[geomesa-users] GeoMesa 2.1.0 released

[Jim Hughes <jhughes@xxxxxxxx>]

Hi all,

The GeoMesa team here at CCRi is happy to announce the release of 
LocationTech GeoMesa version 2.1.0. Artifacts are available from Maven 
Central and on GitHub[1]. The release notes and upgrade guide provide 
key details on what has changed[2]. LocationTech GeoMesa 2.1.0 works 
with GeoServer 2.12.x.

Of note, GeoMesa 2.1.0 adds support for Spark 2.3 and Kafka 2.0. 
GeoMesa’s converter library, which is designed to ease the integration 
of new data sources into an enterprise, features new schema inferencing 
support[3], which means that you no longer need to define your data 
format before ingesting. Improvements to the Kafka DataStore allow for 
‘event-time’ message ordering and log compaction[4].

In terms of storage, there have been a number of improvements. Of note, 
GeoMesa’s FileSystem DataStore now fully supports non-point geometries 
using the Apache ORC file format. For situations where data is spread 
across multiple back-end databases, the new Merged DataStore View 
provides a single point of access[5].

For users working with Accumulo, Cassandra, and HBase, one can now 
manage data ‘partitions’[6]. For instance, entire partitions can be 
easily aged off. Tables which are not going to receive further writes 
can also be tuned to use less memory for write caches. Compactions 
should be smaller and less resource intensive. As this is a new feature, 
there may be room for new ideas and uses!

For GeoMesa 2.2 and beyond, we are continuing efforts to improvement and 
easy data management.  We plan to support Spark 2.4 in GeoMesa 2.2.  In 
the geospatial Java ecosystem, there are a number of interesting things 
going on. Of note, GeoMesa 2.2 will upgrade to GeoTools 20.x which uses 
LocationTech JTS 1.16. Changing from JTS 1.14 to 1.16 will require 
library users to upgrade their code to handle the move from the 
com.vividsolutions.jts package names to the org.locationtech.jts ones.

After that, the upcoming GeoTools version 21.0 (and the corresponding 
GeoServer version) will be compatible with Java 11. GeoMesa’s support 
for Java 9 and beyond will require at least that upgrade (and likely 
some version upgrade for Hadoop and Spark).

Cheers,

Jim

1. Binary downloads: 
https://github.com/locationtech/geomesa/releases/tag/geomesa_2.11-2.1.0

2. Release notes: 
https://geomesa.atlassian.net/wiki/spaces/GEOMESA/pages/465174529/GeoMesa+2.1.0+Release+Notes 

Upgrade guide: http://www.geomesa.org/documentation/user/upgrade.html

3. Details about schema inference
https://www.geomesa.org/documentation/user/cli/ingest.html#ingest

4. Kafka:
Event time: 
https://www.geomesa.org/documentation/user/kafka/index_config.html#feature-event-time
Log Compaction:
https://www.geomesa.org/documentation/user/kafka/data.html#kafka-topic-compaction

5. Merged DataStore View
https://www.geomesa.org/documentation/user/merged_view.html

6. Partition Management:
https://www.geomesa.org/documentation/user/datastores/index_config.html#configuring-partitioned-indices 

https://www.geomesa.org/documentation/user/cli/schemas.html#manage-partitions-cli