no | PostgreSQL database storing authentication server configuration |
auth | 8010 | yes | Keycloak authentication server (API and admin UI) |
entity | 8020 | no | Gremlin database storing application data |
entity | 8021 | no | ACI port of IDOL Content database backend for the Gremlin database |
entity | 8022 | no | Index port of IDOL Content database backend for the Gremlin database |
entity | 8023 | no | Port of Cassandra database backend for the Gremlin database |
filestore | 8030 | no | S3-compatible object storage |
analysis | 8040 | no | NiFi server for media analysis (API and admin UI) |
audit | 8050 | no | PostgreSQL database storing audit logs |
api | 8060 | yes | System HTTP API |
ui | 8070 | yes | LEMA UI |
Docker volumes are created with the prefix micro-focus-idol-lema_
, which can be changed using the
COMPOSE_PROJECT_NAME
setting. The following volumes are created:
Component | Volume name | Purpose |
---|---|---|
auth | auth-db-data | Authentication server configuration |
entity | entity-schemadb-data | Schema for application data |
entity | entity-storagedb-data | Application data |
entity | entity-indexdb-data | Search index for application data |
filestore | filestore-service-data | Uploaded and generated files |
audit | audit-db-data | Audit logs |
All containers connect to a Docker network called micro-focus-idol-lema_main
. The
micro-focus-idol-lema
prefix can be changed using the COMPOSE_PROJECT_NAME
setting.