10:23 a.m.
Hi,
With WildFly 24 almost out, now might be a good time to revisit our community
documentation at https://docs.wildfly.org.
We are lacking a Cloud-oriented guide that describes how to build and deploy WildFly
applications on Kubernetes. We have a whole ecosystem around this (operator, helm charts,
S2I images, Bootable Jar + JKube) but the documentation is spread around all these smaller
projects and we don’t have something that gives the high-level picture.
I would like to provide a “Getting Started on the Cloud” guide to fill that gap.
However I’m not sure where we could write this guide. I don’t think it belongs to
https://github.com/wildfly/wildfly/tree/master/docs.
So another related thing I would like is to evaluate using https://antora.org to manage
our community documentation at https://docs.wildfly.org.
This documentation site generator is able to aggregate documentation across multiple
repositories and provide versioned documentation.
It is used by Smallrye for their docs: https://smallrye.io/docs/index/index/index.html
As you can see each individual projects maintains its documentation but they are all
aggregated in a single place.
There is also the ability to access different version of the documentation:
https://smallrye.io/docs/smallrye-metrics/3.0.0/index.html
https://smallrye.io/docs/smallrye-metrics/2.4.0/index.html
I’m envisioning to restructure our community documentation with something like:
docs.wildfly.org (generated by Antora)>
|
+- WildFly (from https://github.com/wildfly/wildfly docs)
|
+- Galleon (from https://github.com/wildfly/galleon)
|
+- Bootable Jar (from https://github.com/wildfly-extras/wildfly-jar-maven-plugin)
|
+- WildFly on the Cloud
| |
| +- Getting started Guide (from I don’t know where)
| |
| +- WildFly S2I (from https://github.com/wildfly/wildfly-s2i
| |
| +- WildFly Operator (from https://github.com/wildfly/wildfly-operator)
| |
| +- Helm Chart for WildFly (from https://github.com/wildfly/wildfly-charts)
|
+- Other WildFly-related documentation/guide (e.g Elytron, Clustering)
The advantage of that approach is that each project manages its doc individually and do
not have to a kind of Big Bang release when we deliver WildFly releases.
We still have to provide some consistency though.
We could also have a dedicated section for task-oriented guides:
* Connect to Keycloak on Kubernetes
* Clustering Guide for the Cloud
* Integrated with Apache Kafka
* ...
It’s not clear to me how these guides relates to our quickstarts though…
What do you think? Is this something worth investigating?
Jeff
--
Jeff Mesnil
Principal Software Engineer
Red Hat
http://jmesnil.net/