Merge branch 'pull_in_latest_upstream'

CHANGELOG.md

@@ -5,6 +5,16 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/)
 and this project's packages adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.html).
 
+## [v0.5.0]
+
+## Changed
+
+- Update the chart to latest [upstream version 1.0.0](https:/Kong/charts) (#043c2c2)
+
+## Updated
+
+- README contains some more information
+
 ## [v0.4.6]
 
 ## Added

README.md

@@ -2,9 +2,141 @@
 
 # kong-app chart
 
+[Kong for Kubernetes](https:/Kong/kubernetes-ingress-controller) is
+an open-source Ingress Controller for Kubernetes that offers API management capabilities
+with a plugin architecture.
+
 Giant Swarm offers a Kong Managed App which can be installed in tenant clusters.
-Here we define the Kong chart with its templates and default configuration.
+
+## Configuration
+There are 3 ways that Kong can be configured in Kubernetes:
+
+1. Using a Database (PostgreSQL or Cassandra)
+1. Kong Kubernetes Ingress Controller (with CRDs)
+1. DBless (no databases and static Configuration file)
+
+The recommended and supported way is to use Ingress Controller (which works
+alongside DBless).
+
+This app does not by default provide a database and if a database is required,
+then you will need to BYOD (Bring Your Own Database). For testing purposes, it
+is possible launch postgres alongside this App (described below).
+
+For detailed configuration options, please refer to the [configuration list](helm/kong-app/README.md#configuration)
+also the [`values.yaml` file](helm/kong-app/values.yaml)
+
+Any key value put under the `env` section translates to environment variables
+used to control Kong's configuration. Every key is prefixed with KONG_ and
+upper-cased before setting the environment variable.
+
+### Kong Ingress Controller (Recommended)
+The default installation of the App will use Kong Ingress Controller. This
+method uses CRDs to configure various aspects of Kong. Please refer to Kong
+Ingress Controller
+[documentation](https:/Kong/kubernetes-ingress-controller/tree/master/docs)
+for more detailed explanation and usage.
+
+### DBLess Kong
+The [official documentation](https://docs.konghq.com/1.4.x/db-less-and-declarative-config/)
+explains how DBLess Kong works and the possible limitations. To use this method
+of operation with this App, you will need to include the following you your
+values YAML:
+
+Using an existing ConfigMap
+---------------------------
+```YAML
+env:
+ database: "off"
+ingressController:
+ enabled: false
+dblessConfig:
+ configMap: nameOfExistingConfigMap
+```
+
+Inline configuration
+--------------------
+```YAML
+env:
+ database: "off"
+ingressController:
+ enabled: false
+dblessConfig:
+ config:
+ _format_version: "1.1"
+ services:
+ - name: example.com
+ url: http://example.com
+ routes:
+ - name: example
+ paths:
+ - "/example"
+```
+
+### Using your own Database
+Kong supports two databases:
+
+- PostgreSQL: 9.5 and above.
+- Cassandra: 2.2 and above.
+
+This section will focus on PostgreSQL, however the same process can be equally
+applied to Cassandra.
+
+Example database configuration:
+```YAML
+ingressController:
+ enabled: false
+env:
+ database: "postgress" # can be "off" or "cassandra"
+ pg_host: 127.0.0.1
+ pg_user: postgres
+ pg_password: # This can also be a string value, but not recommended
+ valueFrom:
+ secretKeyRef:
+ key: kong
+ name: postgres
+```
+You can use any valid database configuration option inside `env`. This is also
+how Cassandra can be configured.
+
+#### Installing a database alongside the App
+For testing purposes, it is possible to install a PostgreSQL server alongside
+the App. To do this:
+
+```YAML
+postgresql:
+ enabled: true
+ postgresqlUsername: kong
+ postgresqlDatabase: kong
+ service:
+ port: 5432
+```
+(There is no need to add PostgreSQL configuration data to `env`)
+
+Please note: This configuration should only really be used for testing and it
+not something we can support
+
+### Using Kong Ingress Controller with a Database
+When using Kong Ingress Controller, their shouldn't be a need to use a Database.
+However in some cases (for example, plugin support) a data store is
+required. In these situations, it is possible to use a mixture of Kong Ingress
+Controller and a database.
+
+Like in the case of using a database, Giant Swarm do not support the database and
+can only provide best efforts support with this configuration.
+
+To configure, please see ['Using your own Database'](#using-your-own-database).
+But ensure that that `ingressController.enabled` is set to `true`.
+
+Note:
+> Is it possible to create consumers using the Admin API?
+>
+> From version 0.5.0 onwards, Kong Ingress Controller tags each entity that it
+> manages inside Kong's database and only manages the entities that it creates.
+> This means that if consumers and credentials are created dynamically,
+> they won't be deleted by the Ingress Controller.
+(Taken from https:/Kong/kubernetes-ingress-controller/blob/master/docs/faq.md#is-it-possible-to-create-consumers-using-the-admin-api)
+
 
 ## Credit
 
-* https:/helm/charts/tree/master/stable/kong
+* https:/Kong/charts/

helm/kong-app/FAQs.md

@@ -0,0 +1,32 @@
+# Frequently Asked Questions (FAQs)
+
+#### Kong fails to start after `helm upgrade` when Postgres is used. What do I do?
+
+You may be running into this issue: https:/helm/charts/issues/12575.
+This issue is caused due to: https:/helm/helm/issues/3053.
+
+The problem that happens is that Postgres database has the old password but
+the new secret has a different password, which is used by Kong, and password
+based authentication fails.
+
+The solution to the problem is to specify a password to the `postgresql` chart.
+This is to ensure that the password is not generated randomly but is set to
+the same one that is user-provided on each upgrade.
+
+#### Kong fails to start on a fresh installation with Postgres. What do I do?
+
+Please make sure that there is no `PersistentVolumes` present from a previous
+release. If there are, it can lead to data or passwords being out of sync
+and result in connection issues.
+
+A simple way to find out is to use the following command:
+
+```
+kubectl get pv -n <your-namespace>
+```
+
+And then based on the `AGE` column, determine if you have an old volume.
+If you do, then please delete the release, delete the volume, and then
+do a fresh installation. PersistentVolumes can remain in the cluster even if
+you delete the namespace itself (the namespace in which they were present).
+

helm/kong-app/OWNERS

@@ -1,6 +1,8 @@
 approvers:
 - hbagdi
 - shashiranjan84
+- rainest
 reviewers:
 - hbagdi
 - shashiranjan84
+- rainest