May 2022 Platform Release Notes

Here's what's new in the latest HERE platform release.

Highlights

Use Externalized Checkpoint to Activate Stream-5.0 Pipelines

Now you can activate your Stream-5.0 and newer pipelines using the latest Externalized Checkpoint available. This will allow you to recover your pipeline without losing the state of the pipeline if it unexpectedly fails.

Keep in mind that your pipeline code has to be changed to enable Externalized Checkpoints and since this is enabled for quick recovery of a failed pipeline, they will be available in the system only for 7 days. There are also limitations on when a Checkpoint (vs. Savepoint) can be used to activate a pipeline - please familiarize yourself with the Flink documentation around this.

This new functionality is available only through CLI and API.

Improved support for cache configurability, data pre-fetching and pre-loading 

We now offer high-level API functionality which enhances cache configurability, map data (tile) pre-fetching and pre-loading, that allows for considerable performance improvements with application startup. The API also introduces a new cache mechanism which allows re-using blobs that are shared between different catalog versions. This allows for a reduction in memory usage compared to the previous approach. Also, specific tiles can be evicted from the cache and replaced with other specific (or newest) versions.

All of this is offered with Java and Scala support in the Data Client Library and Location Library of the Data SDK for Java & Scala. Please note: this implementation is currently experimental and available for use with selected features only. Please refer to the Developer Guide for the Location Library under "High-Level API v2" for more information.

Added tutorials on developer.here.com for several typical application use cases

Using the HERE Data SDK for Java & Scala, and with the help of several new tutorials, your task to develop applications for typical use cases has just been made easier. Under this link you will find new tutorials for the following use cases:

• develop and debug stream and batch applications (using Apache Flink and Spark) on the HERE Platform, including how to monitor your application using Grafana, Splunk, Spark UI, and the Platform Billing Page
• calculate partition Tile IDs for certain geocoordinate area queries and levels using the HERE Tile Resolver
• retrieve weather conditions from the HERE Live Weather catalog
• retrieve real-time traffic from the HERE Real Time Traffic catalog
• create and manage subscriptions (notifications) to catalog and layer-level changes