Improved User Experience, Performance Gains on Large Datasets and more Developer Options

Release Notes November 2020 – January 2021

The focus of this release has been enabling end users to react and manage real-time streaming data in fewer key/mouse strokes. We’ve added easier column arrangement, smarter data formatting and better real-time search/filter tools. Data selection can now be used to drive complex real-time streaming analytics providing instant rollups on user-interested subsets of data, allowing users to absorb more information faster in rapidly changing market conditions.

We’ve made large performance improvements and eliminated the 2gb/128 column limit on user-driven batch events, meaning users can manage huge tasks on a nearly unlimited number of items at once.

As the adoption of AMI in critical use cases has grown, so have our developers tools grown with better features to identify real-time resource utilization and historical diagnostics. Finally, we’ve continued to evolve the look and feel of AMI to provide crisper-looking dashboards while maintaining the speed and performance users rely on.

User Experience:

• Real-time analytics on user selection —
Example: as a user selects rows in a table, provide real-time aggregation on their selection, even as deltas are streaming in.
• Draggable column arrangement in tables and trees (instead of using arrangement tool)
• A new quick filter input for columns enabling convenient searching/filtering
• User ability to override column formats — End users can now conveniently customize date/time and number formatting
• Added a DateField and DateRangeField input field for table cell editing
• Added option to sort text in autocomplete dropdown of text fields
• Login button disabled when clicked/pushed, shows waiting message while logging in


AMIDB:

• Added DIAGNOSE keyword — See memory usage on a column-by-column basis
• Can add/remove special reserved columns using ALTER clause
• Added DECORATE trigger — A faster alternative to the JOIN clause for the common case of blending across two streams of data without needing to produce a proper cartesian product.


Performance Enhancements:

• The disk defragmentation engine has been redesigned, increasing data throughput 4x
• A single transaction is unlimited in size; prior it had a to 2gb limit.
• Support for real-time updates to an unlimited number of columns; prior it was 128 realtime columns
• Fast AmiScript Compiler, handles query strings up to 2gb
• Minified core JavaScript files, reducing load time.
• Query Planner identifies large OR clauses and process as IN clause for much faster execution


Developer Tools:

• Stdout.log & stderr.log files are preserved/rolled on startup now (instead of being overwritten)
• Added Statistics for Memory and Number of Users — See historical usage
• Added Import/Export data mode — Easily copy datamodels across layout files now.
• Added license info in the help menu
• Allow diffing layout for unsaved changes
• Aded icons for user and memory stats in menu bar — See visual queue in header for resource utilization. Clicking on the icon will provide historical trends


Adapters:

• AMPS Datasource Adapter — Added protobufs support; prior, only real-time streaming adapter existed
• Apache Ignite Datasource Adapter
• HP Vertica DB Datasource Adapter


Configuration Options:

• ami.default.user.timezone — Set the default timezone for users (which users can then override in their preferences)
• ami.font.java.mappings — Assign aliases for font names
• ami.db.onstartup.ondisk.defrag — Allows defragmentation of on disk storage to be skipped, reducing startup time
• ami.web.http.slow.response.warn.ms — Log slow http network communication
• http.slow.session.lock.warn.ms — Log contention between realtime streaming data and user actions
• ami.slow.amiscript.warn.ms — Log slow amiscript methods

If you have any questions about these
new features or would like a demo,
please email info@3forge.com.