SmartOps24x7 Intelligent Agent Platform Architecture

The major components of Smart Operations Management SmartOps24x7™ are are as follows:

1. The Control database which contains all the scheduling data for the Intelligent Agent processes (IAPs) plus all the data to dynamically construct and manage IAPs.
2. A web-browser interface through which system administrators can schedule and manage IAPs by interacting with the Control database.
3. An IAP Launcher, which is a light weight system process which schedules, assembles, runs and monitors IAPs depending on the contents of the Control database.
4. An Operations Management database which is used by IAPs to exchange operational data by dropping off and retrieving the latest updates to operational status and history.
5. An extensive set of error handling mechanisms, which avoid the need for Python script writers to have to worry about handling errors, such as due to dropped internet connections.
6. Support mechanisms, such as the automated translation of data formats from different source systems and to those of different target systems.

A critical concept in enabling SmartOps24x7 to run 24x7, for months on end, is the philosophy that when an IAP finished handling the latest updates from its designated data sources, it kills the process, thereby freeing up the system resources that it has used.

Most Microsoft operating system processes fail to return resources, which are small chunks of memory, as they run. As a result, the computers run out of memory and have to be rebooted under normal circumstances. This is especially true of processes that have to communicate with external devices over unreliable communication channels, such as the internet.

By killing each IAP process as it completes handling a batch of new data and then rescheduling the process to run again, on a periodic basis, SmartOps24x7 is able to conserve resources and keep running for a long time.

Also, the Launcher monitors IAP execution to detect whether it is hung or taking too long and then kills the process, again freeing up resources. This can be a problem with communicating with external systems over unreliable communications channels or using obsolete communications interfaces or where the external system may fail unexpectedly.

As the IAPs are run dynamically, this enables their controlling Python scripts to be changed when needed without needing to shut down the data transfer mechanism or recompile any code.  

Please click here to learn more about the SmartOps24x7 Intelligent Agents.