Overview of Happenings in Teiid Land
There’s a lot happening across the various Teiid related projects. Here’s a brief overview of where things are at.
Teiid / Teiid WildFly
Teiid 13.1 was released a couple of weeks ago. That added support for transactional Infinispan sources, which has already been picked up by Teiid Spring Boot, and will be utilized by the Operator for turn-key materialization. There is also a new policy construct similar to PosgreSQL that allows you to specify row-based security per operation, rather than just for all.
We’re actively working toward Teiid 14.0. It will update to WildFly 19 support ideally with associated updates to the configuration and our documentation. I expect that we’ll address in total what is a new normal amount of issues for a release approximately 40 and target end of April / start of June for a release.
Teiid Spring Boot
Version 1.4 was just released. If you have been following the development you will have seen that the emphasis has been on supporting Teiid Syndesis, existing Teiid sources, etc. rather than on further development of Spring Annotations. 1.5 will continue this trend with Teiid core 14 and adding at least support for SOAP sources and connection pooling of non-JDBC sources.
With all the working going on with Teiid Spring Boot the natural question is - does that mean Teiid WildFly is going away?
The short and intermediate term answers are no. There are quite a few differences between the two projects due to functionality not being ported over, adapting to the Spring way of doing things, and fundamental changes (such as single VDB based deployments). For many users of Teiid WildFly it is simply still a bridge too far to switch as their development and management of VDBs is all geared toward WildFly. So for quite a while Teiid WildFly will be maintained - see above about the update to WildFly 19 - and Teiid Spring Boot will continue to mature and evolve. Longer term there has been an industry wide march toward MicroProfile which has also dovetailed at Red Hat with our Quarkus efforts. For Teiid WildFly / Thorntail we will keep reassessing if/when it makes sense to realign with that or if we should look again toward providing more OpenShift / Kubernetes collateral based upon Teiid WildFly.
There’s been a flurry of activity here, which Ramesh Reddy has been leading. The goal is to keep moving up the Operator Maturity Model and to a version that is applicable to a wide variety of Kubernetes environments, not just OpenShift. Eventually this enables a nirvana state deploying / managing Teiid on OpenShift - the operator will do almost everything for you. For example our most turn-key materialization feature yet is coming soon leveraging the Teiid 13.x work to support transactional Infinispan caches. The operator will simplify the cache cluster management and other steps necessary for that to work.
For a couple of months the basic mechanics of creating a virtualization with Teiid Syndesis have been functional and maturing. Now we’ve pivoting to a lot of new functionality - including a metadata tree to assist in creating views, UI and backend integration for data role management with Keycloak/Red Hat SSO, a new DDL editor with LSP support, and more. There are several other big efforts on the horizon, such as replacing the internal deployment of Teiid Spring Boot VDB images with management by the Teiid Operator, and more of data-centric approach of creating your virtualization (more pervasive and contextualized preview of actual data based upon your UI actions).
A documentation overhaul is still underway. We have one primary set of docs that serve for both Teiid and Teiid Spring Boot, but that require a lot of conditional logic to get the appropriate output given the differences between the two. We’re not yet promoting the Teiid Spring Boot specific docs much yet in the community, but you can seem them under development here: http://teiid.github.io/teiid-documents/master/sb/
There is another language server effort that has been started: https://github.com/teiid/teiid-language-server The reason for another project is that this is the starting point for supporting development with Visual Studio Code and without needing an instance of OpenShift with Teiid Syndesis deployed. Over time we’ll be able to determine how much our LSP efforts can share.
The website teiid.io has a few more changes pending after a large-ish update to clean up some of older content. We’re hoping to greatly expand Teiid splashier quickstarts (not just references to github), better integrated blogs, and further update the look.
Thank you, Steve