Batch Server Monitoring

Recommind has implemented a JMX interface to access general metrics about document processing jobs as well as detailed information on a single document’s processing.

The JMX API is compatible with current monitoring tools.

Batch servers are used for document processing jobs triggered, for example, by conversion, redaction, production, printing, or OCR jobs.

The JMX API provides these types of information:

  • Job statistics contain general metrics about single document jobs. This is the main information source to detect high work load and freezes. You can periodically call statistics information with a third party monitoring tool.
  • Job information contains the information about a single document’s processing. It allows you to identify the index engine and the Axcelerate Review & Analysis job that requested a document’s processing, and find related log file entries.

    Important: Never call job information periodically with a monitoring tool. The data volume might block the system.

Note: The jobs monitored in the JMX API are single document jobs. They do not match the jobs you see on the Jobs page. Often, a single document job matches a job step on the Jobs page. But there are also cases where several single document jobs are required for a job step. This may be the case, for instance, if a conversion exception slip sheet is created after conversion failed, or if, during production, the first 10 pages of a large document are sent to a different batch server than the next pages.