The containers in the stateful set were ignoring the resource requests
when specified in `values.yaml`. Comment out the resources requests
in `values.yaml`, so we only set requests if the user specifically
enables them.
I verified this is working by running `kubectl describe pod POD | grep
cpu` on one of the created pods on the fix branch. It returned `cpu:
"100m"`. On the master branch, the same command returned nothing.
* Update dependency version and change zooKeeper variable to zookeeper
* Update requrement lock
* Change zookeeper services name to the new value
* Update Chart.yaml
Also, update the conditional logic in the mysql.fullname template to more
closely mirror the behaviour of other charts, i.e. if no fullnameOverride is
specified and the release name contains the chart name, use it as the full name.
* Add the tcp port definition to the unbound container
* Add the tcp port definition to the unbound service
* Wrap resources in with block to prevent rendering when empty
* Style: unindent remaining with blocks to left margin
* Bump chart version.
* de-capitalize all the variables of etcd
* remove unused variables
* refactor the image definition
* update the version
* remove trailing spaces
* bump a major version
* fix typo
* Update Chart.yaml
* disable pvc by default, to make the e2e test pass
* add persistent volume enabled
* fix align issue
* update helpers.tpl
* fix cpu and memory resources
* remove default function
* [incubator/horovod] support horovod
* fix lint check
* update the readme
* Separate files per resource
* add use secrets
* add env for using secrets
* move to stable
* change to offical docker image
* [stable/nginx-ingress] Update ports
* Make port configurations consistent
* Use named ports to refer to target ports from services
* Bumped chart version
* [incubator/keycloak] Add miscellaneous improvements
* Update Keycloak to 3.4.3.Final
* Use nc to check for PostgreSQL availability in order to avoid
RBAC configurations
* Allow additional volumes and mounts to be configured
* Allow custom pre-start script to be configured
* Improve affinity configuration
* Add security context
* Update _helpers.tpl
* Only add environment variables for Keycloak credentials on
install. They are only needed for creating the intial user.
This avoids an error in the logs
* Add pod disruption budget
* Make test more robust
* Update Postgresql dependency
* Align ingress configuration with best practices
* Fix for version comparison from strings to semver
See #3002 for more detail
* Include pre-releases in the semver ranges
This is important when testing against alpha and beta builds of
Kubernetes along with environments that use pre-releases to denote
things other than pre-releases (e.g., gke denotes the environment
with a pre-releases)
Use `secretRef` instead of `secretKeyRef`, because `secretKeyRef` is not
a valid key for `envFrom`.
Here are the relevant
[docs](https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.10/#envfromsource-v1-core).
I verified that is working by running `helm install --set
envFromSecret=SOME-EXISTING-SECRET` and verifying that pods successfully
launched on this branch, but did not successfully launch from the master
branch.
* [stable/keel] Add support for SLACK_BOT_NAME
Allow to specify a Bot name different from the default "keel".
* add slack botname to README
* bump release version and app version to latest release
* STOP SLAVE IO_THREAD before changing master in xtrabackup container
* Remove slave existing data before cloning
* incremented chart version
* fix typo
* Updates redis-metrics service
This updates the field specs to be `metrics.service.annotations`, instead
of `metrics.annotations`, to be more in line with how the `master.service`
is configured.
This commit also fixes documentation and values.yaml references to the old
service annotation configurations.
* Increases chart version
* Fixes duplicate key issue
* Updates chart version (new major)
* Removes metrics endpoint from slave service
* Changes chart to update minor version