The key point I want to get across with this guide is:
1. Who can upgrade (what breaking changes we made)
2. That there's no tooling to automate migration yet
3. That you can upgrade and start using v2 features anyway
Signed-off-by: Nic Cope <nicc@rk0n.org>
* Keep them concise (~50 chars) for rendering on index pages
* Adjust them for clarity and accuracy
* Add descriptions to pages that were missing them
Signed-off-by: Nic Cope <nicc@rk0n.org>
* Realtime compositions are beta now
Signed-off-by: Nic Cope <nicc@rk0n.org>
* Composition functions don't run on delete
Signed-off-by: Nic Cope <nicc@rk0n.org>
* You can't patch XR metadata
Signed-off-by: Nic Cope <nicc@rk0n.org>
* Document function response cache
Signed-off-by: Nic Cope <nicc@rk0n.org>
---------
Signed-off-by: Nic Cope <nicc@rk0n.org>
* Drop note about (only) AWS MRs supporting v2
A few of the other providers like GCP and Azure now do. The rest will
fast follow.
Also, it's not misleading to say the others don't support v2. That reads
like they're incompatible, which they're not. They're just not
namespaced yet.
Signed-off-by: Nic Cope <nicc@rk0n.org>
* Only use modern namespaced provider-upjet-aws MRs
* Make sure they're using v2.0.0 providers
* Make sure they use m.example.org style API groups
* Make sure they're namespaced
Signed-off-by: Nic Cope <nicc@rk0n.org>
---------
Signed-off-by: Nic Cope <nicc@rk0n.org>
Large Crossplane providers install hundreds of CRDs consuming significant API
server resources, even when users only need a few resource types. The new
alpha ManagedResourceDefinition (MRD) and ManagedResourceActivationPolicy
(MRAP) features in Crossplane v2.0+ solve this by enabling selective
activation of provider resources, but lack user-facing documentation.
This change adds complete documentation covering both concepts and practical
implementation. The MRD concepts page explains the CRD scaling problem and
selective activation approach. The MRAP concepts page details pattern-based
activation strategies and multiple policy coordination. Two how-to guides
provide end-to-end workflows: one for users wanting to reduce CRD overhead
through selective activation, and another for provider developers implementing
safe-start capability. The user guide was tested with simulated MRDs to verify
the activation workflow and troubleshooting steps.
Signed-off-by: Nic Cope <nicc@rk0n.org>
Nic Cope
Jason Tang
Jared Watts
Matt Welke
Craig D Wilhite
Mark Anderson-Trocme
Steven Borrelli
Yury Tsarev
Pierluigi Lenoci
dalton hill
Cem Mergenci, PhD
Se7en
Sharp
Pete Lumbis
Pete Lumbis
Crossplane