Hub uses configuration files to establish the required source connections, processes and target connections for data being transmitted via a TCP/IP connection from a z/OS system.
There are two sets of files that define a Hub pipeline, a file to define the elements of the pipeline and a file per pipeline element (Source, Process, or Target). All these files are defined in JSON format and are stored in the Hub configuration directory and/or subfolders of this directory.
For example, here is a list of the Component files that would be required if Ironstream is sending SMF Type 110 records to Hub from one source. All the records are to be sent to one Splunk Cloud server and only records for transactions starting with “PRO” are to be sent to another Splunk Cloud server:
- Pipeline configuration file.
- Source configuration file for IBM Z Source.
- Process configuration file that looks for CICS SMF 110 records.
- Process configuration file that looks for transactions matching “PRO*”.
- Target configuration file that sends to a Splunk Cloud server A over TCP/IP.
- Target configuration file that sends to a Splunk Cloud server B over TCP/IP.
The configuration files are subject to validation. When Hub starts it will load the configuration files and check that they are in valid JSON format. If so, then each field will be verified that it contains valid values. If any configuration file contains invalid JSON or invalid fields, then Hub will write a message to its log detailing the error and will then shut down.
If a field has a default value it can be omitted from the configuration file.
Example configuration files are supplied in the Hub deployment package.