* [stable/concourse] Refactor values.yml to map concourse binary commands
* support concourse version 4.1
* update auth
* refactor credentialsManager and metrics sections in values.yml
* allow config concourse parameters through env insteand of args for
flexibility
Signed-off-by: Rui Yang <ryang@pivotal.io>
* [stable/concourse] add postgresql secrets for its sslmod
*remove check guard for required secrets
*typo fixes
Signed-off-by: Rui Yang <ryang@pivotal.io>
* [stable/concourse] fix prometheus port reference
Signed-off-by: Rui Yang <ryang@pivotal.io>
* [stable/concourse] quotes env vars that take a list of values
Signed-off-by: Rui Yang <ryang@pivotal.io>
* [stable/concourse] bump concourse version to 4.2.1
Signed-off-by: Rui Yang <ryang@pivotal.io>
* [stable/concourse] quote int variables
Signed-off-by: Rui Yang <ryang@pivotal.io>
* stable/concourse: use the correct work dir in the container preamble
There's a short preamble (written in shell) in the container that cleans
up before the container starts. It had the work dir hard-coded, while
its value has been configurable since a few chart versions.
Now it uses the same configured value.
Signed-off-by: Norbert Buchmueller <norbert.buchmueller@optiopay.com>
* stable/concourse: bump chart version to 1.16.1
Signed-off-by: Norbert Buchmueller <norbert.buchmueller@optiopay.com>
* stable/concourse: use the correct work dir in the volume's `mountPath`
Signed-off-by: Norbert Buchmueller <norbert.buchmueller@optiopay.com>
* Allow user to specify a static IP for the Concourse loadbalancer
Signed-off-by: Divya Dadlani <ddadlani@pivotal.io>
* update readme with loadBalancerIP entry; bump version to 1.15.0;
* update readme description of looadBalancerIP
* Add optional values for additional volumes and volumeMounts
* I have an addiction to additions...
* Add value to override working directory
* Bump minor instead of patch
* Add readme update for new attributes.
* Version bump to 1.15.0
* revert bump to 1.14.0
* [concourse] add nodeselector for web
* bump chart version
* typo
* update values.yaml and README
* bump chart version
* Bump minor version for new feature
* Add support for AWS Secrets Manager
* Rename credentialManager.secretsmanager to credentialManager.awsSecretsManager
* Bump minor version for new feature
Concourse 3.13.0 adds support for exporting metrics to Datadog (see
https://github.com/concourse/atc/pull/269). This adds support for the
new flags, and bumps the Concourse version.
* adding https option to use ACM in AWS and also whitelisting for service type LoadBalancer
* fixing check for range of IPs
* removing extraneous space
* adding description of value
* bumping version
* bumping to 1.6.0
* [stable/concourse] Add missing release labels to pods
This fixes worker podAntiAffinity which was selecting on the missing release label
* bump patch version of chart
* Create SA for web component (fixes#4767)
We have to create SA account even then kubernetes-in-cluster is disabled.
This change also covers the case then vault is enabled.
* Bump stable/concourse version to 1.2.2
* Fix for version handling with stringe vs semver
See #3002 for more details
* 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)
* Incrementing the chart version
* Fix Vault for `cert` and `token` auth
- `caCert` was renamed to `useCaCert` but not in README or values.yaml
- `vault-client-key` and `vault-client-cert` should only need to be set
when the auth mode has been set to `cert`. Without the check, `approle`
and `token` auth will fail (due to Go not being able to parse the PEM
of "")
* Bump minor
* Added additionalSpec to values and statefulset template
* conflict resolve
* Bumped chart version
* added space to comment
* replaced additionalSpec with more finegrained nodeSelector
* Make secrets optional, eliminate configmap.
Provide a way to opt out of managing secrets in helm values.
The configmap introduced needless repetition, mapping values directly to
environment variables is a simpler approach.
* Don't store postgresql user as a secret
* Add missing documentation for values introduced in this and previous PRs
Also fix encryption key values key in secrets, and rename credentialManager.vault.caCert to credentialManager.vault.useCaCert for clarity
* [stable/concourse] bump postgresql dependency to v0.8.11
* Remove redundant default template function calls
* Flatten secret values
Use a helper to produce more informative errors when secret values are required
* Add support for using k8s secrets as a credential manager.
Add rbac support for secret access from web pods to to team specific
namespaces.
Upgrade to Concourse 3.9.0
* Write the liveness probe log to /tmp instead of /concourse-work-dir
If the worker runs out of disk space, logs will stop tee-ing to
/concourse-work-dir/.liveness_probe because it's on the same persistent
volume, and the condition won't be detected.
* Add OWNERS file
Contributors, reviewers and approvers of PRs to this chart who are also
kubernetes github org members.
* fix typos
* see if btrfs works in e2e test
* Stick with naive FS driver since its the most compatible
* Fix#3407, make docs reflect actual value
* Remove core maintainers from local OWNERS
The maintainers have OWNER on the top level of the repo already
* [stable/concourse] Tune fatal errors.
Remove guardian.api.garden-server.run.failed from the list of fatal errors,
this can be triggered merely from configuring a task to run a command that
doesn't exist.
* Update Chart.yaml
Make necessary improvements to Concourse worker lifecycle management.
Add additional fatal errors emitted as of Concourse 3.8.0 that should
trigger a restart, and remove "unkown volume" as one such error as this
will happen normally when running multiple concourse-web pods.
Try to start workers with a clean slate by cleaning up previous
incarnations of a worker. Call retire-worker before starting. Also
clear the concourse-work-dir before starting.
Call retire-worker in a loop and don't exit that loop until the old
worker is gone. This allows us to remove the fixed worker.postStopDelaySeconds
duration.
Add a note about persistent volumes being necessary.
Add containerPlacementStrategy, default random to better spread load
across workers.
In concourse, to expose prometheus metrics, need to set the env vars:
CONCOURSE_PROMETHEUS_BIND_IP = 0.0.0.0
CONCOURSE_PROMETHEUS_BIND_PORT = portnumber
This provides an HTTP endpoint to scrape the metrics.
Enabling metrics via the values.yaml is half the battle, still need to configure prometheus (potentially via service discovery and annotations) to scrape the data.
Rui Yang
Norbert Buchmüller
Babatunde Olu-Isa
Divya Dadlani
Daniel Bartlett
Yann David
Richard Alberto
Kamil Szczygieł
Chris Hiestand
Kurt McAlpine
AdamDang
Naoki Ainoya
Dan Wendorf
Daniel Whatmuff
Brian Martin
Will Tran
josecv
Chengfei Shang
Bruno
Alexander M
Ma Huihuang
Dylan Arbour
Matt Farina
Kai Chen
James Gregory
Alex M
pontusarfwedson
Adarsh J
Michael Bannister
Lingming Xia
John Wells
matthope