Redis Flex v2 in Redis Cloud extends Redis databases beyond RAM by using both memory and SSD-backed Flash storage. Hot data stays in RAM for low-latency access, while less frequently accessed data is moved to Flash to optimize capacity and cost.

This guide explains how to select and size Flex databases, change the RAM-to-Flash ratio safely, monitor performance, and troubleshoot common behaviors in the Redis Cloud Console or API.

Redis Flex v2 enhances the original auto-tiering model by enabling both keys and values to be tiered to SSD, while keeping frequently accessed data in RAM. Redis Cloud supports configurable RAM-to-SSD ratios from 10% to 50%, allowing customers to optimize performance, throughput, and cost based on workload requirements. This selectable range is available for Pro and Software offerings. For Essentials, the RAM-to-SSD ratio is fixed at the time of database creation.

This article includes At a Glance, Prerequisites, Quick Fix Table, Sizing Guidance, Validation Checklist, Operations & Monitoring, Scaling, and Troubleshooting Summary.

At a Glance / Versioning Summary

Prerequisites

• Plan and size requirements: Verify that your Redis Cloud plan supports Redis Flex. 

  • Redis Flex is available in Redis Cloud Essentials and Pro subscriptions. 

    • For Redis Cloud Pro, Flex databases require a minimum database size of 250 GB, or 500 GB when high availability (HA) is enabled. 

    • Confirm that the selected database size meets these minimum requirements before creating the database.

• Change window: Schedule ratio edits during a maintenance window; resizes are online but can temporarily affect performance.

• Backups: Ensure recent backups and a successful restore test before changes.

• Observability: Use Cloud metrics, Redis Insight, or your connected APM tool.

Quick Fix Table

Deployment Planning

Choosing When to Use Flex v2

Flex v2 is ideal for:

• Large datasets with predictable locality.

• Workloads tolerant of slightly higher latency for cold reads.

• Cost-sensitive deployments that benefit from Flash capacity expansion.

Avoid Flex v2 when:

• Access is uniform across the entire dataset.

• The hot data set ≈ total dataset size.

• Required features (for example, Active-Active) are not supported in your plan but will be available in future.

Plan Selection & Sizing

Sizing Example (10 TB Database, 30 % RAM)

Understanding RAM Ratios

Redis Flex databases support configurable RAM ratios from 10% to 50%.

The RAM ratio determines how much of the database capacity is allocated to memory versus SSD storage.

Lower RAM ratios generally provide lower cost and are often suitable for workloads with a small hot working set.

Higher RAM ratios provide more memory capacity for frequently accessed data and may improve throughput and latency consistency for workloads with larger active datasets.

Choosing the correct RAM ratio depends on:

• Dataset size

• Working set size

• Throughput requirements

• Latency requirements

• Cost objectives

The optimal configuration varies by workload and should be validated using application performance metrics.

Validation Checklist

Operating & Monitoring

Console path: Databases → (your database) → Configuration → Memory tiering

Ratio changes are online but trigger an active resize that moves data between RAM and Flash. Use a maintenance window and re-baseline performance after completion.

Monitor regularly:

• RAM hit ratio, % values in RAM, latency (avg/p99), ops /sec

• Evictions and expirations

• Flash IOPS and throughput

• Replica lag

Use Redis Cloud metrics, Redis Insight, or your APM dashboards.

Monitoring Tip

High SSD utilization alone is not an indication of a problem. Redis Flex is designed to serve colder data from SSD as part of normal operation.

Investigate further only when increased flash activity is accompanied by:

• Increased latency

• Reduced throughput

• Memory pressure

• Application performance degradation

Focus on latency, throughput, memory utilization, and application performance when determining whether a RAM ratio adjustment is necessary.

Scaling

Backups

Redis Flex is not a system of record.
Maintain scheduled backups and verify restore jobs before making ratio or size changes.

Troubleshooting Summary Table

Where to Look

Console metrics, database events, API status, and Redis Insight dashboards.