Predix Edge Applications and Services Release Notes 12-19

Bug Fixes

This release contains the following bug fix.

Cloud Gateway Data Transmission: Resolved an issue that would see the cloud gateway stop sending data to the cloud when it encountered bad network conditions, necessitating a restart.

This release has the following known issues:

Duplicate Client IDs when Connecting to MQTT: Applications with no clientid configured may encounter an issue with duplicate client IDs when connecting to the MQTT broker resulting in the applications not working as expected. To avoid this issue, configure the clientid field either in the PEAS applications or in custom applications when connecting to the broker.
Excessive Logging Depletes Memory: Applications that generate excessive logs can create conditions where Docker (dockerd) consumes a large amount of memory. If memory usage is of concern, configure applications to produce fewer logs.
EGD Failure During Power Cycle: The EGD application may fail to start or deploy if a previous attempt to start/deploy was interrupted by a power cycle. The workaround for this issue is to reboot the device before attempting to start/deploy the application.

OPC UA and Timeseries Output Format: Users of opcuapollflat blocks should not use the flat_to_timeseries block to generate timeseries output as this will result in the loss of tag-based timestamps because the flat_to_timeseries format uses the first timestamp it encounters for the entire tag list and discards the rest.; Instead, the output_format field should be set to time_series.

Cloud Gateway Performance: When the cloud connection is re-established and there is a large store buffer, cloud gateway performance will be slow for a couple of minutes.

Raw Modbus Data

Raw modbus data passed from an adapter into a raw modbus sink will fail with an error similar to:

Couldn't parse {"data":{"tagName":{"val":19}},"timestamp":1562096019014}: std::exception

Raspberry Pi: Cloud Gateway on Raspberry Pi is known to fail after running for an extended period of time, due to insufficient memory on the Raspberry Pi device.
Note: The Raspberry Pi image remains unchanged from the 2.1.0 release.