elasticache-cluster-deployer
Provisions Amazon ElastiCache clusters with production defaults: Redis OSS self-designed clusters vs cluster mode enabled (shards and replicas), Memcached clusters, replication groups (primary + read replicas), multi-AZ with automatic failover, node type sizing, subnet group creation, security groups, parameter groups, at-rest encryption (KMS), in-transit encryption (TLS), Redis AUTH tokens, Redis shard/slot mapping, backup and snapshot management (automated + manual), Global Datastore (cross-region replication), auto-scaling, and online cluster resizing (scale up/down with use-online-resharding). Emits a READY_TO_DEPLOY checklist with verification commands. Use when creating an ElastiCache cluster, setting up Redis replication groups, enabling multi-AZ. Triggers: create elasticache cluster, redis replication group, cluster mode enabled, elasticache multi-az failover, elasticache encryption kms tls, redis auth token, elasticache global datastore, online resharding, elasticache auto scaling, memcached...
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-24
- Stars
- ★ 4
- Forks
- 0
01 前言 Front Matter
- name
- elasticache-cluster-deployer
- license
- Apache-2.0
- metadata
- { "tags": "aws, elasticache, redis, memcached, cloudops, deploy, databases, provisioning, replication-group, cluster-mode, multi-az, encryption, global-datastore", "phase": "1", "author": "Jacky Chan — AWS Community Builder", "domain": "aws-cloudops", "family": "Databases", "version": "0.1.0", "keywords": "aws, elasticache, redis, memcached, cloudops, deploy, provisioning, replication group, cluster mode, multi-az, failover, encryption, kms, tls, auth token, global datastore, online resharding, auto scaling", "task_type": "deploy", "complexity": "high", "entry_point": "false", "skill_class": "capability", "when_to_use": "Invoke when the user wants to create an Amazon ElastiCache cluster (Redis OSS or Memcached), configure a Redis replication group with multi-AZ automatic failover, enable cluster mode with shards and read replicas, configure at-rest encryption (KMS) or in-transit encryption (TLS), set Redis AUTH tokens, create a Global Datastore for cross-region replication, perform online cluster resizing, or configure ElastiCache auto-scaling. Do NOT invoke for Amazon MemoryDB for Redis, Amazon DynamoDB, or self-managed Redis/Memcached on EC2.", "dependencies": "aws-orchestrator", "requires_llm": "true", "verdict_shape": "READY_TO_DEPLOY | PREREQUISITES_MISSING", "lifecycle_status": "active", "supports_pipeline": "true" }
- description
- Provisions Amazon ElastiCache clusters with production defaults: Redis OSS self-designed clusters vs cluster mode enabled (shards and replicas), Memcached clusters, replication groups (primary + read replicas), multi-AZ with automatic failover, node type sizing, subnet group creation, security groups, parameter groups, at-rest encryption (KMS), in-transit encryption (TLS), Redis AUTH tokens, Redis shard/slot mapping, backup and snapshot management (automated + manual), Global Datastore (cross-region replication), auto-scaling, and online cluster resizing (scale up/down with use-online-resharding). Emits a READY_TO_DEPLOY checklist with verification commands. Use when creating an ElastiCache cluster, setting up Redis replication groups, enabling multi-AZ. Triggers: create elasticache cluster, redis replication group, cluster mode enabled, elasticache multi-az failover, elasticache encryption kms tls, redis auth token, elasticache global datastore, online resharding, elasticache auto scaling, memcached...
- compatibility
- Agent runtime that reads SKILL.md (Claude Code, Cursor, Windsurf, Codex, Gemini). For live deployment: AWS CLI v2 with elasticache access. Works with Terraform aws_elasticache_replication_group / aws_elasticache_cluster / aws_elasticache_subnet_group resources and CloudFormation AWS::ElastiCache::ReplicationGroup / AWS::ElastiCache::CacheCluster / AWS::ElastiCache::SubnetGroup templates.
02 资源文件 20
- 参考文档eval/test-cases.yaml640 B
- 参考文档evals/baselines/memcached-cluster-basic.md313 B
- 参考文档evals/baselines/memcached-with-encryption.md284 B
- 参考文档evals/baselines/missing-subnet-group.md279 B
- 参考文档evals/baselines/redis-global-datastore.md312 B
- 参考文档evals/baselines/redis-replication-group-multiaz.md321 B
- 参考文档evals/eval_config.yaml48 B
- 参考文档evals/evals.json1.2 KB
- 参考文档evals/prompts/memcached-cluster-basic.md260 B
- 参考文档evals/prompts/memcached-with-encryption.md231 B
- 参考文档evals/prompts/missing-subnet-group.md222 B
- 参考文档evals/prompts/redis-global-datastore.md317 B
- 参考文档evals/prompts/redis-replication-group-multiaz.md466 B
- 参考文档examples/README.md280 B
- 参考文档references/advanced-patterns.md2.9 KB
- 参考文档references/config-patterns.md1.8 KB
- 参考文档references/error-handling.md1.0 KB
- 参考文档references/provisioning-cli-commands.md6.9 KB
- 参考文档references/worked-examples.md1.3 KB
- 参考文档skill.usf.yaml4.4 KB
03 SKILL.md 正文
ElastiCache Cluster Deployer
An AWS CloudOps agent skill that provisions Amazon ElastiCache clusters with correct defaults. The skill walks the operator through Redis OSS cluster mode vs non-cluster mode, Memcached topology, replication groups, multi-AZ automatic failover, node type sizing, subnet groups, security groups, parameter groups, encryption, AUTH tokens, shard/slot mapping, backup management, Global Datastore, auto-scaling, and online cluster resizing — then emits a READY_TO_DEPLOY checklist with verification commands.
Activation keywords
create ElastiCache cluster, Redis replication group, cluster mode enabled, ElastiCache multi-AZ failover, ElastiCache encryption KMS TLS, Redis AUTH token, ElastiCache Global Datastore, online resharding, ElastiCache auto scaling, Memcached cluster.
STRICT output contract
When this skill is invoked with an ElastiCache-provisioning request
(create a cluster, configure a replication group, enable multi-AZ
failover, set up encryption, create a Global Datastore, scale a cluster
online, or a partial configuration), the agent MUST respond with the
READY_TO_DEPLOY checklist defined in the "Output format" section using
the literal all-caps labels ELASTICACHE_CLUSTER:, VERDICT:,
CHECKLIST:, and VERIFICATION_COMMANDS:. Do NOT preface the checklist
with prose, headings, or disclaimers — emit the block as the first lines
of the response. This contract is what assertion-based evals and
downstream provisioning pipelines rely on; deviating from the literal
labels breaks automation silently.
If any prerequisite is missing, the verdict is PREREQUISITES_MISSING
with a specific gap citation in the checklist (marked [✗]), and
READY_TO_DEPLOY MUST NOT also appear.
FORBIDDEN output patterns
-
NEVER emit
VERDICT: READY_TO_DEPLOYwhen any CHECKLIST item is[✗]. If even one prerequisite is unmet, the verdict MUST bePREREQUISITES_MISSING. A mixed-verdict block is a contract violation. -
NEVER omit the
ELASTICACHE_CLUSTER:header line. It is the parse anchor for downstream provisioning pipelines. Substituting a markdown heading (## ElastiCache) or a lowercase variant breaks automation silently. -
NEVER mark encryption as "TBD", "optional", or "post-creation." At-rest (KMS) and in-transit (TLS) encryption are creation-time-only for Redis replication groups — they CANNOT be toggled on after the cluster exists. The CHECKLIST MUST show an explicit encryption decision (
EnabledorDisabled), never a deferred one. -
NEVER show multi-AZ failover as
Enabledwhenreplicas-per-node-groupis 0. Multi-AZ automatic failover requires at least one replica per shard as a promotion target.Multi-AZ: Enabledwith zero replicas is a silent no-op — failover has no target to promote. -
NEVER allow the snapshot window to overlap the maintenance window in the CHECKLIST. Overlap causes skipped backups or delayed maintenance. The output MUST include both windows and confirm non-overlap with at least 1 hour of separation.
-
NEVER use
--num-cache-clusters(non-cluster mode) and--num-node-groups(cluster mode) interchangeably in the CHECKLIST. These are different API paths requiring different client libraries (cluster-aware vs standard Redis client). The topology line MUST explicitly identify which mode is selected. -
NEVER omit the
VERIFICATION_COMMANDS:block. The copy-pasteable AWS CLI commands are how the operator confirms the deployment succeeded. A checklist without verification commands is incomplete.
Quick navigation
| Section | When to read |
|---|---|
| Prerequisites | Always — verify before provisioning |
| Step 1 — Redis vs Memcached engine selection | Engine choice |
| Step 2 — Cluster mode enabled vs self-designed (Redis) | Redis topology |
| Step 3 — Replication groups and multi-AZ failover | High availability |
| Step 4 — Node type sizing | Capacity planning |
| Step 5 — Subnet groups and security groups | Network + security |
| Step 6 — Parameter groups | Engine configuration |
| Step 7 — Encryption at rest (KMS) and in transit (TLS) | Security |
| Step 8 — Redis AUTH tokens | Authentication |
| Step 9 — Backup and snapshot management | Data durability |
| Step 10 — Global Datastore (cross-region replication) | DR / cross-region |
| Step 11 — Auto-scaling | Capacity elasticity |
| Step 12 — Online cluster resizing (use-online-resharding) | Scale operations |
| Step 13 — Recent features | Latest |
| NEVER do these things | Review before signing off |
| Output format | The literal checklist template |
| references/redis-topology-and-failover.md | Cluster mode + failover detail |
| references/security-and-encryption.md | Encryption + AUTH detail |
Mindset
One-line takeaway: ElastiCache for Redis cluster mode enabled partitions data across shards using hash slots (16,384 total), each shard has a primary and 0-5 read replicas, multi-AZ failover promotes a replica automatically when the primary fails. Encryption at rest and in transit must be enabled at creation time — they CANNOT be toggled on after the cluster exists. Online resharding lets you add/remove shards with zero downtime using the use-online-resharding flag.
Three misconceptions dominate ElastiCache misdesign at provisioning time:
Moved verbatim → references/advanced-patterns.md § "Mindset misconceptions". Load when: you need the full argument behind encryption/cluster-mode/engine misconceptions.
Configuration dependency graph (novel heuristic)
ElastiCache configurations are NOT independent. Encryption must be enabled at creation time. Cluster mode is immutable without online resharding. Subnet groups must exist before the cluster. Use this graph to sequence provisioning.
| Configuration | Hard dependencies | Silent failure / immutability | Enables downstream |
|---|---|---|---|
| Subnet group | Subnets in >= 2 AZs (multi-AZ); same VPC | Cross-VPC subnets rejected at API | cluster creation |
| Security group | VPC exists; port 6379 (Redis) or 11211 (Memcached) | No ingress rule = client timeout (not API error) | client connectivity |
| Replication group (Redis) | Subnet group + security group exist | Encryption must be set at creation — CANNOT toggle | cluster endpoints |
| Cluster mode enabled | --num-node-groups > 1 at creation |
Switching modes requires migration | horizontal scaling |
| Multi-AZ + failover | >= 1 replica per shard; nodes across >= 2 AZs | Multi-AZ without replicas = no failover target | high availability |
| At-rest encryption (KMS) | KMS key exists; set at creation | CANNOT enable on existing cluster | compliance |
| In-transit encryption (TLS) | Set at creation | CANNOT enable on existing cluster | compliance |
| AUTH token | In-transit encryption MUST be enabled first | AUTH without TLS = token in cleartext | authentication |
| Backup (automated) | Retention period > 0; snapshot window set | Overlap with maintenance window = skipped snapshots | data durability |
| Global Datastore | Primary cluster in region A; same engine version | Encryption config must match across regions | cross-region DR |
| Online resharding | Cluster mode enabled; async operation | Non-cluster mode CANNOT be resharded | scale without downtime |
| Auto-scaling | Replication group exists; scaling policy defined | Min/max must be within node-group limits | capacity elasticity |
Moved verbatim → references/advanced-patterns.md § "Configuration dependency graph". Load when: the cross-dependency gotchas behind the table.
Expert heuristic: cluster mode enabled topology + replica promotion priority + snapshot window vs maintenance window overlap
Moved verbatim → references/config-patterns.md § "Expert heuristic: cluster mode topology, promotion priority, window overlap". Load when: designing shard/replica/AZ layout, promotion priorities, or non-overlapping windows.
Prerequisites (verify before provisioning)
Before emitting provisioning commands, verify these prerequisites. If any are missing, the verdict is PREREQUISITES_MISSING.
| Prerequisite | Why it matters | How to verify |
|---|---|---|
| VPC and subnets (>= 2 AZs) | Subnet group needs multi-AZ subnets | aws ec2 describe-subnets |
| Subnet group exists or creatable | Cluster requires a subnet group | aws elasticache describe-cache-subnet-groups |
| Security group (port 6379/11211) | Client connectivity | aws ec2 describe-security-groups |
| Node type selected | Memory, CPU, network capacity | aws elasticache describe-cache-engine-versions |
| Engine + version | Encryption needs Redis >= 6.0 for TLS 1.3 | Confirm feature compatibility |
| Encryption decision | MUST be at creation — cannot toggle later | Assess compliance requirements |
| AUTH token decision | Requires in-transit encryption first | Assess authentication requirements |
| Multi-AZ + failover decision | Requires >= 1 replica per shard | Assess availability requirements |
| Cluster mode decision | Different API path and client libs | Assess scaling requirements |
| Snapshot vs maintenance window | Must not overlap | Plan windows |
If any prerequisite is missing, output VERDICT: PREREQUISITES_MISSING
and cite the specific gap.
Step 1 — Redis vs Memcached engine selection
| Feature | Redis OSS | Memcached |
|---|---|---|
| Persistence | Yes (RDB, AOF) | No |
| Replication | Yes (primary + replicas) | No |
| Multi-AZ failover | Yes (auto promotion) | No |
| Cluster mode / sharding | Yes (cluster mode enabled) | Yes (auto-discovery) |
| Encryption at rest | Yes (KMS) | No |
| Encryption in transit | Yes (TLS) | No |
| AUTH token | Yes | No |
| Snapshots / backups | Yes | No |
| Global Datastore | Yes (cross-region) | No |
| Online resharding | Yes | No |
| Multi-threaded | No (per shard) | Yes |
| Data structures | Rich (lists, sets, hashes, streams) | Key-value only |
Choose Redis for persistence, replication, encryption, HA, or rich data structures. Choose Memcached for simple ephemeral multi- threaded caching with no durability.
Step 2 — Cluster mode enabled vs self-designed (Redis)
| Feature | Non-cluster (self-designed) | Cluster mode enabled |
|---|---|---|
| API params | --num-cache-clusters |
--num-node-groups + --replicas-per-node-group |
| Partitioning | Single primary (no sharding) | Hash slots across shards |
| Max memory | One node's capacity | Sum of all shards |
| Client library | Standard Redis client | Cluster-aware client |
| Online resharding | Not supported | Supported |
Non-cluster mode:
Moved verbatim → references/provisioning-cli-commands.md § "Step 2". Load when: creating a non-cluster replication group.
Cluster mode enabled:
Moved verbatim → references/provisioning-cli-commands.md § "Step 2". Load when: creating a cluster-mode-enabled replication group.
Common mistake: using --num-cache-clusters when you need cluster
mode. Cluster mode uses --num-node-groups and --replicas-per-node-group.
Step 3 — Replication groups and multi-AZ failover
A Redis replication group has a primary and 0-5 read replicas per shard. Multi-AZ with automatic failover promotes a replica in a different AZ when the primary fails.
Moved verbatim → references/provisioning-cli-commands.md § "Step 3". Load when: enabling multi-AZ + automatic failover at creation or via modify.
Critical: multi-AZ without at least one replica per shard is ineffective — failover has no target to promote.
Step 4 — Node type sizing
| Workload | Recommended type | Rationale |
|---|---|---|
| Small cache (< 5 GB) | cache.r6g.large | Cost-effective |
| Medium (5-25 GB) | cache.r6g.xlarge or cluster mode | Balance memory + CPU |
| Large (25-100 GB) | cache.r6g.2xlarge or more shards | Distribute memory |
| High throughput | cache.r6g.4xlarge+ or more shards | CPU + network |
For cluster mode, total capacity = (node memory) x (num shards). Per-shard throughput scales with replicas (reads) and shards (writes).
Step 5 — Subnet groups and security groups
Create a subnet group (requires >= 2 AZs for multi-AZ):
Moved verbatim → references/provisioning-cli-commands.md § "Step 5". Load when: creating the multi-AZ subnet group.
Security group (Redis 6379, Memcached 11211):
Moved verbatim → references/provisioning-cli-commands.md § "Step 5". Load when: creating the SG and authorizing inbound from the app SG.
Critical: the SG must allow inbound from the application SG (not CIDR) for least-privilege. Port differs: Redis = 6379, Memcached = 11211.
Step 6 — Parameter groups
Moved verbatim → references/provisioning-cli-commands.md § "Step 6". Load when: creating and modifying a Redis parameter group.
| Parameter | Default | Effect |
|---|---|---|
| maxmemory-policy | volatile-lru | Eviction policy when full |
| timeout | 0 | Idle client timeout |
| cluster-enabled | yes (cluster mode) | Redis cluster mode |
Step 7 — Encryption at rest (KMS) and in transit (TLS)
Encryption is creation-time-only for Redis replication groups. It CANNOT be toggled on after the cluster exists.
Moved verbatim → references/provisioning-cli-commands.md § "Step 7". Load when: creating with KMS + TLS encryption at creation.
| Encryption type | Parameter | Post-creation |
|---|---|---|
| At-rest (KMS) | --at-rest-encryption-enabled |
CANNOT toggle |
| In-transit (TLS) | --transit-encryption-enabled |
CANNOT toggle |
| AUTH token | --auth-token |
CANNOT toggle |
Critical: if you need encryption on an existing cluster, you must create a new encrypted cluster and migrate data.
Step 8 — Redis AUTH tokens
AUTH tokens require in-transit encryption (TLS) — otherwise the token is sent in cleartext.
Moved verbatim → references/provisioning-cli-commands.md § "Step 8". Load when: setting an AUTH token and storing it for rotation.
Step 9 — Backup and snapshot management
| Backup type | How | Retention |
|---|---|---|
| Automated | Daily snapshot in snapshot window | 0-35 days (0 = disabled) |
| Manual | On-demand via API | Until manually deleted |
Moved verbatim → references/provisioning-cli-commands.md § "Step 9". Load when: creating with automated snapshots or taking a manual snapshot.
Critical: the snapshot window MUST NOT overlap the maintenance window. If they overlap, snapshots may be skipped.
Step 10 — Global Datastore (cross-region replication)
Global Datastore provides cross-region replication for DR and low- latency multi-region reads.
Moved verbatim → references/provisioning-cli-commands.md § "Step 10". Load when: building a Global Datastore with primary + secondary.
Constraints: same engine version across regions; encryption config must match; secondary is read-only until failover.
Step 11 — Auto-scaling
Moved verbatim → references/provisioning-cli-commands.md § "Step 11". Load when: registering a scalable target and attaching the policy.
Auto-scaling uses CloudWatch metrics (EngineCPUUtilization, DatabaseMemoryUsagePercentage) as triggers. Scale-out adds shards; scale-in removes shards. Online resharding is performed automatically.
Step 12 — Online cluster resizing (use-online-resharding)
Online resharding adds/removes shards from a cluster-mode-enabled replication group with ZERO downtime.
Moved verbatim → references/provisioning-cli-commands.md § "Step 12". Load when: adding/removing shards and polling resharding status.
Key implication: online resharding is ONLY for cluster-mode-enabled groups. Non-cluster mode requires migration to a new cluster.
Step 13 — Recent features
Moved verbatim → references/advanced-patterns.md § "Step 13". Load when: deciding on r7g, serverless, TLS 1.3, or online vertical scaling.
NEVER do these things
-
NEVER assume encryption can be enabled after cluster creation. At-rest (KMS) and in-transit (TLS) encryption are creation-time-only for Redis replication groups. Existing non-encrypted clusters require migration to a new encrypted cluster.
-
NEVER set an AUTH token without enabling in-transit encryption. AUTH requires TLS. Setting AUTH without TLS sends the token in cleartext — a security vulnerability.
-
NEVER enable multi-AZ failover without at least one replica per shard. Multi-AZ without a replica to promote is a no-op. Always verify
--replicas-per-node-group >= 1. -
NEVER confuse cluster mode and non-cluster mode. Cluster mode uses
--num-node-groups+--replicas-per-node-group. Non-cluster uses--num-cache-clusters. Client libraries differ (cluster-aware vs standard). -
NEVER allow the snapshot window to overlap the maintenance window. Overlap causes skipped snapshots or delayed maintenance. Separate windows by at least 1 hour.
-
NEVER attempt online resharding on a non-cluster-mode cluster. Online resharding is ONLY for cluster-mode-enabled groups. Non- cluster mode requires migration.
-
NEVER create a Global Datastore with mismatched engine versions. Primary and secondary must use the same engine version. Mismatches cause replication failures.
-
NEVER use Memcached when you need persistence, replication, or encryption. Memcached does NOT support these features. Use Redis.
-
NEVER forget the security group port. Redis = 6379; Memcached = 11211. Wrong port is a silent failure (client timeout, not API error).
-
NEVER scale a cluster without monitoring the resharding status. Online resharding is asynchronous. Check
describe-replication-groupsstatus until "available" before further modifications.
Output format
ELASTICACHE_CLUSTER: <cluster-id> (<engine>, <node-type>, <topology>)
VERDICT: READY_TO_DEPLOY | PREREQUISITES_MISSING
CHECKLIST:
[✓|✗] Engine: Redis OSS | Memcached
[✓|✗] Topology: Cluster mode enabled (<N> shards x <R> replicas) | Non-cluster (<N> nodes) | Memcached (<N> nodes)
[✓|✗] Node type: <node-type>
[✓|✗] Subnet group: <subnet-group-name> (<N> AZs)
[✓|✗] Security group: <sg-id> (port <6379|11211>)
[✓|✗] Parameter group: <param-group-name>
[✓|✗] Multi-AZ failover: Enabled (<N> AZs) | Disabled
[✓|✗] At-rest encryption (KMS): Enabled (key <kms-key-id>) | Disabled
[✓|✗] In-transit encryption (TLS): Enabled | Disabled
[✓|✗] AUTH token: Set (Secrets Manager) | Not set
[✓|✗] Backup: Automated (retention <N> days, window <window>) | Manual | Disabled
[✓|✗] Snapshot window: <window> (no overlap with maintenance <window>)
[✓|✗] Global Datastore: <global-id> (primary <region>, secondary <region>) | Not configured
[✓|✗] Auto-scaling: Target tracking (<metric>, target <value>%) | Not configured
[✓|✗] Online resharding: Available (cluster mode) | Not available (non-cluster)
[✓|✗] Tags: <key=value list>
VERIFICATION_COMMANDS:
aws elasticache describe-replication-groups --replication-group-id <cluster-id> --region <region>
aws elasticache describe-cache-clusters --show-cache-node-info --region <region>
aws elasticache describe-global-replication-groups --show-global-nodegroups --region <region>
Worked example — Redis cluster mode enabled, 3 shards, KMS encryption, multi-AZ
ELASTICACHE_CLUSTER: session-store-prod (Redis OSS 7.0, cache.r6g.large, cluster mode enabled)
VERDICT: READY_TO_DEPLOY
CHECKLIST:
[✓] Engine: Redis OSS (version 7.0)
[✓] Topology: Cluster mode enabled (3 shards x 1 replica = 6 nodes total)
Shard 1: slots 0-5460, primary AZ-a, replica AZ-b (priority 100)
Shard 2: slots 5461-10922, primary AZ-b, replica AZ-c (priority 100)
Shard 3: slots 10923-16383, primary AZ-c, replica AZ-a (priority 100)
[✓] Node type: cache.r6g.large (2 vCPU, 12.93 GB RAM per node)
[✓] Subnet group: prod-redis-subnet-group (3 AZs: us-east-1a, us-east-1b, us-east-1c)
[✓] Security group: sg-0abc123def456 (port 6379, inbound from sg-app-workload only)
[✓] Parameter group: prod-redis-params (maxmemory-policy=allkeys-lru, timeout=300)
[✓] Multi-AZ failover: Enabled (3 AZs, automatic-failover + multi-az flags set)
[✓] At-rest encryption (KMS): Enabled (key arn:aws:kms:us-east-1:123456789012:key/a1b2c3d4-5678-90ef-1234-567890abcdef)
[✓] In-transit encryption (TLS): Enabled (creation-time-only — cannot toggle post-creation)
[✓] AUTH token: Set (Secrets Manager: elasticache/session-store-prod-auth-token, requires TLS)
[✓] Backup: Automated (retention 7 days, snapshot window 01:00-03:00 UTC daily)
[✓] Snapshot window: 01:00-03:00 UTC (no overlap with maintenance wed:04:00-wed:05:00 UTC — 1h gap confirmed)
[✓] Global Datastore: Not configured
[✓] Auto-scaling: Target tracking (EngineCPUUtilization, target 60%, min 3 shards, max 10 shards)
[✓] Online resharding: Available (cluster mode enabled — can add/remove shards with zero downtime)
[✓] Tags: Environment=production, Application=session-store, Owner=platform-team
VERIFICATION_COMMANDS:
aws elasticache describe-replication-groups --replication-group-id session-store-prod --region us-east-1
aws elasticache describe-cache-clusters --show-cache-node-info --region us-east-1
aws kms describe-key --key-id arn:aws:kms:us-east-1:123456789012:key/a1b2c3d4-5678-90ef-1234-567890abcdef --region us-east-1
Worked example — PREREQUISITES_MISSING (subnet group absent)
Moved verbatim → references/worked-examples.md § "Worked example". Load when: emitting a PREREQUISITES_MISSING checklist with a cited gap.
Error handling
Moved verbatim → references/error-handling.md § "Error handling". Load when: encryption unsupported, failover not triggering, clients cannot connect, resharding stuck.
References (load on demand)
- references/worked-examples.md — PREREQUISITES_MISSING worked example (subnet group absent)
- references/error-handling.md — triage: encryption unsupported, failover not triggering, connect failures, resharding stuck, skipped snapshots
- references/advanced-patterns.md — misconceptions, dependency-graph deep dive, recent AWS features
- references/config-patterns.md — topology + promotion priority + window overlap pattern (extended)
- references/provisioning-cli-commands.md — copy-pasteable CLI sequences for Steps 2-12 (extended)
Domain
AWS CloudOps / Amazon ElastiCache Cluster Provisioning & In-Memory Data Store Management.
AWS documentation
- ElastiCache for Redis — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/WhatIs.html
- Replication groups — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/Replication.html
- Cluster mode enabled — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/ClusterMode.html
- Multi-AZ with automatic failover — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/AutoFailover.html
- Encryption at rest and in transit — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/at-rest-encryption.html
- AUTH tokens — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/auth.html
- Global Datastore — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/Redis-Global-Datastore.html
- Online resharding — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/cluster-mode-resharding.html
- ElastiCache auto-scaling — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/AutoScaling.html
- Memcached — https://docs.aws.amazon.com/AmazonElastiCache/latest/mem-ug/WhatIs.html
- Snapshots and backups — https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/Snapshots.html
04 同仓库更多 Skills
xray-tracing-deployer
Deploys AWS X-Ray distributed tracing with production-grade configuration: X-Ray daemon deployment (EC2 systemd, ECS sidecar, EKS DaemonSet, Lambda built-in), SDK instrumentation (Java, Python, Node.js, Go, .NET) with annotations and metadata, sampling rules (default reservoir + rate vs custom rules with URL / method / service-name predicates), service map generation, trace query and retrieval, groups and insights, CloudWatch ServiceLens integration, IAM permissions (PutTraceSegments, GetSamplingRules), and latest features (Lambda Powertools tracing, OpenTelemetry / ADOT Collector support, W3C trace context). Emits a READY_TO_DEPLOY checklist with every configuration item verified. Use when enabling X-Ray on a new application, adding tracing to ECS / EKS / Lambda, validating a sampling configuration, or generating daemon deployment CLI and IaC templates. Triggers: X-Ray, distributed tracing, X-Ray daemon, sampling rules, service map, X-Ray SDK, OpenTelemetry, ADOT, Lambda Powertools, CloudWatch ServiceLens.
wellarchitected-workload-auditor
Audits AWS Well-Architected Tool workloads for review staleness, per-pillar high-risk issue counts, milestone tracking gaps, and remediation plan completeness. Emits a deterministic verdict (STALE_REVIEW | HIGH_RISK | CONFIG_GAP | OK) per workload with per-pillar risk breakdown, milestone trend analysis, and specific remediation. Use when reviewing Well-Architected workloads, checking WA review freshness, auditing high-risk issues per pillar, validating milestone coverage, or tracking remediation plans.
wellarchitected-review-operator
Operates AWS Well-Architected Tool reviews end-to-end — creates workloads, runs pillar reviews (operational excellence, security, reliability, performance, cost optimization, sustainability, Prosperity), answers questions with risk-tier rationale, generates reports and improvement plans, creates milestones, and integrates with Trusted Advisor and Well-Architected Labs. Runs deterministic pre-checks (workload existence, lens availability, pillar coverage), emits the exact update-answer / get-consolidated-report CLI behind a CONFIRM gate, and verifies risk-tier transitions. Emits READY | BLOCKED | COMPLETED. Use when creating a workload, answering pillar questions, generating an improvement plan, creating a milestone, or integrating with Trusted Advisor. Triggers: well-architected tool, well architected review, workload review, pillar, operational excellence, security, reliability, performance, cost optimization, sustainability, prosperity, improvement plan, milestone, trusted advisor.
wafv2-web-acl-deployer
Provisions AWS WAFv2 Web ACLs with correct production defaults: scope selection (CLOUDFRONT vs REGIONAL), managed rule groups (AWSManagedRulesCommonRuleSet, SQLiRuleSet, LinuxRuleSet, WindowsRuleSet, IPReputationList, BotControlRuleSet, AmazonIpReputationList, ATP), custom rules (byte match, IP set, regex pattern, geo match, size constraint, rate-based), rule priority ordering, logging destinations (Kinesis Firehose, CloudWatch Logs, S3), resource association (ALB, API Gateway, CloudFront), rate-based rules with aggregate keys (IP, forwarded IP), Challenge and CAPTCHA actions, and account takeover prevention. Emits a READY_TO_DEPLOY checklist. Use when creating a Web ACL, attaching managed rules, configuring rate limiting, enabling WAF logging, associating a Web ACL with an ALB/API Gateway/CloudFront, or configuring CAPTCHA/Challenge actions. Triggers: create WAF Web ACL, WAFv2, managed rule groups, WAF rate limiting, WAF logging, WAF CAPTCHA, WAF Challenge action, bot control, ATP managed rules.
wafv2-web-acl-auditor
Audits AWS WAFv2 Web ACL configurations to determine whether the ACL provides effective protection: default-action posture (Allow vs Block), managed-rule-group coverage gaps, rule effectiveness (BLOCK vs COUNT, shadow/bypass rules, stale exclusions), rate-based rule correctness, logging and visibility configuration, and text-transformation bypass vectors. Emits a deterministic verdict per Web ACL. Use when reviewing a WAFv2 Web ACL for security posture, checking managed-rule coverage, validating rate-limit configuration, auditing WAF logging, or hardening a WAF before production deployment. Triggers: WAF, WAFv2, Web ACL, managed rule group, AWSManagedRules, rate-based rule, CAPTCHA, Challenge, OverrideAction, Count mode, forwarded IP, Bot Control, ATP, text transformation, WAF logging, WAF visibility, OWASP, SQLi, XSS, RFI.
waf-rule-deployer
Provisions AWS WAFv2 rule sets with production defaults: Web ACL creation (CloudFront scope in us-east-1 vs Regional scope), rule groups (managed vs custom) with capacity planning in WCU (Web ACL Capacity Units, 1500 default per ACL), managed rule groups (AWSManagedRulesCommonRuleSet, AWSManagedRulesSQLiRuleSet, AWSManagedRulesAmazonIpReputationList, AWSManagedRulesLinuxRuleSet, AWSManagedRulesWindowsRuleSet, AWSManagedRulesUnixRuleSet), custom rules (byte-match, regex pattern set, geo-match, rate-based), IP set and regex pattern set resources, rule priority ordering (lowest number evaluated first), action types (allow, block, count, CAPTCHA, Challenge), bot control managed rule group, account takeover. Triggers: create WAF rule, WAF managed rule group, WAF custom rule, WAF byte match, WAF geo match, WAF rate based rule, WAF IP set, WAF regex pattern set, WAF WCU budget, WAF CloudFront ACL, WAF logging Kinesis Firehose, WAF bot control, WAF ATP, WAF label matching, WAF Challenge action, WAF CAPTCHA.