Releases: rhinoman/odmp
Releases · rhinoman/odmp
OpenDMP 0.5.0
- Python and Clojure script processors now run as separate services.
- Improved error handling on service calls
- Bumped Keycloak version to 11
- Elasticsearch now supported as a data destination!
OpenDMP 0.4.0
Significant changes here
- Added Plugin support!
- Added Consul to the stack for plugin service discovery
- Started on a Plugin development library.
- Added a plugins folder to the repo.
- Some bug fixes
- Some code reorganization - created a parent POM.
Known issues:
- When ingesting from S3, if a plugin throws a 500 error, the flow will start over trying to re-ingest the same file. Ad infinitum. This can be caused either by a problem with the plugin itself, or corrupt data. I haven't observed this behavior when ingesting from a file drop.
OpenDMP 0.3.0
- S3 Ingest now supported. You can set up a S3 bucket and key prefix to "watch" for new data.
- Raw Data is now tagged when it enters a flow and history of individual data sets (i.e., what processors handle the data as it traverses the flow) are tracked.
- Updated to Apache Camel 3.5. This Camel bug has been fixed, which has allowed me to remove an ugly hack-around in the OpenDMP code.
- External Processor implemented - can execute external scripts and tools as part of a dataflow.
- Collection interface is more fleshed out
- Can delete collections
- Can examine properties and history of individual datasets (and delete dataset records as well)
OpenDMP 0.2.0
This is a good point to cut a (very alpha) release:
- S3 Destination for collect processors works (still can't ingest from S3 yet, working on it)
- Can download datasets through the UI now
- Python script processors work
- Datasets are now sortable and paginated
- Errors during runs are now displayed in the Dataflow view (requires refresh - need to implement push via SSE, or perhaps be lazy and do polling).
Still to do:
Lots