Best practices Limitations Pricing Monitoring Using data tiering Restoring data from backup into clusters with data tiering enabled

Data tiering in ElastiCache

ElastiCache for Valkey or Redis OSS clusters that comprise a replication group and use a node type from the r6gd family have their data tiered between memory and local SSD (solid state drives) storage. Data tiering provides a new price-performance option for Valkey or Redis OSS workloads by utilizing lower-cost solid state drives (SSDs) in each cluster node in addition to storing data in memory. It is ideal for workloads that access up to 20 percent of their overall dataset regularly, and for applications that can tolerate additional latency when accessing data on SSD.

On ElastiCache clusters with data tiering, ElastiCache monitors the last access time of every item it stores. When available memory (DRAM) is fully consumed, ElastiCache uses a least-recently used (LRU) algorithm to automatically move infrequently accessed items from memory to SSD. When data on SSD is subsequently accessed, ElastiCache automatically and asynchronously moves it back to memory before processing the request. If you have a workload that accesses only a subset of its data regularly, data tiering is an optimal way to scale your capacity cost-effectively.

Note that when using data tiering, keys themselves always remain in memory, while the LRU governs the placement of values on memory vs. disk. In general, we recommend that your key sizes are smaller than your value sizes when using data tiering.

Data tiering is designed to have minimal performance impact to application workloads. For example, assuming 500-byte String values, you can expect an additional 300 microseconds of latency on average for requests to data stored on SSD compared to requests to data in memory.

With the largest data tiering node size (cache.r6gd.16xlarge), you can store up to 1 petabyte in a single 500-node cluster (500 TB when using 1 read replica). Data tiering is compatible with all Valkey or Redis OSS commands and data structures supported in ElastiCache. You don't need any client-side changes to use this feature.

Topics

Best practices
Limitations
Pricing
Monitoring
Using data tiering
Restoring data from backup into clusters with data tiering enabled

Best practices

We recommend the following best practices:

Data tiering is ideal for workloads that access up to 20 percent of their overall dataset regularly, and for applications that can tolerate additional latency when accessing data on SSD.
When using SSD capacity available on data-tiered nodes, we recommend that value size be larger than the key size. When items are moved between DRAM and SSD, keys will always remain in memory and only the values are moved to the SSD tier.

Limitations

Data tiering has the following limitations:

You can only use data tiering on clusters that are part of a replication group.
The node type you use must be from the r6gd family, which is available in the following regions: us-east-2, us-east-1, us-west-2, us-west-1, eu-west-1, eu-central-1, eu-north-1, eu-west-3, ap-northeast-1, ap-southeast-1, ap-southeast-2, ap-south-1, ca-central-1 and sa-east-1.
You must use an engine that is Valkey 7.2 or later, or a Redis OSS 6.2 or later.
You cannot restore a backup of an r6gd cluster into another cluster unless it also uses r6gd.
You cannot export a backup to Amazon S3 for data-tiering clusters.
Online migration is not supported for clusters running on the r6gd node type.
Scaling is not supported from a data tiering cluster (for example, a cluster using an r6gd node type) to a cluster that does not use data tiering (for example, a cluster using an r6g node type). For more information, see Scaling ElastiCache.
Auto scaling is supported on clusters using data tiering for Valkey version 7.2 and later, and Redis OSS version 7.0.7 and later. For more information, see Auto Scaling Valkey and Redis OSS clusters
Data tiering only supports volatile-lru, allkeys-lru, volatile-lfu, allkeys-lfu and noeviction maxmemory policies.
Forkless save is supported for Valkey version 7.2 and later, and Redis OSS version 7.0.7 and later. For more information, see How synchronization and backup are implemented.
Items larger than 128 MiB are not moved to SSD.

Pricing

R6gd nodes have 4.8x more total capacity (memory + SSD) and can help you achieve over 60 percent savings when running at maximum utilization compared to R6g nodes (memory only). For more information, see ElastiCache pricing.

Monitoring

ElastiCache offers metrics designed specifically to monitor the performance clusters that use data tiering. To monitor the ratio of items in DRAM compared to SSD, you can use the CurrItems metric at Metrics for Valkey and Redis OSS. You can calculate the percentage as: (CurrItems with Dimension: Tier = Memory * 100) / (CurrItems with no dimension filter).

If the configured eviction policy allows, then ElastiCache will start evicting items when the percentage of items in memory decreases below 5 percent. On nodes configured with noeviction policy, write operations will receive an out of memory error.

It is still recommended that you consider scaling out for Cluster Mode Enabled clusters or scaling up for Cluster Mode disabled clusters when the percentage of items in memory decreases below 5 percent. For more information on scaling see Scaling clusters in Valkey or Redis OSS (Cluster Mode Enabled). For more information on metrics for Valkey or Redis OSS clusters that use data tiering see Metrics for Valkey and Redis OSS.

Using data tiering

When creating a cluster as part of a replication group, you use data tiering by selecting a node type from the r6gd family, such as cache.r6gd.xlarge. Selecting that node type automatically enables data tiering.

For more information on creating a cluster, see Creating a cluster for Valkey or Redis OSS.

When creating a replication group using the AWS CLI, you use data tiering by selecting a node type from the r6gd family, such as cache.r6gd.xlarge and setting the --data-tiering-enabled parameter.

You cannot opt out of data tiering when selecting a node type from the r6gd family. If you set the --no-data-tiering-enabled parameter, the operation will fail.

For Linux, macOS, or Unix:


aws elasticache create-replication-group \
   --replication-group-id redis-dt-cluster \
   --replication-group-description "Redis OSS cluster with data tiering" \
   --num-node-groups 1 \
   --replicas-per-node-group 1 \
   --cache-node-type cache.r6gd.xlarge \
   --engine redis \   
   --cache-subnet-group-name default \
   --automatic-failover-enabled \
   --data-tiering-enabled

For Windows:


aws elasticache create-replication-group ^
   --replication-group-id redis-dt-cluster ^
   --replication-group-description "Redis OSS cluster with data tiering" ^
   --num-node-groups 1 ^
   --replicas-per-node-group 1 ^
   --cache-node-type cache.r6gd.xlarge ^
   --engine redis ^   
   --cache-subnet-group-name default ^
   --automatic-failover-enabled ^
   --data-tiering-enabled

After running this operation, you will see a response similar to the following:


{
    "ReplicationGroup": {
        "ReplicationGroupId": "redis-dt-cluster",
        "Description": "Redis OSS cluster with data tiering",
        "Status": "creating",           
        "PendingModifiedValues": {},
        "MemberClusters": [
            "redis-dt-cluster"
        ],
        "AutomaticFailover": "enabled",
        "DataTiering": "enabled",
        "SnapshotRetentionLimit": 0,
        "SnapshotWindow": "06:00-07:00",
        "ClusterEnabled": false,
        "CacheNodeType": "cache.r6gd.xlarge",       
        "TransitEncryptionEnabled": false,
        "AtRestEncryptionEnabled": false
    }
}

Restoring data from backup into clusters with data tiering enabled

You can restore a backup to a new cluster with data tiering enabled using the (Console), (AWS CLI) or (ElastiCache API). When you create a cluster using node types in the r6gd family, data tiering is enabled.

To restore a backup to a new cluster with data tiering enabled (console)

Sign in to the AWS Management Console and open the ElastiCache console at https://console.aws.amazon.com/elasticache/.
From the navigation pane, choose Backups.
In the list of backups, choose the box to the left of the backup name you want to restore from.
Choose Restore.
Complete the Restore Cluster dialog box. Be sure to complete all the Required fields and any of the others you want to change from the defaults.
1. Cluster ID – Required. The name of the new cluster.
2. Cluster mode enabled (scale out) – Choose this for a Valkey or Redis OSS (cluster mode enabled) cluster.
3. Node Type – Specify cache.r6gd.xlarge or any other node type from the r6gd family.
4. Number of Shards – Choose the number of shards you want in the new cluster (API/CLI: node groups).
5. Replicas per Shard – Choose the number of read replica nodes you want in each shard.
6. Slots and keyspaces – Choose how you want keys distributed among the shards. If you choose to specify the key distributions complete the table specifying the key ranges for each shard.
7. Availability zone(s) – Specify how you want the cluster's Availability Zones selected.
8. Port – Change this only if you want the new cluster to use a different port.
9. Choose a VPC – Choose the VPC in which to create this cluster.
10. Parameter Group – Choose a parameter group that reserves sufficient memory for Valkey or Redis OSS overhead for the node type you selected.
When the settings are as you want them, choose Create.

For more information on creating a cluster, see Creating a cluster for Valkey or Redis OSS.

When creating a replication group using the AWS CLI, data tiering is by default used by selecting a node type from the r6gd family, such as cache.r6gd.xlarge and setting the --data-tiering-enabled parameter.

You cannot opt out of data tiering when selecting a node type from the r6gd family. If you set the --no-data-tiering-enabled parameter, the operation will fail.

For Linux, macOS, or Unix:


aws elasticache create-replication-group \
   --replication-group-id redis-dt-cluster \
   --replication-group-description "Redis OSS cluster with data tiering" \
   --num-node-groups 1 \
   --replicas-per-node-group 1 \
   --cache-node-type cache.r6gd.xlarge \
   --engine redis \   
   --cache-subnet-group-name default \
   --automatic-failover-enabled \
   --data-tiering-enabled \
   --snapshot-name my-snapshot

For Linux, macOS, or Unix:


aws elasticache create-replication-group ^
   --replication-group-id redis-dt-cluster ^
   --replication-group-description "Redis OSS cluster with data tiering" ^
   --num-node-groups 1 ^
   --replicas-per-node-group 1 ^
   --cache-node-type cache.r6gd.xlarge ^
   --engine redis ^   
   --cache-subnet-group-name default ^
   --automatic-failover-enabled ^
   --data-tiering-enabled ^
   --snapshot-name my-snapshot

After running this operation, you will see a response similar to the following:


{
    "ReplicationGroup": {
        "ReplicationGroupId": "redis-dt-cluster",
        "Description": "Redis OSS cluster with data tiering",
        "Status": "creating",           
        "PendingModifiedValues": {},
        "MemberClusters": [
            "redis-dt-cluster"
        ],
        "AutomaticFailover": "enabled",
        "DataTiering": "enabled",
        "SnapshotRetentionLimit": 0,
        "SnapshotWindow": "06:00-07:00",
        "ClusterEnabled": false,
        "CacheNodeType": "cache.r6gd.xlarge",        
        "TransitEncryptionEnabled": false,
        "AtRestEncryptionEnabled": false
    }
}

Warning Javascript is disabled or is unavailable in your browser.

To use the Amazon Web Services Documentation, Javascript must be enabled. Please refer to your browser's Help pages for instructions.

Document Conventions

DNS names and underlying IP

Preparing a cluster in ElastiCache