You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Make job, initContainer, metrics resources configurable (#71)
It is a good practice to configure the ephemeral-storage limits and
requests for containers in addition to memory and CPU.
Also some containers had hard coded values for the resource limits and
requests.
Metrics container was not possible to configure separately from the main
container.
I also need to set the annotations for the pod created by the Job.
I hope i set some sane default to the requests, please adjust as You see
fit.
A Helm chart for managing a wireguard vpn in kubernetes
6
6
@@ -38,7 +38,7 @@ A Helm chart for managing a wireguard vpn in kubernetes
38
38
| healthSideCar.image.pullPolicy | string |`"Always"`| Pull Policy always to avoid cached rolling tags, if you change this you should use a non rolling tag |
39
39
| healthSideCar.image.repository | string |`"ghcr.io/bryopsida/http-healthcheck-sidecar"`| Override repo if you prefer to use your own image |
40
40
| healthSideCar.image.tag | string |`"main"`| Rolling tag used by default to take patches automatically |
41
-
| healthSideCar.resources | object |`{"limits":{"cpu":"100m","memory":"256Mi"},"requests":{"cpu":"100m","memory":"256Mi"}}`| set resource constraints, set to nil to remove |
41
+
| healthSideCar.resources | object |`{"limits":{"cpu":"100m","ephemeral-storage":"256Mi","memory":"256Mi"},"requests":{"cpu":"100m","ephemeral-storage":"8Mi","memory":"256Mi"}}`| set resource constraints, set to nil to remove |
42
42
| healthSideCar.securityContext | object |`{"allowPrivilegeEscalation":false,"capabilities":{"drop":["ALL"]},"readOnlyRootFilesystem":true,"runAsGroup":10001,"runAsNonRoot":true,"runAsUser":10001,"seccompProfile":{"type":"RuntimeDefault"}}`| Secure settings by default, can be overriden to reduce security posture if needed |
43
43
| healthSideCar.service.enabled | bool |`true`| Toggle to enable the service, if the pod is a daemonset healthSideCar.useHostPort can be used instead |
44
44
| healthSideCar.service.nodePort | int |`31313`| The port for the service exposed on each node |
@@ -51,6 +51,12 @@ A Helm chart for managing a wireguard vpn in kubernetes
| keygenJob.useWireguardManager | bool |`false`| when enabled, uses a image with go bindings for k8s and wg to create the secret if it does not exist, on re-runs it it leaves the existing secret in place and exits succesfully |
70
83
| keygenJob.wireguardMgrImage | object |`{"pullPolicy":"Always","repository":"ghcr.io/bryopsida/k8s-wireguard-mgr","tag":"main"}`| When useWireguardManager is enabled this image is used instead of the kubectl image |
71
84
| labels | object |`{}`||
@@ -87,6 +100,12 @@ A Helm chart for managing a wireguard vpn in kubernetes
87
100
| metrics.prometheusRule.groups | list |`[]`| Groups, containing the alert rules. Example: groups: - name: Wireguard rules: - alert: WireguardInstanceNotAvailable annotations: message: "Wireguard instance in namespace {{ `{{` }} $labels.namespace {{ `}}` }} has not been available for the last 5 minutes." expr: | absent(kube_pod_status_ready{namespace="{{ include "common.names.namespace" . }}", condition="true"} * on (pod) kube_pod_labels{pod=~"{{ include "common.names.fullname" . }}-\\d+", namespace="{{ include "common.names.namespace" . }}"}) != 0 for: 5m labels: severity: critical |
88
101
| metrics.prometheusRule.labels | object |`{}`| Additional labels that can be used so PrometheusRule will be discovered by Prometheus |
89
102
| metrics.prometheusRule.namespace | string |`""`| Namespace of the ServiceMonitor. If empty, current namespace is used |
0 commit comments