Redis Performance & Scaling Expert Needed for Real Time Casino Platform
Budget: $15 – $25 USD
I run a real-time casino platform that processes high-frequency bets and instant wallet updates. At peak, we push around 50 k operations per second, and I need the p99 latency to stay below 5 ms. My top priority is eliminating the hot-key contention and uneven shard load that are currently causing unpredictable spikes.
What I’ve already tried
• Adjusted maxmemory settings and chosen data structures that suit our access patterns.
• Introduced pipelining, Lua scripts for atomic wallet operations, and basic client-side sharding.
Despite these efforts, wallet-related keys still become bottlenecks, and query latency climbs whenever traffic concentrates on them.
What I need from you
• Deep tuning of Redis for low-latency, high-throughput workloads—fixing hot keys, refining TTL usage, and validating our sharding strategy.
• A production-grade cluster, replication, and HA design that lets us scale seamlessly while safeguarding against data loss (RDB/AOF).
• Memory-use optimisation and eviction rules that keep hot data resident without starving the rest of the dataset.
• A solid monitoring and alerting approach (e.g., Redis-exporter, Grafana, or any stack you trust) with clear thresholds and runbooks.
Deliverables
1. An optimised redis.conf (or equivalent parameter set) ready for production.
2. A detailed cluster & failover blueprint, including scaling and maintenance procedures.
3. A monitoring/alerting plan with recommended dashboards and alert rules.
4. Concise documentation for my DevOps team explaining every decision and change.
Experience with iGaming, trading, or other high-load systems is a strong plus, as is familiarity with Lua scripting, Pub/Sub, and managed offerings such as ElastiCache or Redis Enterprise. If you’ve taken a platform from “works” to “works flawlessly under pressure,” I’d love to see how you can do the same here.
What I’ve already tried
• Adjusted maxmemory settings and chosen data structures that suit our access patterns.
• Introduced pipelining, Lua scripts for atomic wallet operations, and basic client-side sharding.
Despite these efforts, wallet-related keys still become bottlenecks, and query latency climbs whenever traffic concentrates on them.
What I need from you
• Deep tuning of Redis for low-latency, high-throughput workloads—fixing hot keys, refining TTL usage, and validating our sharding strategy.
• A production-grade cluster, replication, and HA design that lets us scale seamlessly while safeguarding against data loss (RDB/AOF).
• Memory-use optimisation and eviction rules that keep hot data resident without starving the rest of the dataset.
• A solid monitoring and alerting approach (e.g., Redis-exporter, Grafana, or any stack you trust) with clear thresholds and runbooks.
Deliverables
1. An optimised redis.conf (or equivalent parameter set) ready for production.
2. A detailed cluster & failover blueprint, including scaling and maintenance procedures.
3. A monitoring/alerting plan with recommended dashboards and alert rules.
4. Concise documentation for my DevOps team explaining every decision and change.
Experience with iGaming, trading, or other high-load systems is a strong plus, as is familiarity with Lua scripting, Pub/Sub, and managed offerings such as ElastiCache or Redis Enterprise. If you’ve taken a platform from “works” to “works flawlessly under pressure,” I’d love to see how you can do the same here.