The Linux Scheduler extends the reach Nathean Analytics to collect data from Linux based devices. Moreover it is redesigned to operate in a full disconnected mode with only occasional access to a Nathean Analytics server.
Enhancements are based upon extending the Multi-source Remote connections
Remembering that this existing functionally allows the configuration of a single data-marting dataset which extends over a range of database connections and is executed locally on each machine by the Nathean Schedule Service.
We set up the schedule service and provide it with the scheduler with the db_connection_id for which it’s responsible.
Currently the service is set up to only process for one db_connection_id but you can set up multiple instances of the scheduler on a single server, each to look after a single db_connection_id.
The enhancements include
- Added device management to describe remote devices
- Added device connection to DB Connection so that when executed we can refer to the device
- New Linux based Linux Scheduler Service (LSS) which runs on various Linux distributions including Raspberry PI device distributions. Initially the system supports the following data source types – SQLLite, Python Dataset, Text Log File, Json datasets
- The new scheduler operates in a disconnected mode and will occasionally sync up with the server for syncing the object definitions and the data prepared for datamarting. This is a fundamental change to how the Windows Scheduler Service (WSS) operates. The WSS is always connected to the server. When it needs the object definition, it makes an SQL call to the server database and brings the definition into memory. The LSS has a sync process which periodically connects to the server. If available, it checks for available object definition updates and uploads data cached locally.
- Information about the operating system is recorded in the Device definition. The database connection type will .indicate supported types.
- Currently the multisource data connectors communicate with the Nathean server using SQL. The new Python based service will connect using the REST API and will operate in a disconnected mode assuming only an occasional connection back to the server, so it will download the schedule and dataset definition it needs to process locally in it’s own SQLLite database and when the schedule is executed, it will process the dataset query and stored this in the local SQLLite database.
When it later is able to connect to the Nathean server it will then pass the data up.
- The installer/upgrader will be based upon APT (Advanced Packaging Tool) which supports initial deployment to the device and then looks after being upgraded from that point onwards when it finds new versions are available.