Add SNI support in release notes for 0.8 (#1389)

* Add SNI support in release notes for 0.8

* Add SNI support in release notes for 0.8

* Add SNI support in release notes for 0.8

* Removed TBD.

* envoy apis
This commit is contained in:
Andra Cismaru 2018-05-31 16:18:29 -07:00 committed by istio-bot
parent a37fc3bf81
commit c245855c36
1 changed files with 8 additions and 4 deletions

View File

@ -14,13 +14,17 @@ This model adds many new features and addresses usability issues
with the prior model. There is a conversion tool built into `istioctl` to help migrate your config from
the old model. [Try out the new traffic management model](/docs/tasks/traffic-management).
- **Envoy V2**. Users can choose to inject Envoy V2 as the sidecar. In this mode, Pilot uses Envoy's new API to push configuration to the data plane. This new approach
increases effective scalability and should eliminate spurious 404 errors. [TBD: docs on how to control this?]
- **Gateway for Ingress/Egress**. We no longer support combining Kubernetes Ingress specs with Istio route rules,
as it has led to several bugs and reliability issues. Istio now supports a platform independent ingress/egress
[Gateway](/docs/concepts/traffic-management/rules-configuration/#gateways) that works across
Kubernetes and Cloud Foundry and works seamlessly with the routing rules.
Kubernetes and Cloud Foundry and integrates seamlessly with the routing rules.
- **The new Gateway supports [Server Name Indication](https://en.wikipedia.org/wiki/Server_Name_Indication)** based routing,
as well as serving a certificate based on the SNI value. HTTPS access to external services will be auto configured based on SNI.
- **Envoy v2 APIs**. Users can choose to inject a sidecar using Envoy's v2 APIs. In this mode, Pilot uses [Envoy's v2 Aggregated Discovery Service
API](https://www.envoyproxy.io/docs/envoy/latest/configuration/overview/v2_overview) to push configuration to the data plane.
This new approach increases scalability of the control plane.
- **Constrained Inbound Ports**. We now restrict the inbound ports in a pod to the ones declared by the apps running inside that pod.