[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-google-cloud-gke-workload-scaling":3,"mdc--8oqchg-key":34,"related-org-google-cloud-gke-workload-scaling":862,"related-repo-google-cloud-gke-workload-scaling":1051},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":12,"stars":24,"repoUrl":25,"updatedAt":26,"license":27,"forks":28,"topics":29,"repo":30,"sourceUrl":32,"mdContent":33},"gke-workload-scaling","scale GKE workloads with HPA and VPA","Specific workflows for scaling GKE workloads using HPA and VPA, as well as best practices for autoscaling configuration.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},"google-cloud","Google Cloud","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fgoogle-cloud.png","GoogleCloudPlatform",[13,17,20,23],{"name":14,"slug":15,"type":16},"Performance","performance","tag",{"name":18,"slug":19,"type":16},"Deployment","deployment",{"name":21,"slug":22,"type":16},"Kubernetes","kubernetes",{"name":9,"slug":8,"type":16},161,"https:\u002F\u002Fgithub.com\u002FGoogleCloudPlatform\u002Fgke-mcp","2026-07-12T07:39:44.456612",null,78,[],{"repoUrl":25,"stars":24,"forks":28,"topics":31,"description":27},[],"https:\u002F\u002Fgithub.com\u002FGoogleCloudPlatform\u002Fgke-mcp\u002Ftree\u002FHEAD\u002Fskills\u002Fgke-workload-scaling","---\nname: gke-workload-scaling\ndescription: Specific workflows for scaling GKE workloads using HPA and VPA, as well as best practices for autoscaling configuration.\n---\n\n# GKE Workload Scaling\n\nThis skill provides workflows and best practices for scaling applications on Google Kubernetes Engine (GKE). It covers manual scaling, Horizontal Pod Autoscaling (HPA), and Vertical Pod Autoscaling (VPA).\n\n## Workflows\n\n### 1. Manual Scaling\n\nQuickly scale a deployment to a fixed number of replicas. Useful for immediate manual intervention or testing.\n\n**Command:**\n\n```bash\nkubectl scale deployment \u003Cdeployment-name> --replicas=\u003Cnumber> -n \u003Cnamespace>\n```\n\n### 2. Horizontal Pod Autoscaling (HPA)\n\nAutomatically scale the number of pods based on observed CPU utilization, memory utilization, or custom metrics.\n\n**Prerequisites:**\n\n- Metrics Server must be running (enabled by default on GKE).\n- Containers clearly define resource requests\u002Flimits.\n\n**Quick Command:**\n\n```bash\nkubectl autoscale deployment \u003Cdeployment-name> --cpu-percent=50 --min=1 --max=10\n```\n\n**Manifest Approach (Recommended):**\nUse a YAML manifest for version-controlled configuration.\nSee [assets\u002Fhpa-example.yaml](assets\u002Fhpa-example.yaml) for a template.\n\n```bash\nkubectl apply -f assets\u002Fhpa-example.yaml\n```\n\n**Custom Metrics & External Metrics:**\nFor GKE, the modern and recommended approach for scaling based on Cloud Monitoring metrics (e.g., Pub\u002FSub queue length) is to use the **External** metric type, which is natively supported by the GKE control plane without requiring the Custom Metrics Adapter. For application-specific metrics exposed via Prometheus, you can use **Google Cloud Managed Service for Prometheus** or the Prometheus Adapter.\n\n### 3. Vertical Pod Autoscaling (VPA)\n\nAutomatically adjust the CPU and memory reservations for your pods to match actual usage. This is critical for right-sizing workloads.\n\n**Prerequisites:**\n\n- VPA must be enabled on the cluster.\n  - **Autopilot:** Enabled by default.\n  - **Standard:** Must be enabled manually.\n\n**Enable VPA on Standard Cluster:**\n\n```bash\ngcloud container clusters update \u003Ccluster-name> --enable-vertical-pod-autoscaling --zone \u003Czone>\n```\n\n**Update Modes:**\n\n- `Off`: Calculates recommendations but does not apply them. Good for \"dry run\" analysis.\n- `Initial`: Assigns resources only at pod creation time.\n- `Auto`: Updates running pods by restarting them if recommendations differ significantly from requests.\n- `InPlaceOrRecreate`: Attempts to update Pod resources without recreating the Pod. If in-place update is not possible, it reverts to `Auto` mode (requires GKE 1.34+).\n\n**Example:**\nSee [assets\u002Fvpa-example.yaml](assets\u002Fvpa-example.yaml) for a configuration template.\n\n### 4. Cluster Autoscaler\n\nWhile not a workload-level scaler, the Cluster Autoscaler is essential for ensuring your cluster has enough nodes to run the scaled pods.\n\n**Enable on a Node Pool:**\n\n```bash\ngcloud container clusters update \u003Ccluster-name> \\\n    --enable-autoscaling \\\n    --node-pool \u003Cnode-pool-name> \\\n    --min-nodes \u003Cmin> \\\n    --max-nodes \u003Cmax> \\\n    --zone \u003Czone>\n```\n\n## Best Practices\n\n1. **Define Resource Requests:** HPA and VPA rely on accurate resource requests. Always define them in your container specs.\n2. **Avoid Metric Conflicts:** Do not configure HPA and VPA to use the same metric (e.g., both CPU). This causes thrashing.\n   - _Typical Pattern:_ HPA on CPU, VPA on Memory.\n3. **Pod Disruption Budgets (PDBs):** Define PDBs to ensure application availability during scaling events or node upgrades.\n4. **HPA Lag:** HPA has a stabilization window (default 5 mins) to prevent rapid fluctuation.\n5. **VPA \"Auto\" Mode Risks:** In \"Auto\" mode, VPA restarts pods to change resources. Ensure your application handles restarts gracefully (e.g., handles SIGTERM).\n   - _Note:_ By default, VPA requires at least 2 replicas to perform evictions. In GKE 1.22+, you can override this by setting `minReplicas` in `PodUpdatePolicy`.\n",{"data":35,"body":36},{"name":4,"description":6},{"type":37,"children":38},"root",[39,47,53,60,67,72,81,180,186,191,199,214,222,276,294,323,347,353,358,365,396,404,478,486,540,557,563,568,576,753,759,856],{"type":40,"tag":41,"props":42,"children":43},"element","h1",{"id":4},[44],{"type":45,"value":46},"text","GKE Workload Scaling",{"type":40,"tag":48,"props":49,"children":50},"p",{},[51],{"type":45,"value":52},"This skill provides workflows and best practices for scaling applications on Google Kubernetes Engine (GKE). It covers manual scaling, Horizontal Pod Autoscaling (HPA), and Vertical Pod Autoscaling (VPA).",{"type":40,"tag":54,"props":55,"children":57},"h2",{"id":56},"workflows",[58],{"type":45,"value":59},"Workflows",{"type":40,"tag":61,"props":62,"children":64},"h3",{"id":63},"_1-manual-scaling",[65],{"type":45,"value":66},"1. Manual Scaling",{"type":40,"tag":48,"props":68,"children":69},{},[70],{"type":45,"value":71},"Quickly scale a deployment to a fixed number of replicas. Useful for immediate manual intervention or testing.",{"type":40,"tag":48,"props":73,"children":74},{},[75],{"type":40,"tag":76,"props":77,"children":78},"strong",{},[79],{"type":45,"value":80},"Command:",{"type":40,"tag":82,"props":83,"children":88},"pre",{"className":84,"code":85,"language":86,"meta":87,"style":87},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","kubectl scale deployment \u003Cdeployment-name> --replicas=\u003Cnumber> -n \u003Cnamespace>\n","bash","",[89],{"type":40,"tag":90,"props":91,"children":92},"code",{"__ignoreMap":87},[93],{"type":40,"tag":94,"props":95,"children":98},"span",{"class":96,"line":97},"line",1,[99,105,111,116,122,127,133,138,143,148,153,157,162,166,171,175],{"type":40,"tag":94,"props":100,"children":102},{"style":101},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[103],{"type":45,"value":104},"kubectl",{"type":40,"tag":94,"props":106,"children":108},{"style":107},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[109],{"type":45,"value":110}," scale",{"type":40,"tag":94,"props":112,"children":113},{"style":107},[114],{"type":45,"value":115}," deployment",{"type":40,"tag":94,"props":117,"children":119},{"style":118},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[120],{"type":45,"value":121}," \u003C",{"type":40,"tag":94,"props":123,"children":124},{"style":107},[125],{"type":45,"value":126},"deployment-nam",{"type":40,"tag":94,"props":128,"children":130},{"style":129},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[131],{"type":45,"value":132},"e",{"type":40,"tag":94,"props":134,"children":135},{"style":118},[136],{"type":45,"value":137},">",{"type":40,"tag":94,"props":139,"children":140},{"style":107},[141],{"type":45,"value":142}," --replicas=",{"type":40,"tag":94,"props":144,"children":145},{"style":118},[146],{"type":45,"value":147},"\u003C",{"type":40,"tag":94,"props":149,"children":150},{"style":107},[151],{"type":45,"value":152},"number",{"type":40,"tag":94,"props":154,"children":155},{"style":118},[156],{"type":45,"value":137},{"type":40,"tag":94,"props":158,"children":159},{"style":107},[160],{"type":45,"value":161}," -n",{"type":40,"tag":94,"props":163,"children":164},{"style":118},[165],{"type":45,"value":121},{"type":40,"tag":94,"props":167,"children":168},{"style":107},[169],{"type":45,"value":170},"namespac",{"type":40,"tag":94,"props":172,"children":173},{"style":129},[174],{"type":45,"value":132},{"type":40,"tag":94,"props":176,"children":177},{"style":118},[178],{"type":45,"value":179},">\n",{"type":40,"tag":61,"props":181,"children":183},{"id":182},"_2-horizontal-pod-autoscaling-hpa",[184],{"type":45,"value":185},"2. Horizontal Pod Autoscaling (HPA)",{"type":40,"tag":48,"props":187,"children":188},{},[189],{"type":45,"value":190},"Automatically scale the number of pods based on observed CPU utilization, memory utilization, or custom metrics.",{"type":40,"tag":48,"props":192,"children":193},{},[194],{"type":40,"tag":76,"props":195,"children":196},{},[197],{"type":45,"value":198},"Prerequisites:",{"type":40,"tag":200,"props":201,"children":202},"ul",{},[203,209],{"type":40,"tag":204,"props":205,"children":206},"li",{},[207],{"type":45,"value":208},"Metrics Server must be running (enabled by default on GKE).",{"type":40,"tag":204,"props":210,"children":211},{},[212],{"type":45,"value":213},"Containers clearly define resource requests\u002Flimits.",{"type":40,"tag":48,"props":215,"children":216},{},[217],{"type":40,"tag":76,"props":218,"children":219},{},[220],{"type":45,"value":221},"Quick Command:",{"type":40,"tag":82,"props":223,"children":225},{"className":84,"code":224,"language":86,"meta":87,"style":87},"kubectl autoscale deployment \u003Cdeployment-name> --cpu-percent=50 --min=1 --max=10\n",[226],{"type":40,"tag":90,"props":227,"children":228},{"__ignoreMap":87},[229],{"type":40,"tag":94,"props":230,"children":231},{"class":96,"line":97},[232,236,241,245,249,253,257,261,266,271],{"type":40,"tag":94,"props":233,"children":234},{"style":101},[235],{"type":45,"value":104},{"type":40,"tag":94,"props":237,"children":238},{"style":107},[239],{"type":45,"value":240}," autoscale",{"type":40,"tag":94,"props":242,"children":243},{"style":107},[244],{"type":45,"value":115},{"type":40,"tag":94,"props":246,"children":247},{"style":118},[248],{"type":45,"value":121},{"type":40,"tag":94,"props":250,"children":251},{"style":107},[252],{"type":45,"value":126},{"type":40,"tag":94,"props":254,"children":255},{"style":129},[256],{"type":45,"value":132},{"type":40,"tag":94,"props":258,"children":259},{"style":118},[260],{"type":45,"value":137},{"type":40,"tag":94,"props":262,"children":263},{"style":107},[264],{"type":45,"value":265}," --cpu-percent=50",{"type":40,"tag":94,"props":267,"children":268},{"style":107},[269],{"type":45,"value":270}," --min=1",{"type":40,"tag":94,"props":272,"children":273},{"style":107},[274],{"type":45,"value":275}," --max=10\n",{"type":40,"tag":48,"props":277,"children":278},{},[279,284,286,292],{"type":40,"tag":76,"props":280,"children":281},{},[282],{"type":45,"value":283},"Manifest Approach (Recommended):",{"type":45,"value":285},"\nUse a YAML manifest for version-controlled configuration.\nSee ",{"type":40,"tag":287,"props":288,"children":290},"a",{"href":289},"assets\u002Fhpa-example.yaml",[291],{"type":45,"value":289},{"type":45,"value":293}," for a template.",{"type":40,"tag":82,"props":295,"children":297},{"className":84,"code":296,"language":86,"meta":87,"style":87},"kubectl apply -f assets\u002Fhpa-example.yaml\n",[298],{"type":40,"tag":90,"props":299,"children":300},{"__ignoreMap":87},[301],{"type":40,"tag":94,"props":302,"children":303},{"class":96,"line":97},[304,308,313,318],{"type":40,"tag":94,"props":305,"children":306},{"style":101},[307],{"type":45,"value":104},{"type":40,"tag":94,"props":309,"children":310},{"style":107},[311],{"type":45,"value":312}," apply",{"type":40,"tag":94,"props":314,"children":315},{"style":107},[316],{"type":45,"value":317}," -f",{"type":40,"tag":94,"props":319,"children":320},{"style":107},[321],{"type":45,"value":322}," assets\u002Fhpa-example.yaml\n",{"type":40,"tag":48,"props":324,"children":325},{},[326,331,333,338,340,345],{"type":40,"tag":76,"props":327,"children":328},{},[329],{"type":45,"value":330},"Custom Metrics & External Metrics:",{"type":45,"value":332},"\nFor GKE, the modern and recommended approach for scaling based on Cloud Monitoring metrics (e.g., Pub\u002FSub queue length) is to use the ",{"type":40,"tag":76,"props":334,"children":335},{},[336],{"type":45,"value":337},"External",{"type":45,"value":339}," metric type, which is natively supported by the GKE control plane without requiring the Custom Metrics Adapter. For application-specific metrics exposed via Prometheus, you can use ",{"type":40,"tag":76,"props":341,"children":342},{},[343],{"type":45,"value":344},"Google Cloud Managed Service for Prometheus",{"type":45,"value":346}," or the Prometheus Adapter.",{"type":40,"tag":61,"props":348,"children":350},{"id":349},"_3-vertical-pod-autoscaling-vpa",[351],{"type":45,"value":352},"3. Vertical Pod Autoscaling (VPA)",{"type":40,"tag":48,"props":354,"children":355},{},[356],{"type":45,"value":357},"Automatically adjust the CPU and memory reservations for your pods to match actual usage. This is critical for right-sizing workloads.",{"type":40,"tag":48,"props":359,"children":360},{},[361],{"type":40,"tag":76,"props":362,"children":363},{},[364],{"type":45,"value":198},{"type":40,"tag":200,"props":366,"children":367},{},[368],{"type":40,"tag":204,"props":369,"children":370},{},[371,373],{"type":45,"value":372},"VPA must be enabled on the cluster.\n",{"type":40,"tag":200,"props":374,"children":375},{},[376,386],{"type":40,"tag":204,"props":377,"children":378},{},[379,384],{"type":40,"tag":76,"props":380,"children":381},{},[382],{"type":45,"value":383},"Autopilot:",{"type":45,"value":385}," Enabled by default.",{"type":40,"tag":204,"props":387,"children":388},{},[389,394],{"type":40,"tag":76,"props":390,"children":391},{},[392],{"type":45,"value":393},"Standard:",{"type":45,"value":395}," Must be enabled manually.",{"type":40,"tag":48,"props":397,"children":398},{},[399],{"type":40,"tag":76,"props":400,"children":401},{},[402],{"type":45,"value":403},"Enable VPA on Standard Cluster:",{"type":40,"tag":82,"props":405,"children":407},{"className":84,"code":406,"language":86,"meta":87,"style":87},"gcloud container clusters update \u003Ccluster-name> --enable-vertical-pod-autoscaling --zone \u003Czone>\n",[408],{"type":40,"tag":90,"props":409,"children":410},{"__ignoreMap":87},[411],{"type":40,"tag":94,"props":412,"children":413},{"class":96,"line":97},[414,419,424,429,434,438,443,447,451,456,461,465,470,474],{"type":40,"tag":94,"props":415,"children":416},{"style":101},[417],{"type":45,"value":418},"gcloud",{"type":40,"tag":94,"props":420,"children":421},{"style":107},[422],{"type":45,"value":423}," container",{"type":40,"tag":94,"props":425,"children":426},{"style":107},[427],{"type":45,"value":428}," clusters",{"type":40,"tag":94,"props":430,"children":431},{"style":107},[432],{"type":45,"value":433}," update",{"type":40,"tag":94,"props":435,"children":436},{"style":118},[437],{"type":45,"value":121},{"type":40,"tag":94,"props":439,"children":440},{"style":107},[441],{"type":45,"value":442},"cluster-nam",{"type":40,"tag":94,"props":444,"children":445},{"style":129},[446],{"type":45,"value":132},{"type":40,"tag":94,"props":448,"children":449},{"style":118},[450],{"type":45,"value":137},{"type":40,"tag":94,"props":452,"children":453},{"style":107},[454],{"type":45,"value":455}," --enable-vertical-pod-autoscaling",{"type":40,"tag":94,"props":457,"children":458},{"style":107},[459],{"type":45,"value":460}," --zone",{"type":40,"tag":94,"props":462,"children":463},{"style":118},[464],{"type":45,"value":121},{"type":40,"tag":94,"props":466,"children":467},{"style":107},[468],{"type":45,"value":469},"zon",{"type":40,"tag":94,"props":471,"children":472},{"style":129},[473],{"type":45,"value":132},{"type":40,"tag":94,"props":475,"children":476},{"style":118},[477],{"type":45,"value":179},{"type":40,"tag":48,"props":479,"children":480},{},[481],{"type":40,"tag":76,"props":482,"children":483},{},[484],{"type":45,"value":485},"Update Modes:",{"type":40,"tag":200,"props":487,"children":488},{},[489,500,511,522],{"type":40,"tag":204,"props":490,"children":491},{},[492,498],{"type":40,"tag":90,"props":493,"children":495},{"className":494},[],[496],{"type":45,"value":497},"Off",{"type":45,"value":499},": Calculates recommendations but does not apply them. Good for \"dry run\" analysis.",{"type":40,"tag":204,"props":501,"children":502},{},[503,509],{"type":40,"tag":90,"props":504,"children":506},{"className":505},[],[507],{"type":45,"value":508},"Initial",{"type":45,"value":510},": Assigns resources only at pod creation time.",{"type":40,"tag":204,"props":512,"children":513},{},[514,520],{"type":40,"tag":90,"props":515,"children":517},{"className":516},[],[518],{"type":45,"value":519},"Auto",{"type":45,"value":521},": Updates running pods by restarting them if recommendations differ significantly from requests.",{"type":40,"tag":204,"props":523,"children":524},{},[525,531,533,538],{"type":40,"tag":90,"props":526,"children":528},{"className":527},[],[529],{"type":45,"value":530},"InPlaceOrRecreate",{"type":45,"value":532},": Attempts to update Pod resources without recreating the Pod. If in-place update is not possible, it reverts to ",{"type":40,"tag":90,"props":534,"children":536},{"className":535},[],[537],{"type":45,"value":519},{"type":45,"value":539}," mode (requires GKE 1.34+).",{"type":40,"tag":48,"props":541,"children":542},{},[543,548,550,555],{"type":40,"tag":76,"props":544,"children":545},{},[546],{"type":45,"value":547},"Example:",{"type":45,"value":549},"\nSee ",{"type":40,"tag":287,"props":551,"children":553},{"href":552},"assets\u002Fvpa-example.yaml",[554],{"type":45,"value":552},{"type":45,"value":556}," for a configuration template.",{"type":40,"tag":61,"props":558,"children":560},{"id":559},"_4-cluster-autoscaler",[561],{"type":45,"value":562},"4. Cluster Autoscaler",{"type":40,"tag":48,"props":564,"children":565},{},[566],{"type":45,"value":567},"While not a workload-level scaler, the Cluster Autoscaler is essential for ensuring your cluster has enough nodes to run the scaled pods.",{"type":40,"tag":48,"props":569,"children":570},{},[571],{"type":40,"tag":76,"props":572,"children":573},{},[574],{"type":45,"value":575},"Enable on a Node Pool:",{"type":40,"tag":82,"props":577,"children":579},{"className":84,"code":578,"language":86,"meta":87,"style":87},"gcloud container clusters update \u003Ccluster-name> \\\n    --enable-autoscaling \\\n    --node-pool \u003Cnode-pool-name> \\\n    --min-nodes \u003Cmin> \\\n    --max-nodes \u003Cmax> \\\n    --zone \u003Czone>\n",[580],{"type":40,"tag":90,"props":581,"children":582},{"__ignoreMap":87},[583,623,636,666,697,728],{"type":40,"tag":94,"props":584,"children":585},{"class":96,"line":97},[586,590,594,598,602,606,610,614,618],{"type":40,"tag":94,"props":587,"children":588},{"style":101},[589],{"type":45,"value":418},{"type":40,"tag":94,"props":591,"children":592},{"style":107},[593],{"type":45,"value":423},{"type":40,"tag":94,"props":595,"children":596},{"style":107},[597],{"type":45,"value":428},{"type":40,"tag":94,"props":599,"children":600},{"style":107},[601],{"type":45,"value":433},{"type":40,"tag":94,"props":603,"children":604},{"style":118},[605],{"type":45,"value":121},{"type":40,"tag":94,"props":607,"children":608},{"style":107},[609],{"type":45,"value":442},{"type":40,"tag":94,"props":611,"children":612},{"style":129},[613],{"type":45,"value":132},{"type":40,"tag":94,"props":615,"children":616},{"style":118},[617],{"type":45,"value":137},{"type":40,"tag":94,"props":619,"children":620},{"style":129},[621],{"type":45,"value":622}," \\\n",{"type":40,"tag":94,"props":624,"children":626},{"class":96,"line":625},2,[627,632],{"type":40,"tag":94,"props":628,"children":629},{"style":107},[630],{"type":45,"value":631},"    --enable-autoscaling",{"type":40,"tag":94,"props":633,"children":634},{"style":129},[635],{"type":45,"value":622},{"type":40,"tag":94,"props":637,"children":639},{"class":96,"line":638},3,[640,645,649,654,658,662],{"type":40,"tag":94,"props":641,"children":642},{"style":107},[643],{"type":45,"value":644},"    --node-pool",{"type":40,"tag":94,"props":646,"children":647},{"style":118},[648],{"type":45,"value":121},{"type":40,"tag":94,"props":650,"children":651},{"style":107},[652],{"type":45,"value":653},"node-pool-nam",{"type":40,"tag":94,"props":655,"children":656},{"style":129},[657],{"type":45,"value":132},{"type":40,"tag":94,"props":659,"children":660},{"style":118},[661],{"type":45,"value":137},{"type":40,"tag":94,"props":663,"children":664},{"style":129},[665],{"type":45,"value":622},{"type":40,"tag":94,"props":667,"children":669},{"class":96,"line":668},4,[670,675,679,684,689,693],{"type":40,"tag":94,"props":671,"children":672},{"style":107},[673],{"type":45,"value":674},"    --min-nodes",{"type":40,"tag":94,"props":676,"children":677},{"style":118},[678],{"type":45,"value":121},{"type":40,"tag":94,"props":680,"children":681},{"style":107},[682],{"type":45,"value":683},"mi",{"type":40,"tag":94,"props":685,"children":686},{"style":129},[687],{"type":45,"value":688},"n",{"type":40,"tag":94,"props":690,"children":691},{"style":118},[692],{"type":45,"value":137},{"type":40,"tag":94,"props":694,"children":695},{"style":129},[696],{"type":45,"value":622},{"type":40,"tag":94,"props":698,"children":700},{"class":96,"line":699},5,[701,706,710,715,720,724],{"type":40,"tag":94,"props":702,"children":703},{"style":107},[704],{"type":45,"value":705},"    --max-nodes",{"type":40,"tag":94,"props":707,"children":708},{"style":118},[709],{"type":45,"value":121},{"type":40,"tag":94,"props":711,"children":712},{"style":107},[713],{"type":45,"value":714},"ma",{"type":40,"tag":94,"props":716,"children":717},{"style":129},[718],{"type":45,"value":719},"x",{"type":40,"tag":94,"props":721,"children":722},{"style":118},[723],{"type":45,"value":137},{"type":40,"tag":94,"props":725,"children":726},{"style":129},[727],{"type":45,"value":622},{"type":40,"tag":94,"props":729,"children":731},{"class":96,"line":730},6,[732,737,741,745,749],{"type":40,"tag":94,"props":733,"children":734},{"style":107},[735],{"type":45,"value":736},"    --zone",{"type":40,"tag":94,"props":738,"children":739},{"style":118},[740],{"type":45,"value":121},{"type":40,"tag":94,"props":742,"children":743},{"style":107},[744],{"type":45,"value":469},{"type":40,"tag":94,"props":746,"children":747},{"style":129},[748],{"type":45,"value":132},{"type":40,"tag":94,"props":750,"children":751},{"style":118},[752],{"type":45,"value":179},{"type":40,"tag":54,"props":754,"children":756},{"id":755},"best-practices",[757],{"type":45,"value":758},"Best Practices",{"type":40,"tag":760,"props":761,"children":762},"ol",{},[763,773,797,807,817],{"type":40,"tag":204,"props":764,"children":765},{},[766,771],{"type":40,"tag":76,"props":767,"children":768},{},[769],{"type":45,"value":770},"Define Resource Requests:",{"type":45,"value":772}," HPA and VPA rely on accurate resource requests. Always define them in your container specs.",{"type":40,"tag":204,"props":774,"children":775},{},[776,781,783],{"type":40,"tag":76,"props":777,"children":778},{},[779],{"type":45,"value":780},"Avoid Metric Conflicts:",{"type":45,"value":782}," Do not configure HPA and VPA to use the same metric (e.g., both CPU). This causes thrashing.\n",{"type":40,"tag":200,"props":784,"children":785},{},[786],{"type":40,"tag":204,"props":787,"children":788},{},[789,795],{"type":40,"tag":790,"props":791,"children":792},"em",{},[793],{"type":45,"value":794},"Typical Pattern:",{"type":45,"value":796}," HPA on CPU, VPA on Memory.",{"type":40,"tag":204,"props":798,"children":799},{},[800,805],{"type":40,"tag":76,"props":801,"children":802},{},[803],{"type":45,"value":804},"Pod Disruption Budgets (PDBs):",{"type":45,"value":806}," Define PDBs to ensure application availability during scaling events or node upgrades.",{"type":40,"tag":204,"props":808,"children":809},{},[810,815],{"type":40,"tag":76,"props":811,"children":812},{},[813],{"type":45,"value":814},"HPA Lag:",{"type":45,"value":816}," HPA has a stabilization window (default 5 mins) to prevent rapid fluctuation.",{"type":40,"tag":204,"props":818,"children":819},{},[820,825,827],{"type":40,"tag":76,"props":821,"children":822},{},[823],{"type":45,"value":824},"VPA \"Auto\" Mode Risks:",{"type":45,"value":826}," In \"Auto\" mode, VPA restarts pods to change resources. Ensure your application handles restarts gracefully (e.g., handles SIGTERM).\n",{"type":40,"tag":200,"props":828,"children":829},{},[830],{"type":40,"tag":204,"props":831,"children":832},{},[833,838,840,846,848,854],{"type":40,"tag":790,"props":834,"children":835},{},[836],{"type":45,"value":837},"Note:",{"type":45,"value":839}," By default, VPA requires at least 2 replicas to perform evictions. In GKE 1.22+, you can override this by setting ",{"type":40,"tag":90,"props":841,"children":843},{"className":842},[],[844],{"type":45,"value":845},"minReplicas",{"type":45,"value":847}," in ",{"type":40,"tag":90,"props":849,"children":851},{"className":850},[],[852],{"type":45,"value":853},"PodUpdatePolicy",{"type":45,"value":855},".",{"type":40,"tag":857,"props":858,"children":859},"style",{},[860],{"type":45,"value":861},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":863,"total":1050},[864,882,898,920,934,945,959,976,993,1006,1022,1032],{"slug":865,"name":865,"fn":866,"description":867,"org":868,"tags":869,"stars":879,"repoUrl":880,"updatedAt":881},"kb-search","search and extract local knowledge base documents","Allows listing, searching and extracting information from local knowledge base documents for information about tables\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[870,873,876],{"name":871,"slug":872,"type":16},"Documentation","documentation",{"name":874,"slug":875,"type":16},"Knowledge Base","knowledge-base",{"name":877,"slug":878,"type":16},"Search","search",6749,"https:\u002F\u002Fgithub.com\u002FGoogleCloudPlatform\u002Fknowledge-catalog","2026-07-12T07:38:52.157375",{"slug":883,"name":884,"fn":885,"description":886,"org":887,"tags":888,"stars":879,"repoUrl":880,"updatedAt":897},"knowledgecatalogdiscoveryagent","knowledge_catalog_discovery_agent","search and rank Knowledge Catalog data entries","Analyzes user queries, extracts relevant predicates, and utilizes Knowledge Catalog Search to find and rank the most relevant data entries. Engages with the user throughout the process.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[889,892,893,896],{"name":890,"slug":891,"type":16},"Data Analysis","data-analysis",{"name":9,"slug":8,"type":16},{"name":894,"slug":895,"type":16},"Knowledge Management","knowledge-management",{"name":877,"slug":878,"type":16},"2026-07-12T07:38:22.196851",{"slug":899,"name":899,"fn":900,"description":901,"org":902,"tags":903,"stars":917,"repoUrl":918,"updatedAt":919},"contributing","contribute to Cloud Foundation Fabric","End-to-end workflow for contributing to Cloud Foundation Fabric: triaging GitHub issues, proactive feature development, validating with tests and Policy Troubleshooter, and submitting sanitized Pull Requests. Use when addressing a Fabric GitHub issue, developing a module or FAST stage change, or preparing a branch for a pull request.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[904,907,910,913,914],{"name":905,"slug":906,"type":16},"Automation","automation",{"name":908,"slug":909,"type":16},"Engineering","engineering",{"name":911,"slug":912,"type":16},"GitHub","github",{"name":9,"slug":8,"type":16},{"name":915,"slug":916,"type":16},"Pull Requests","pull-requests",2062,"https:\u002F\u002Fgithub.com\u002FGoogleCloudPlatform\u002Fcloud-foundation-fabric","2026-07-31T06:23:36.935005",{"slug":921,"name":921,"fn":922,"description":923,"org":924,"tags":925,"stars":917,"repoUrl":918,"updatedAt":933},"fabric-builder","generate Terraform code for Google Cloud","Generates idiomatic Cloud Foundation Fabric (CFF) Terraform code using CFF modules. Use when users ask to create GCP resources, use Fabric modules, or generate Terraform code for Google Cloud.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[926,927,930],{"name":9,"slug":8,"type":16},{"name":928,"slug":929,"type":16},"Infrastructure as Code","infrastructure-as-code",{"name":931,"slug":932,"type":16},"Terraform","terraform","2026-07-12T07:38:23.514555",{"slug":935,"name":935,"fn":936,"description":937,"org":938,"tags":939,"stars":917,"repoUrl":918,"updatedAt":944},"fast-0-org-setup-prereqs","prepare prerequisites for FAST 0-org-setup","Guides the user step-by-step through the prerequisites for the FAST 0-org-setup stage, supporting both Standard GCP and Google Cloud Dedicated (GCD) environments. Use when a user asks to prepare or run prerequisites for 0-org-setup or bootstrap the FAST landing zone.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[940,941],{"name":9,"slug":8,"type":16},{"name":942,"slug":943,"type":16},"Operations","operations","2026-07-12T07:38:28.127148",{"slug":946,"name":946,"fn":947,"description":948,"org":949,"tags":950,"stars":956,"repoUrl":957,"updatedAt":958},"agent-aware-cli","design agent-aware command-line interfaces","Guide for designing and implementing command-line interfaces (CLIs) that are equally usable by human developers and automated coding agents. Use when the user wants to build a CLI, apply CLI best practices, or use Go with Cobra and Viper.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[951,954,955],{"name":952,"slug":953,"type":16},"CLI","cli",{"name":908,"slug":909,"type":16},{"name":9,"slug":8,"type":16},1150,"https:\u002F\u002Fgithub.com\u002FGoogleCloudPlatform\u002Fvertex-ai-creative-studio","2026-07-12T07:39:08.41406",{"slug":960,"name":960,"fn":961,"description":962,"org":963,"tags":964,"stars":956,"repoUrl":957,"updatedAt":975},"build-mcp-genmedia","build and configure GenAI MCP servers","Builds the mcp-genmedia Go MCP servers (nanobanana, veo, lyria, gemini-multimodal, chirp3-hd, avtool) from source and wires them into settings.json. Use this skill whenever the MCP tools are missing or broken — typically at the start of a new session, after a container restart, or when \u002Ftmp has been wiped. The prebuilt binaries in \u002Fworkspace\u002F.local\u002Fbin\u002F have no exec bit and live on a noexec mount; this skill compiles fresh executables into \u002Ftmp\u002Fbin\u002F where execution is allowed.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[965,968,969,972],{"name":966,"slug":967,"type":16},"API Development","api-development",{"name":9,"slug":8,"type":16},{"name":970,"slug":971,"type":16},"LLM","llm",{"name":973,"slug":974,"type":16},"MCP","mcp","2026-07-12T07:39:10.911302",{"slug":977,"name":977,"fn":978,"description":979,"org":980,"tags":981,"stars":956,"repoUrl":957,"updatedAt":992},"genmedia-audio-engineer","synthesize and mix audio content","Expert in audio synthesis, music generation, and mixing. Use when creating podcasts, background scores, or multi-track audio layering using mcp-chirp3-go, mcp-lyria-go, mcp-gemini-go, mcp-nanobanana-go, and mcp-avtool-go.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[982,985,988,989],{"name":983,"slug":984,"type":16},"Audio","audio",{"name":986,"slug":987,"type":16},"Creative","creative",{"name":9,"slug":8,"type":16},{"name":990,"slug":991,"type":16},"Vertex AI","vertex-ai","2026-07-12T07:39:16.623879",{"slug":994,"name":994,"fn":995,"description":996,"org":997,"tags":998,"stars":956,"repoUrl":957,"updatedAt":1005},"genmedia-image-artist","generate and edit AI images","Expert in AI image generation and editing. Use when the user needs high-quality textures, character-consistent visuals, or image-to-image editing using mcp-nanobanana-go.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[999,1000,1001,1004],{"name":986,"slug":987,"type":16},{"name":9,"slug":8,"type":16},{"name":1002,"slug":1003,"type":16},"Image Generation","image-generation",{"name":990,"slug":991,"type":16},"2026-07-12T07:39:15.372822",{"slug":1007,"name":1007,"fn":1008,"description":1009,"org":1010,"tags":1011,"stars":956,"repoUrl":957,"updatedAt":1021},"genmedia-producer","produce multi-step media content","Expert media production assistant. Use when requested to help with storyboarding, podcast creation, audio assembly, or complex multi-step media workflows using the GenMedia MCP servers (Veo, Lyria, Gemini TTS, NanoBanana).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1012,1013,1014,1015,1018],{"name":983,"slug":984,"type":16},{"name":986,"slug":987,"type":16},{"name":9,"slug":8,"type":16},{"name":1016,"slug":1017,"type":16},"Media","media",{"name":1019,"slug":1020,"type":16},"Video","video","2026-07-12T07:39:09.672849",{"slug":1023,"name":1023,"fn":1024,"description":1025,"org":1026,"tags":1027,"stars":956,"repoUrl":957,"updatedAt":1031},"genmedia-video-editor","edit and compose video content","Expert in video composition, editing, and format conversion. Use when the user wants to generate high-quality video, overlay images on video, concatenate clips, create GIFs, or sync audio to video using mcp-avtool-go and mcp-veo-go.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1028,1029,1030],{"name":986,"slug":987,"type":16},{"name":9,"slug":8,"type":16},{"name":1019,"slug":1020,"type":16},"2026-07-12T07:39:13.749081",{"slug":1033,"name":1033,"fn":1034,"description":1035,"org":1036,"tags":1037,"stars":956,"repoUrl":957,"updatedAt":1049},"genmedia-voice-director","generate expressive text-to-speech with Gemini","Expert in casting, directing, and generating expressive text-to-speech using Gemini TTS. Use this when the user needs virtual voice actor personas, expressive speech generation, or multiple variations of a voiceover (like \"take 3 on the bounce\").",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1038,1039,1040,1043,1046],{"name":983,"slug":984,"type":16},{"name":986,"slug":987,"type":16},{"name":1041,"slug":1042,"type":16},"Gemini","gemini",{"name":1044,"slug":1045,"type":16},"Speech","speech",{"name":1047,"slug":1048,"type":16},"Text-to-Speech","text-to-speech","2026-07-12T07:39:17.86673",80,{"items":1052,"total":1140},[1053,1063,1076,1089,1102,1113,1129],{"slug":1054,"name":1054,"fn":1055,"description":1056,"org":1057,"tags":1058,"stars":24,"repoUrl":25,"updatedAt":1062},"custom-golden-image-discovery","discover golden base images for GKE nodes","Expert at discovering golden base images for GKE custom nodes using technical specs or context clues.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1059,1060,1061],{"name":18,"slug":19,"type":16},{"name":9,"slug":8,"type":16},{"name":21,"slug":22,"type":16},"2026-07-12T07:39:30.888879",{"slug":1064,"name":1064,"fn":1065,"description":1066,"org":1067,"tags":1068,"stars":24,"repoUrl":25,"updatedAt":1075},"gke-ai-troubleshooting-handle-disruption-gpu-tpu","diagnose GPU and TPU workload disruptions","Diagnose and predict node disruption during Compute Engine host maintenance for GPU and TPU workloads.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1069,1072,1073,1074],{"name":1070,"slug":1071,"type":16},"Debugging","debugging",{"name":9,"slug":8,"type":16},{"name":21,"slug":22,"type":16},{"name":14,"slug":15,"type":16},"2026-07-28T05:34:18.149515",{"slug":1077,"name":1077,"fn":1078,"description":1079,"org":1080,"tags":1081,"stars":24,"repoUrl":25,"updatedAt":1088},"gke-ai-troubleshooting-jobset-interruption","diagnose GKE JobSet interruptions","Systematically diagnose GKE JobSet interruptions, restarts, and preemptions for AI\u002FML training workloads. Identifies preemption events, maintenance interruptions, bad host VMs, unhealthy pods, and coordinator worker failures.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1082,1083,1084,1085],{"name":1070,"slug":1071,"type":16},{"name":9,"slug":8,"type":16},{"name":21,"slug":22,"type":16},{"name":1086,"slug":1087,"type":16},"Observability","observability","2026-07-12T07:40:04.511878",{"slug":1090,"name":1090,"fn":1091,"description":1092,"org":1093,"tags":1094,"stars":24,"repoUrl":25,"updatedAt":1101},"gke-ai-troubleshooting-skill-creation-guide","create GKE troubleshooting skill bundles","Expert instructions for building high-quality GKE troubleshooting skills. Codifies Step 0 context rules, zero-hallucination signatures, and explicit LQL\u002FPromQL query requirements.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1095,1096,1097,1098],{"name":871,"slug":872,"type":16},{"name":908,"slug":909,"type":16},{"name":9,"slug":8,"type":16},{"name":1099,"slug":1100,"type":16},"Technical Writing","technical-writing","2026-07-12T07:39:50.73484",{"slug":1103,"name":1103,"fn":1104,"description":1105,"org":1106,"tags":1107,"stars":24,"repoUrl":25,"updatedAt":1112},"gke-ai-troubleshooting-tpu-connection-failure-vbar-oom","diagnose GKE TPU connection failures","Diagnose and prevent `vbar_control_agent` segfaults and OOMs caused by race conditions during TPU device resets and frequent metrics collection (e.g. every 3s). Use when TPU slice initialization fails or `vbar_control_agent` crashes on TPU v6e nodes.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1108,1109,1110,1111],{"name":1070,"slug":1071,"type":16},{"name":9,"slug":8,"type":16},{"name":21,"slug":22,"type":16},{"name":14,"slug":15,"type":16},"2026-07-12T07:39:49.482979",{"slug":1114,"name":1114,"fn":1115,"description":1116,"org":1117,"tags":1118,"stars":24,"repoUrl":25,"updatedAt":1128},"gke-app-onboarding","containerize and deploy apps to GKE","Workflows for containerizing and deploying applications to GKE for the first time.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1119,1122,1123,1124,1125],{"name":1120,"slug":1121,"type":16},"Containers","containers",{"name":18,"slug":19,"type":16},{"name":9,"slug":8,"type":16},{"name":21,"slug":22,"type":16},{"name":1126,"slug":1127,"type":16},"Onboarding","onboarding","2026-07-12T07:39:41.935837",{"slug":1130,"name":1130,"fn":1131,"description":1132,"org":1133,"tags":1134,"stars":24,"repoUrl":25,"updatedAt":1139},"gke-backup-dr","configure GKE backup and disaster recovery","Workflows for configuring Backup for GKE and disaster recovery.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":11},[1135,1136,1137,1138],{"name":18,"slug":19,"type":16},{"name":9,"slug":8,"type":16},{"name":21,"slug":22,"type":16},{"name":942,"slug":943,"type":16},"2026-07-12T07:39:34.806995",25]