Product roadmap

Adopting a new technology, especially foundational software like a database engine—even a virtual one like Kubling—is a long process that involves some degree of uncertainty.

As developers and DevOps professionals, one thing we often value, but don’t always have, is a clear idea of where the products we rely on are heading. This is why we’re sharing our high-level roadmap with you.

Please keep in mind that we only list features here that might directly impact you, and not all of them will necessarily be implemented.

Support Raw-Text Data Source Formats Analysis High Prio

We are working on a new data source that will allow Kubling to load files from various origins (e.g., local filesystems, HTTP, S3, Azure Files) in formats such as CSV, Excel, and JSON. Kubling will create lightweight, ephemeral indexes for fast searches, including vector support, providing powerful data access without the need for a full database load.

Support all PostgreSQL Metadata Objects High Prio

We plan to support missing PostgreSQL metadata objects (tables, views, and columns) in order to enable connectivity with a wide variety of tools, including some legacy systems where adding the Kubling driver is not an option, such as PowerBI and Spotfire.

Create a new GraphQL Data Source Analysis High Prio

We are currently developing a new Data Source for GraphQL, which is expected to make a significant impact. Many of the existing Script modules that do not require manual interception will likely be transitioned to GraphQL. Since the new Data Source will be statically compiled, it will offer performance comparable to that of the Kubernetes Data Source, ensuring both efficiency and scalability.

Nov-24 Update: Initial tests revealed the complexity of creating a GraphQL Data Source, as APIs do not typically adhere to a standard for schema generation, complicating query formation. We will soon decide whether to continue with this approach or to let users implement it within Script Modules.

Support for ElasticSearch Data Source Analysis

We are currently evaluating the feasibility of supporting ElasticSearch as a data source without importing specific client libraries, in order to avoid increasing the binary size.

File Logs High Prio

We plan to support saving logs in files, with the ability to define some rolling strategy.

Log Routers Implemented

We are analyzing whether to support routing logs, based on a predefined criteria, to custom script functions.

SQL functions defined in modules

Currently, SQL functions are defined in specific modules, forcing to have everything centralized. We plan to support defining SQL functions also in Script Bundle Modules.

Smaller kdv CLI

We’ll release a smaller version of the CLI without the test command, which requires lots of libraries to be added to the utility.

Load DDL from modules in Integration tests Analysis

Loading DDL files defined inside modules is not yet supported. We are currently analyzing whether the integration tests should read the internal module metadata, or if the ddlFilePath property should support schemes like module: and/or bundle:.

Upgrade Hibernate dialect to version 7 Low Prio

Support Couchbase Data Source Implemented Low Prio

We will support Couchbase, possibly with some limitations, in the next a future release.