Update dashboard VictoriaMetrics - vmagent ( 20 → 21 )

.taskfiles/volsync/Taskfile.yaml

@@ -2,11 +2,10 @@
 # yaml-language-server: $schema=https://taskfile.dev/schema.json
 version: '3'
 
-# This taskfile is used to manage certain VolSync tasks for a given application, limitations are described below.
+# Taskfile used to manage certain VolSync tasks for a given application, limitations are as followed.
 #   1. Fluxtomization, HelmRelease, PVC, ReplicationSource all have the same name (e.g. plex)
 #   2. ReplicationSource and ReplicationDestination are a Restic repository
-#   3. Applications are deployed as either a Kubernetes Deployment or StatefulSet
+#   3. Each application only has one PVC that is being replicated
-#   4. Each application only has one PVC that is being replicated
 
 vars:
   VOLSYNC_RESOURCES_DIR: '{{.ROOT_DIR}}/.taskfiles/volsync/resources'
@@ -14,39 +13,34 @@ vars:
 tasks:
 
   state-*:
-    desc: Suspend or Resume Volsync
+    desc: Suspend or resume Volsync [CLUSTER=main]
-    summary: |-
-      CLUSTER: Cluster to run command against (default: main)
-      STATE: resume or suspend (required)
     cmds:
-      # - until kubectl wait jobs --all --all-namespaces --for=condition=complete --timeout=5m &>/dev/null; do sleep 5; done
+      - flux --namespace flux-system {{.STATE}} kustomization volsync
-      - flux {{.STATE}} kustomization volsync
+      - flux --namespace volsync-system {{.STATE}} helmrelease volsync
-      - flux --namespace {{.NS}} {{.STATE}} helmrelease volsync
+      - kubectl --namespace volsync-system scale deployment volsync --replicas {{if eq .STATE "suspend"}}0{{else}}1{{end}}
-      - kubectl --namespace {{.NS}} scale deployment --all --replicas {{if eq .STATE "suspend"}}0{{else}}1{{end}}
     vars:
-      NS: '{{.NS | default "volsync-system"}}'
       STATE: '{{index .MATCH 0}}'
     requires:
       vars: [CLUSTER]
+    preconditions:
+      - '[[ "{{.STATE}}" == "suspend" || "{{.STATE}}" == "resume" ]]'
+      - which flux kubectl
 
   unlock:
-    desc: Unlock all Restic repositories
+    desc: Unlock all restic source repos [CLUSTER=main]
-    summary: |-
+    cmds:
-      CLUSTER: Cluster to run command against (default: main)
+      - for: { var: SOURCES, split: "\n" }
-    cmd: >
+        cmd: kubectl --namespace {{splitList "," .ITEM | first}} patch --field-manager=flux-client-side-apply replicationsources {{splitList "," .ITEM | last}} --type merge --patch "{\"spec\":{\"restic\":{\"unlock\":\"{{now | unixEpoch}}\"}}}"
-      kubectl get replicationsources --all-namespaces --no-headers -A | awk '{print $1, $2}'
+    vars:
-      | xargs --max-procs=2 -l bash -c 'kubectl --namespace "$0" patch --field-manager=flux-client-side-apply replicationsources "$1" --type merge --patch "{\"spec\":{\"restic\":{\"unlock\":\"{{now | unixEpoch}}\"}}}"'
+      SOURCES:
+        sh: kubectl get replicationsources --all-namespaces --no-headers --output=jsonpath='{range .items[*]}{.metadata.namespace},{.metadata.name}{"\n"}{end}'
     requires:
       vars: [CLUSTER]
+    preconditions:
+      - which kubectl
 
-  # To run backup jobs in parallel for all replicationsources:
-  #   - kubectl get replicationsources --all-namespaces --no-headers | awk '{print $2, $1}' | xargs --max-procs=4 -l bash -c 'task volsync:snapshot APP=$0 NS=$1'
   snapshot:
-    desc: Snapshot an application
+    desc: Snapshot an app [CLUSTER=main] [NS=default] [APP=required]
-    summary: |-
-      CLUSTER: Cluster to run command against (default: main)
-      NS: Namespace the application is in (default: default)
-      APP: Application to snapshot (required)
     cmds:
       - kubectl --namespace {{.NS}} patch replicationsources {{.APP}} --type merge -p '{"spec":{"trigger":{"manual":"{{now | unixEpoch}}"}}}'
       - until kubectl --namespace {{.NS}} get job/{{.JOB}} &>/dev/null; do sleep 5; done
@@ -58,47 +52,34 @@ tasks:
       vars: [CLUSTER, APP]
     preconditions:
       - kubectl --namespace {{.NS}} get replicationsources {{.APP}}
+      - which kubectl
 
-  # To run restore jobs in parallel for all replicationdestinations:
-  #    - kubectl get replicationsources --all-namespaces --no-headers | awk '{print $2, $1}' | xargs --max-procs=4 -l bash -c 'task volsync:restore APP=$0 NS=$1'
   restore:
-    desc: Restore an application
+    desc: Restore an app [CLUSTER=main] [NS=default] [APP=required] [PREVIOUS=required]
-    summary: |-
-      CLUSTER: Cluster to run command against (default: main)
-      NS: Namespace the application is in (default: default)
-      APP: Application to restore (required)
-      PREVIOUS: Previous number of snapshots to restore (default: 2)
-    cmds:
-      - task: .suspend
-      - task: .restore
-      - task: .resume
-    requires:
-      vars: [CLUSTER, APP]
-
-  .suspend:
-    internal: true
     cmds:
+      # Suspend
       - flux --namespace flux-system suspend kustomization {{.APP}}
       - flux --namespace {{.NS}} suspend helmrelease {{.APP}}
       - kubectl --namespace {{.NS}} scale {{.CONTROLLER}}/{{.APP}} --replicas 0
       - kubectl --namespace {{.NS}} wait pod --for=delete --selector="app.kubernetes.io/name={{.APP}}" --timeout=5m
+      # Restore
+      - minijinja-cli {{.VOLSYNC_RESOURCES_DIR}}/replicationdestination.yaml.j2 | kubectl apply --server-side --filename -
+      - until kubectl --namespace {{.NS}} get job/volsync-dst-{{.APP}}-manual &>/dev/null; do sleep 5; done
+      - kubectl --namespace {{.NS}} wait job/volsync-dst-{{.APP}}-manual --for=condition=complete --timeout=120m
+      - kubectl --namespace {{.NS}} delete replicationdestination {{.APP}}-manual
+      # Resume
+      - flux --namespace flux-system resume kustomization {{.APP}}
+      - flux --namespace {{.NS}} resume helmrelease {{.APP}}
+      - flux --namespace {{.NS}} reconcile helmrelease {{.APP}} --force
+      - kubectl --namespace {{.NS}} wait pod --for=condition=ready --selector="app.kubernetes.io/name={{.APP}}" --timeout=5m
     vars:
       NS: '{{.NS | default "default"}}'
-      APP: '{{.APP}}'
       CONTROLLER:
         sh: kubectl --namespace {{.NS}} get deployment {{.APP}} &>/dev/null && echo deployment || echo statefulset
-
+    env:
-  .restore:
+      NS: '{{.NS}}'
-    internal: true
+      APP: '{{.APP}}'
-    cmds:
+      PREVIOUS: '{{.PREVIOUS}}'
-      - minijinja-cli --env --trim-blocks --lstrip-blocks --autoescape=none {{.VOLSYNC_RESOURCES_DIR}}/replicationdestination.yaml.j2 | kubectl apply --server-side --filename -
-      - until kubectl --namespace {{.NS}} get job/{{.JOB}} &>/dev/null; do sleep 5; done
-      - kubectl --namespace {{.NS}} wait job/{{.JOB}} --for=condition=complete --timeout=120m
-      - kubectl --namespace {{.NS}} delete replicationdestination {{.JOB}}
-    vars:
-      NS: '{{.NS | default "default"}}'
-      JOB: volsync-dst-{{.APP}}
-      PREVIOUS: '{{.PREVIOUS | default 2}}'
       CLAIM:
         sh: kubectl --namespace {{.NS}} get replicationsources/{{.APP}} --output=jsonpath="{.spec.sourcePVC}"
       ACCESS_MODES:
@@ -109,28 +90,8 @@ tasks:
         sh: kubectl --namespace {{.NS}} get replicationsources/{{.APP}} --output=jsonpath="{.spec.restic.moverSecurityContext.runAsUser}"
       PGID:
         sh: kubectl --namespace {{.NS}} get replicationsources/{{.APP}} --output=jsonpath="{.spec.restic.moverSecurityContext.runAsGroup}"
-    env:
+    requires:
-      NS: '{{.NS}}'
+      vars: [CLUSTER, APP, PREVIOUS]
-      JOB: '{{.JOB}}'
-      APP: '{{.APP}}'
-      PREVIOUS: '{{.PREVIOUS}}'
-      CLAIM: '{{.CLAIM}}'
-      ACCESS_MODES: '{{.ACCESS_MODES}}'
-      STORAGE_CLASS_NAME: '{{.STORAGE_CLASS_NAME}}'
-      PUID: '{{.PUID}}'
-      PGID: '{{.PGID}}'
     preconditions:
       - test -f {{.VOLSYNC_RESOURCES_DIR}}/replicationdestination.yaml.j2
-
+      - which flux kubectl minijinja-cli
-  .resume:
-    internal: true
-    cmds:
-      - flux --namespace {{.NS}} resume helmrelease {{.APP}}
-      - flux --namespace flux-system resume kustomization {{.APP}}
-      - kubectl --namespace {{.NS}} scale {{.CONTROLLER}}/{{.APP}} --replicas 1
-      - kubectl --namespace {{.NS}} wait pod --for=condition=ready --selector="app.kubernetes.io/name={{.APP}}" --timeout=5m
-    vars:
-      NS: '{{.NS | default "default"}}'
-      APP: '{{.APP}}'
-      CONTROLLER:
-        sh: kubectl --namespace {{.NS}} get deployment {{.APP}} &>/dev/null && echo deployment || echo statefulset

.taskfiles/volsync/resources/replicationdestination.yaml.j2

@@ -2,7 +2,7 @@
 apiVersion: volsync.backube/v1alpha1
 kind: ReplicationDestination
 metadata:
-  name: {{ ENV.JOB }}
+  name: {{ ENV.APP }}-manual
   namespace: {{ ENV.NS }}
 spec:
   trigger:

Taskfile.yaml

@@ -7,6 +7,8 @@ vars:
   CLUSTER_SETTINGS_FILE: "{{.CLUSTER_DIR}}/flux/vars/cluster-settings.env"
   K8S_CLUSTER: '{{.K8S_CLUSTER | default "theshire"}}'
   K8S_CLUSTER_DIR: '{{.KUBERNETES_DIR}}'
+  CLUSTER: '{{.CLUSTER | default "theshire"}}'
+  CLUSTER_DIR: '{{.KUBERNETES_DIR}}'
 
 env:
   KUBECONFIG: "{{.ROOT_DIR}}/kubeconfig"

kubernetes/apps/observability/grafana/app/helmrelease.yaml

@@ -247,7 +247,7 @@ spec:
         victoria-agent:
           # renovate: depName="VictoriaMetrics - vmagent"
           gnetId: 12683
-          revision: 20
+          revision: 21
           datasource: Prometheus
         victoria-alert:
           # renovate: depName="VictoriaMetrics - vmalert"

kubernetes/bootstrap/talos/talconfig.yaml

@@ -2,7 +2,7 @@
 # yaml-language-server: $schema=https://ks.hsn.dev/talconfig.json
 clusterName: theshire
 
-talosVersion: v1.8.1
+talosVersion: v1.8.2
 kubernetesVersion: 1.31.1
 endpoint: "https://10.1.1.57:6444"