Talend Metadata Manager application server architecture
Talend Metadata Manager is a three-tier application.
- The Client tier, with a Web browser and a scripting client.
The scripting client allows you to automate administration tasks, such as create a folder and harvest a model upon changes, and execute model management operations, such as compare, migrate or stitch models. All basic capabilities available to the user interface are scriptable.
- The Server tier is a web J2EE application (WAR) that can run in a web container/application server. It uses a tomcat server by default and runs as a standard web application named MM. By default, this server is accessible on port 11480.
- The Database tier is a database used to store data and run computations on. As a rule,
Talend Metadata Manager uses the database to
perform computations on data that can potentially exceed the amount of available server
By default, this is a PostgreSQL database server where all repository information will be stored. Oracle and MS SQLServer are also supported.
When multiple users access Talend Metadata Manager simultaneously, the database can become the key feature for client performance. In this case, you should consider scaling up the database performance by adding more computational resources (such as CPUs) or speeding up the database storage (such as RAID, faster disks, etc.).
Talend Metadata Manager deployment architecture
- Talend Metadata Manager application server
- Database server
The database server should be in "close proximity" to the application server in terms of network connectivity, as there is a high degree of traffic between these two.
- One or more Metadata Harvesting agents
A metadata harvesting agent is a metadata harvesting server, installed on a machine remote to the application server. These are often necessary when it is not possible to install APIs, which are required for harvesting bridge function, on the application server itself.
In this case, a metadata harvesting agent is installed on this machine as well as the API for the metadata source technology (generally a client installation). Then, Talend Metadata Manager uses this machine as an agent to harvest from or export to that source technology.
All administrative and analysis activities are generally conducted from personal workstations with a supported Web browser.