Menu
Le prompt qui vérifie d'abord la source est sélectionné par défaut. Vous pouvez passer à une commande directe ou télécharger une copie locale.
Lisez SKILL.md et les fichiers associés affichés par SkillsMP avant de décider de l'installer.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
npx skills add https://github.com/genlayerlabs/genlayer-studio --skill hosted-studio-debugLa commande reste sur une seule ligne. Faites défiler horizontalement pour la vérifier avant de la copier.
Vous préférez une copie locale ? Téléchargez les fichiers actuellement disponibles dans SkillsMP.
Basé sur la classification professionnelle SOC
Affichage de SKILL.md
| name | hosted-studio-debug |
| description | Debug GenLayer Studio deployments via ArgoCD CLI |
Debug GenLayer Studio deployments via ArgoCD CLI.
Kubernetes manifests are in sibling repo ../devexp-apps-workload (assume by default, ask user if not found):
devexp-apps-workload/workload/
├── dev/ # studio-dev, rally-studio-dev
├── stg/ # studio-stg
├── prd/ # studio-prd, rally-studio-prd
Each contains Deployments, Services, Ingresses, ExternalSecrets managed by ArgoCD.
To get complete visibility into Studio status, check all components across all replicas:
# 1. Overall app health
argocd app get <app>-workload
# 2. List all pods and their status
argocd app resources <app>-workload --kind Pod
# 3. Check consensus worker logs (all 4 replicas in prd)
argocd app logs <app>-workload --name studio-consensus-worker --tail 500 2>&1 | grep -iE "(error|exception|timeout|failed)"
# 4. Check JSON-RPC logs (all 2 replicas in prd)
argocd app logs <app>-workload --name studio-jsonrpc --tail 500 2>&1 | grep -iE "(error|exception|timeout|failed)"
# 5. Check webdriver logs
argocd app logs <app>-workload --name studio-webdriver --tail 200 2>&1 | grep -iE "(error|exception|crash)"
# Check app health
argocd app get <app>-workload
argocd app resources <app>-workload
argocd app resources <app>-workload --kind Pod
argocd app logs <app>-workload --name studio-consensus-worker -- 200
argocd app logs <app>-workload --name studio-jsonrpc -- 200
argocd app logs <app>-workload --name studio-consensus-worker -- 500 2>&1 | grep -i error
argocd app logs <app>-workload --name studio-jsonrpc -- 500 2>&1 | grep -i error
Important: The argocd app logs --name <container> command automatically aggregates logs from ALL pod replicas. This ensures full visibility across the entire deployment.
# Get logs from all consensus worker replicas (production has 4)
argocd app logs <app>-workload --name studio-consensus-worker --tail 500
# Get logs from all JSON-RPC replicas (production has 2)
argocd app logs <app>-workload --name studio-jsonrpc --tail 500
# Get logs from a specific pod (if needed for isolation)
argocd app logs <app>-workload --pod <pod-name> --tail 200
# List pods first to get pod names
argocd app resources <app>-workload --kind Pod
Consensus worker timeouts usually mean GenVM Manager is unresponsive. Check all 4 workers in production:
# Check for GenVM timeouts across all consensus worker replicas
argocd app logs <app>-workload --name studio-consensus-worker --tail 500 2>&1 | grep -E "(timeout|SocketTimeoutError|127.0.0.1:3999)"
# Empty stdout = GenVM never started
argocd app logs <app>-workload --name studio-consensus-worker --tail 500 2>&1 | grep "stdout=''"
Root cause: GenVM Manager (genvm-modules manager --port 3999) becomes unresponsive.
Fix: Worker restart (auto or manual) restarts GenVM Manager.
Key files:
backend/node/genvm/origin/base_host.py:463 - where timeouts occurbackend/node/base.py - Manager.create() spawns GenVMbackend/consensus/worker_service.py - worker startup/health# Check restart timestamps in logs
argocd app logs <app>-workload --name studio-consensus-worker --tail 500 2>&1 | grep -E "(Started|Uvicorn running)"
# Via kubectl
kubectl get pods -n <namespace> -o wide
kubectl get events -n <namespace> --sort-by='.lastTimestamp'
Contracts calling external APIs (Twitter, etc.) through proxies:
# Check for HTTP errors in contract execution
argocd app logs <app>-workload --name studio-consensus-worker --tail 500 2>&1 | grep -E "(HTTP|fetch|proxy|api)"
| Env | App | Namespace |
|---|---|---|
| dev | studio-dev-workload | studio-dev |
| stg | studio-stg-workload | studio-stg |
| prd | studio-prd-workload | studio-prd |
| rally-prd | rally-studio-prd-workload | rally-studio-prd |
| Component | Purpose | Prd Replicas | Common Issues |
|---|---|---|---|
| studio-consensus-worker | Tx processing | 4 | Timeouts, GenVM crashes |
| studio-jsonrpc | RPC API | 2 | DB connections |
| studio-webdriver | Browser sandbox | 1 | Memory, crashes |
| database-migration | Schema updates | 1 (job) | Lock contention |
Note: When debugging, always check logs from ALL replicas to get complete visibility. The argocd app logs --name <container> command handles this automatically.