convez.ai
AgenteDevOps & InfraestruturaAvançado
aerospike-ce-ecosystem

Aerospike CE Ecosystem Plugins

aerospike-ce-ecosystem/aerospike-ce-ecosystem-plugins

Deploy Aerospike em K8s e desenvolva apps Python com suporte completo

Ver no GitHub
InicialPluginScore · de 0 a 100

Agent Claude para deploy e gerenciamento de clusters Aerospike em Kubernetes via ACKO e desenvolvimento de aplicações Python com aerospike-py. Oferece skills para implantação, operações dia-2, testes e-2-e, acesso ao cliente Python completo e diagnóstico de problemas.

Instalação

$npx convez-ai aerospike-ce-ecosystem/aerospike-ce-ecosystem-plugins

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
37%
Último commit
há 3 semanas
Versão
v1.7.3
Issues abertas
13
Contribuidores
1

Casos de uso

  • Deployar clusters Aerospike em Kubernetes com operador ACKO
  • Escalar, atualizar e gerenciar configurações dinâmicas de clusters
  • Desenvolver aplicações FastAPI integradas com Aerospike
  • Diagnosticar e resolver problemas em clusters ACKO
  • Gerenciar índices, UDFs Lua e registros via CLI ackoctl

Veredito

Plugin robusto para quem trabalha com Aerospike em ambientes containerizados, cobrindo tanto deployment em K8s quanto desenvolvimento backend. Requer familiaridade com Kubernetes e bancos dados distribuídos; a curva de aprendizado é íngreme mas necessária para casos de uso enterprise.

README

A tradução para o português deste README ainda está na fila.

aerospike-ce-ecosystem

Claude Code plugin for the Aerospike CE ecosystem — deploy clusters on Kubernetes with ACKO and build Python applications with aerospike-py.

Installation

From GitHub (recommended)

Add the repository as a marketplace, then install:

# Step 1: Add as marketplace
claude plugin marketplace add aerospike-ce-ecosystem/aerospike-ce-ecosystem-plugins

# Step 2: Install the plugin
claude plugin install aerospike-ce-ecosystem

Project-scoped install

To install only for the current project:

claude plugin marketplace add aerospike-ce-ecosystem/aerospike-ce-ecosystem-plugins
claude plugin install aerospike-ce-ecosystem -s project

Verify installation

claude plugin list
# Should show: aerospike-ce-ecosystem@aerospike-ce-ecosystem ✔ enabled

Skills

ACKO (Aerospike CE Kubernetes Operator)

SkillTriggerDescription
acko-deploy"deploy Aerospike on Kubernetes"Step-by-step guide to deploy CE clusters via AerospikeCluster CR — 9 scenarios from minimal to full-featured
acko-operations"scale Aerospike cluster"Day-2 operations: scale, upgrade, dynamic config, warm restart, ACL, pause/resume, troubleshooting
acko-config-reference(background)CE 8.1 parameter reference, CRD-to-conf mapping, webhook validation rules
acko-e2e-test"ACKO e2e test", "kind cluster test"End-to-end test playbook — canonical scenarios, Ginkgo labels, mandatory helm install operator setup, release-verification performance checks

aerospike-py (Python Client)

SkillTriggerDescription
aerospike-py-api"use aerospike-py to ..."Full API reference — AsyncClient, CRUD, batch, CDT, query, expressions, observe, admin
aerospike-py-fastapi"build FastAPI app with Aerospike"Production-ready FastAPI patterns — lifespan, DI, CRUD endpoints, error handling, metrics

Cluster Manager CLI (ackoctl)

SkillTriggerDescription
ackoctl"ackoctl", "manage Aerospike connection", "browse records", "register UDF", "scale cluster"Drive aerospike-cluster-manager via the ackoctl CLI — connections, cluster info, records/sets, queries, secondary indexes, operator notes, raw asinfo, K8s AerospikeCluster CRs, admin (users/roles), and Lua UDF modules. Multi-cluster ACKO friendly via kubeconfig-style contexts.
acko-debugging"CrashLoopBackOff", "phase=Error", "reconcile failure", "migration stuck"Systematic 6-step diagnosis procedure for ACKO clusters with CE 8.1 pitfalls and a remediation matrix. Routes both data-plane and K8s-plane probes through ackoctl (ackoctl cluster info, ackoctl info, ackoctl query exec, ackoctl k8s cluster get/list, ackoctl k8s cluster logs, ackoctl k8s cluster events); falls back to kubectl/asinfo when ackoctl is unavailable.

Ecosystem support

SkillTriggerDescription
bug-reporter"버그 제보", "where do I file this issue", "report this to GitHub"Routes a bug report to the correct aerospike-ce-ecosystem repo by symptom and generates a ready-to-paste GitHub issue body with the required reproduction context

ackoctl integration

Live cluster access goes through the ackoctl Go CLI, which calls aerospike-cluster-manager's REST API at /api/v1/*. There is no MCP HTTP server in the loop, so there is no DNS-rebinding allowlist, no Origin allowlist, and no per-cluster claude mcp add step — the security surface is just the existing cluster-manager FastAPI auth (bearer JWT + workspace ACL).

# 1. Install ackoctl
curl -fsSL https://raw.githubusercontent.com/aerospike-ce-ecosystem/ackoctl/main/install.sh | sh

# 2. Point a context at a running cluster-manager
ackoctl config set-context kind-local \
  --server=http://localhost:8000/api \
  --workspace-id=default
ackoctl config use-context kind-local

# 3. Use the cluster from any skill or chat
#    "in kind-local, list connections and get a sample record from sample_set"
ackoctl connection list
ackoctl record list <CONN_ID> --namespace=test --set=sample_set --page-size=5

For multi-cluster ACKO, register one context per cluster-manager instance — naming convention <env> or <region>:

ackoctl config set-context prod-us \
  --server=https://acm.prod-us.example.com/api \
  --token=$ACKOCTL_TOKEN \
  --workspace-id=prod-us
ackoctl --context=prod-us k8s cluster list

Auth is bearer-token only — users bring their own OIDC JWT (Keycloak CLI, browser device flow, …). The workspace ACL on cluster-manager scopes every call; destructive verbs (delete, remove, mutation info --allow-write, k8s cluster scale) require --yes for non-interactive runs.

The ackoctl skill covers install, configuration, and every command (connection / cluster / set / record / query / index / note / k8s / info / admin / udf). See skills/ackoctl/reference/commands.md for a one-line-per-command cheat sheet.

Prerequisites

For ACKO skills

  • Kubernetes cluster (kind, minikube, EKS, GKE, etc.)
  • ACKO operator installed
  • kubectl configured

For aerospike-py skills

  • Python 3.10+ (aerospike-py requires >=3.10; wheels cover 3.10–3.14)
  • Running Aerospike server (localhost:3000 by default)

Benchmark Results

Evaluated with Claude Sonnet on real deployment tasks:

Metricwith skillwithout skillImprovement
Pass Rate100%87.5%+12.5%
Token Usage25,07736,262-30.8%
Completion Time73.8s137.4s-46.3%
Tool Calls8.030.0-73.3%

Key findings:

  • Skills prevent the #1 CE mistake (enterprise-only config in CE clusters)
  • aerospike-py-specific patterns (NamedTuple access, Depends injection) are taught by skills
  • 7 production bonus features generated with skill vs 0 without

Development

Validate the plugin

claude plugin validate /path/to/aerospike-ce-ecosystem-plugins

Run eval tests

# With skill
claude -p "Deploy a single-node Aerospike cluster for development on a kind cluster" --model sonnet

# Without skill (baseline comparison)
claude -p "Deploy a single-node Aerospike cluster for development on a kind cluster" --model sonnet --disable-plugins

The e2e_pytest/ scenario suite includes a multi-cluster + Keycloak OIDC scenario (common-cluster web + dev/prod operator-cluster API + bitnami/keycloak realm bootstrap). See ACKO multi-cluster docs.

License

MIT

Ver o original no GitHub

Bom para

Relacionados