208.4.2 Performance-based Scheduling

Etere DataMover can schedule jobs based on the performance counters of the server; this feature prevents the execution of jobs (e.g. transcoding) when the Concurrency limit or CPU usage is reached.

The system manages the performance counters to determine the system load status; it is displayed in the top part of the server (File > Show/Hide performance counters) to visualize a real-time system performance:

Real-time system performance

Machine total CPU - Total CPU amount currently used by all servers and the higher peak reached.
DM CPU - Specific CPU currently used by Etere DataMover, and higher peak reached.
DM Private bytes - The Amount currently used by Etere DataMover for internal processing and the higher peak reached.
DM I/O Bytes - Amount of bytes currently used by Etere Media Manager in input/output transfers and the higher peak reached.
DM Handles Count - Amount of handles currently consumed by Etere Media Manager and the higher peak reached.

The performance-based scheduling of jobs can be configured under the Configuration > Options section. In the Schedule section, you can set the database fetch interval, job check interval, and max. Retry on copy error, etc.

Load Balancing
Load balancing uses a data dispatcher to manage the processes in the system. The data dispatcher maintains a list of functions to be executed and assigns available system resources to accomplish the process. The data dispatcher works hand-in-hand with the scheduler; when a process is scheduled, the data dispatcher initiates the process and gives system resources to the process. With a data dispatcher, the maximum CPU time is allocated to each process in the system to achieve maximum throughput by minimizing the latency and response time to complete the process.  

The benefits of load balancing can be summarized into the following:
•Load balancing is based on the real-time monitoring of servers; it allows the distribution of processes and avoids overwhelming servers.
•Load balancing equally distributes processes based on the resource pool designed during T-workflow generation.
•Load balancing is available for Etere data management functions and services, including the following:

•Data Moving
•Transcoding
•Quality Control
•Export
•Conforming
•Closed Captions

•Load balancing is also available for integrated third-party components.