This is the architecture that you will implement during this trial.The box in the center, the ec2 trial environment, represents the machine you will connect to. The environment includes an installation of HVR that will act as the hub, as well as the PostgreSQL 10 database software. PostgreSQL database hvrhub will be used to store the HVR repository tables, and database pgtgt will be a target for transactions flowing from the Oracle Database. You will use a browser to control the HVR GUI through a remote desktop environment that is running on the server. HVR users will typically run the HVR GUI application on their laptop or desktop (MacOS, Windows and Linux are all supported) but this is not required for this trial.
On the left is the source server ec2-source that runs Oracle 11.2 Express with a database XE that will be used as the source. Server ec2-source includes an installation of HVR to act as an agent, and the connection to the source will use AES256 encryption with a unique certificate.
On the right is the target server ec2-target that runs MariaDB 10.1. This server also includes an installation of HVR as an agent, with encryption configured to connect to the agent.
HVR’s architecture is modular and flexible, leveraging agents for optimum efficiency. This diagram shows the most distributed setup with separate HVR installations for capture and integration.
HVR uses a distributed architecture with installations of the HVR software (agents) that are located as close as possible to the source and destination technologies. The use of agents optimizes network communication, distributes load, and secures data on the wire. All HVR capabilities, from initial load through replication through compare/repair, take advantage of the agents.
The HVR Hub – another installation of the same HVR software – runs a scheduler whose main task is to start jobs that perform continuous capture and integration as well as, periodic or one-time, refresh and compare/repair jobs. The scheduler automatically restarts jobs if they fail and gathers all logs so you have a single location for monitoring and troubleshooting. The scheduler requires a connection to a repository database to maintain high level job status information.
The HVR GUI is a client application that can either run remotely or on the hub machine itself. Through the hub the GUI also connects to the repository database to store connection information for end points as so-called Locations, and data replication definitions in so-called Channels. The GUI also provides high-level monitoring capabilities on the scheduler, and access to Insights including a topology overview and a historical view into replicated data. All interactions with the HVR hub, through the GUI or through the command line, are recorded as audit records in an event table to facilitate change tracking and regulatory compliance.
Any single installation of HVR can play multiple roles, namely capture, integration, act as a hub, and serve up the GUI.