Knowledge Base

Q: What are the key configuration options? Key parameters (CLI arg / env var / default): - --listen / CACHEE_LISTEN / 0.0.0.0:6380 — Client listen address - --upstream / CACHEE_UPSTREAM / redis://127

Q: How does memory management work? Set maxmemory via config or CONFIG SET maxmemory 512mb. Graduated watermarks: admission tightens at 80%, aggressive eviction at 90%, safety-only writes at 95%. Evi

Q: Does Cachee support Lua scripting? Yes. EVAL script numkeys key [key ...] arg [arg ...] executes Lua scripts atomically. EVALSHA runs pre-loaded scripts by SHA1 hash. SCRIPT LOAD pre-compiles scri

Q: Does Cachee support Pub/Sub? Yes. SUBSCRIBE channel, UNSUBSCRIBE, PUBLISH channel message. Pattern subscriptions via PSUBSCRIBE pattern (glob patterns like user.*). Messages delivered as ["pmessag

Q: How many tests does Cachee have? 227 automated tests: 60 backend unit tests, 10 cluster tests, and 157 proxy integration tests covering all commands, data structures, eviction policies, transactio

Q: What data structures does Cachee support? Eight data structures with full Redis command coverage: 1. Strings — GET, SET, MSET, INCR, APPEND, GETDEL, GETEX, SETEX, PSETEX. Use cases: Key-value pai

Q: How fast is the L1 cache? L1 cache reads complete in under 1 microsecond (~0.085µs measured). This is approximately 100,000x faster than a typical Redis network round-trip (100-500µs). L1 runs in-

Q: Does Cachee support transactions? Yes. MULTI starts a transaction, commands are queued, EXEC executes atomically. WATCH provides optimistic locking — if watched keys change before EXEC, the transa

Q: How many Redis commands does Cachee support? Total: 133+ unique Redis commands across 8 data structures: - Strings: ~25 commands - Hashes: ~15 commands - Lists: ~15 commands - Sets: ~12 commands -

Q: What monitoring is available? Built-in Prometheus metrics on configurable /metrics endpoint: cache hit/miss rates, latency percentiles, eviction counts, memory usage, key counts by type, connectio

Q: What is the L1/L2 tiered architecture? L1 is an in-process adaptive cache in your application's memory space. L2 is the upstream Redis (your existing Redis, or Cachee's managed service). Cache hit

Q: What is the connection pool architecture? Cachee maintains a configurable connection pool (default 32, max 1000) to the upstream Redis. Pool timeout is configurable (default 5000ms). A circuit bre

Q: What is Cachee? Cachee is a high-performance tiered caching system with full Redis RESP protocol compatibility. It features a sub-microsecond L1 in-process cache that sits in front of any Redis-co

Q: What deployment models does Cachee support? Three models: (1) Managed — Cachee-hosted multi-tenant service with API key authentication, namespace isolation, and per-tenant rate limiting. (2) Sidec

Q: What are the connection limits? Default max clients: 10,000 total. Default max per IP: 100. Both configurable via CLI args or environment variables. Connections exceeding limits receive -ERR max n

Q: What protocol does Cachee use? Standard Redis RESP (REdis Serialization Protocol). Any Redis client library works without modification — redis-py, ioredis, Jedis, go-redis, StackExchange.Redis, et

Q: What eviction policies are available? Eight policies, configurable at runtime via CONFIG SET maxmemory-policy: 1. tiny-cachee (default) — Proprietary adaptive policy using Count-Min Sketch freque

Q: Does Cachee support consumer groups? Yes. Full consumer group support: - XGROUP CREATE key group $ MKSTREAM — Create group with auto-stream creation - XGROUP DESTROY key group — Delete group - XRE

Q: What runtime debugging commands are available? - SLOWLOG GET [count] — Commands exceeding configurable threshold (default 10ms). Each entry includes ID, timestamp, duration, full command, client a

Q: What session management exists? Configurable max session lifetime (default 86400s / 24 hours). Connections must re-authenticate after session expiry.

Q: What rate limiting is available? Per-tenant token bucket rate limiting with configurable RPS and burst limits. Default burst = 2x RPS. Exceeding the limit returns -ERR rate limit exceeded.

Q: How does multi-tenant authentication work? When CACHEE_REQUIRE_AUTH=true, clients must AUTH api_key before sending commands. API keys are SHA-256 hashed and mapped to tenant configs with namespace

Q: What cache services does Cachee support? Cachee supports Redis and Redis-compatible services (AWS ElastiCache, Azure Cache, Google Cloud Memorystore), Memcached, Cloudflare KV, KeyDB, Aerospike, H

Q: What is the L1/L2 cache architecture? L1 (Memory Cache) is an ultra-fast, in-memory cache layer with sub-millisecond response times. L2 (Distributed Cache) is your existing distributed cache like

Q: Is Cachee a replacement for Redis or Memcached? No. Cachee is designed to work alongside your existing Redis, Memcached, or other cache infrastructure. It acts as an intelligent routing layer that

Q: Do I need to change my existing cache infrastructure? No. Cachee works as an overlay on top of your existing cache services. You keep your Redis, Memcached, or other cache infrastructure exactly a

Q: What protocols does Cachee support? Cachee supports standard cache protocols including Redis protocol (RESP), Memcached protocol, and REST APIs for cloud-based caches like Cloudflare KV.

Q: Is there an API for Cachee? Yes. Cachee provides a full REST API for configuration, monitoring, management, and analytics. API documentation is available at cachee.ai/docs with SDKs for major lang

Q: Can I try Cachee before committing? Yes. We offer a Free tier (1,000 AI credits, 1M requests/month) at no cost, plus an interactive sandbox demo. The sandbox is a simulation for demonstration purp

Q: What credentials do I need to connect Cachee to my cache? You'll need to provide credentials, tokens, network routes, and permissions required to connect the Service to your cache services. Cachee

Q: How long does integration take? Most integrations can be completed in under 60 seconds. Our setup guide walks you through the process step by step. Complex enterprise deployments may require addit

Q: Does Cachee store my data? No. Cachee is designed as a stateless gateway. We do not intentionally persist the contents of your cached data (keys, values, payloads) to our storage. Your data reside

Q: How do I integrate Cachee with my application? Integration typically requires updating your cache client configuration to point to the Cachee gateway endpoint -- often just one line of code. We pr

Q: Can Cachee guarantee specific performance improvements? No. Performance gains, cost savings, and routing outcomes vary by workload, provider behavior, network conditions, and customer configuratio

Q: How does Cachee work technically? Cachee functions as a managed, stateless proxy/gateway that receives cache requests from your application, applies AI-powered routing and optimization logic, forw

Q: What makes Cachee different from other caching solutions? Unlike traditional caching solutions, Cachee doesn't replace your cache -- it enhances it. Our Rust-native engine achieves 95%+ L1 hit rat

Q: What is Cachee? Cachee is a managed, AI-powered caching gateway that enhances your existing cache infrastructure. It acts as a stateless proxy that sits in front of your cache services (Redis, Mem

Q: What happens if my payment fails? If a subscription or reload charge fails, your service will be automatically suspended immediately without further notice until payment is successfully processed.

Q: What is Auto-Reload and how does it work? Auto-Reload is enabled by default and automatically purchases additional credits when your balance reaches a threshold or is depleted. This ensures uninte

Q: Is there a free trial? Yes! We offer a Free tier with 1,000 AI credits and 1M requests per month at no cost. You can also try our interactive sandbox demo to experience Cachee instantly (simulatio

Q: What are the support hours? Standard support is available during US business hours. Enterprise plans can include 24/7 support options. Check your plan details for specific coverage.

Q: How do I get technical support? Support options vary by plan: Free (community support), Growth (email support), Scale (priority support), Volume/Enterprise (dedicated support). Contact support@cac

Q: How do I contact support? Email: support@cachee.ai. Chat: Use the widget on cachee.ai. For urgent enterprise issues, contact your dedicated support representative.

Q: Is there an SLA for support response times? Service level agreements are only available through Enterprise plans and must be specified in an Order Form. Standard plans do not include guaranteed re

Q: Where can I find documentation? Documentation is available at cachee.ai/docs, including quick start guides, API reference, SDK documentation, configuration guides, and best practices.

Q: How do I report a bug or issue? Report bugs via support@cachee.ai with details including your account ID, timestamp, error messages, and steps to reproduce the issue.

Q: Can I request new features? Yes. Submit feature requests to support@cachee.ai. We review all feedback, though we cannot guarantee implementation of specific requests.

Q: Is there a community forum or Slack channel? Free tier users have access to community support channels. Check cachee.ai/community for available forums and discussion groups.

Q: How do I escalate an urgent issue? For urgent issues on paid plans, email support@cachee.ai with "URGENT" in the subject line. Enterprise customers should contact their dedicated support represent

Q: How does credit-based billing work? Credits are prepaid usage units consumed based on actual usage as measured by our systems. You can monitor your credit balance in the dashboard. Credits have no