Resiliency Monitoring and Config for of analytics offload

New in API Connect v10.0.5.5 - Analytics ingestion resiliencyAnalytics persistent queue feature updated for better resiliency when both internal storage and offload is configured. The offload processes are now separated from the internal storage processes.

This RFE is a mix of an defect and two actual enhancements. The two enhancements that will be reviewed for future consideration are:

• change buffer sizes of the logstash pipelines

• monitor buffer levels

The following part of the request is considered a product defects and removed from the RFE description:

With APIC v10.0.5.3 we use internal analytics and analytics offload in parallel, offloading analytics data of specific spaces only to API provider specific analytics systems.

At the same time, all other users still work with internal analytics.

We found, that if the receiving system on the offload pipeline cannot be reached, that internal analytics stops receiving data, too.

That's really annoying, as all other users of internal analytics are affected, in case of issues in the offload pipeline. Additionaly we wanted to use internal analytics as a fallback, in case the offload has some problems, which is not possible, given the limitations described.

We believe, that an enterprise-grade mission-critical system like API Connect should provide better decoupling and reliability.