Redis for Very Large Datasets

Redis for Very Large Datasets

Unprecedented Performance. Unparalleled Savings

As the world moves to analytics at the speed of transactions, applications are required to process and analyze extremely large datasets instantaneously. Older generations of big data tools that took hours and days are increasingly outdated. Redis on Flash allows you to meet the new seconds and minutes expectation for large dataset processing, storage and analytics while optimizing costs by using a combination of RAM and Flash memory. Next generation persistent memory technologies like NVMe-based SSDs are benchmarked to deliver 3 Million operations/second at sub-millisecond latencies with 90% of the data in Flash and 10% in RAM, on a single x86 server

Advantages of Analyzing Very Large Datasets in memory with Redis on Flash

  • Blazing fast throughput for operational and analytical workloads, delivering analytics at the speed of business transactions
  • Optimal cost to performance tradeoffs with configurability of RAM: Flash ratios depending on the criticality of the output
  • Redis data structures and Lua scripting offer performance boosts to operational and analytical scenarios due to in-database execution, reduction of code complexity and lower resource requirements
  • Available as Redis on Flash with standard x86 servers as well as standard public cloud instances that use SSDs
  • Available in AWS private clouds as Redis Enterprise VPC
  • Also available with IBM Power systems

>3 Million Operations /sec, <1ms Latency

Redis Labs – Intel Benchmark Snapshot

>2 Million Operations /sec, <1ms Latency

Redis Labs – Samsung Benchmark Snapshot

Redis on Flash in Google Cloud with Kubernetes

Related Resources

Redis on Flash benchmark with Intel NVMe
Redis on Flash benchmark with Samsung NVMe
Spot.IM Case Study : Redis on Flash for very large datasets
Redis Enterprise - IBM Solution Brief: See more about Redis Enterprise and IBM Power with CAPI Flash