AWS Security ChangesHomeSearch

AWS AmazonElastiCache: Clarify ElastiCache Multi-AZ and automatic failover behavior

Service: AmazonElastiCache · 2026-09-27 · Documentation low

File: AmazonElastiCache/latest/dg/AutoFailover.md

Summary

Adds a new 'Automatic failover and Multi-AZ' section explaining how Multi-AZ distributes replicas across Availability Zones and triggers automatic failover to the replica with the least replication lag, documents the behavior when Multi-AZ is disabled (longer write interruption, potential data loss without replicas), notes that failover without Multi-AZ does not protect against AZ-level failures, differentiates cluster mode enabled/disabled defaults, corrects the cluster-mode caveat for planned node replacements, adds a navigation entry, and adds a note distinguishing manual replica promotion from TestFailover-based failover testing.

Security assessment

The change is purely availability/resilience documentation: it describes replica placement across AZs, failover promotion, and data-loss risk when no replicas exist. It references no vulnerability, CVE, or advisory and gives no attacker/exploit-relevant guidance, so it is not security documentation in the authz/encryption/credential sense — only tangentially related via resilience and availability.

Evidence

+Multi-AZ is the recommended configuration for high availability in ElastiCache. When you enable Multi-AZ, ElastiCache distributes your replicas across different Availability Zones and enables automatic failover.

Diff

diff --git a/AmazonElastiCache/latest/dg/AutoFailover.md b/AmazonElastiCache/latest/dg/AutoFailover.md
index 786c31138..51f3b85be 100644
--- a//AmazonElastiCache/latest/dg/AutoFailover.md
+++ b//AmazonElastiCache/latest/dg/AutoFailover.md
@@ -7 +7 @@
-Enabling Multi-AZFailure scenarios with Multi-AZ responsesTesting automatic failoverNotes on Multi-AZ
+Automatic failover and Multi-AZEnabling Multi-AZFailure scenarios with Multi-AZ responsesTesting automatic failoverNotes on Multi-AZ
@@ -19 +19 @@ In case of planned node replacements initiated due to maintenance updates or sel
-  * For Valkey and Redis OSS clusters, the planned node replacements complete while the cluster serves incoming write requests. 
+  * For Valkey and Redis OSS clusters with cluster mode enabled, the planned node replacements complete while the cluster serves incoming write requests. 
@@ -37,0 +38,2 @@ Enabling ElastiCache Multi-AZ on your Valkey or Redis OSS cluster (in the API an
+  * Automatic failover and Multi-AZ
+
@@ -48,0 +51,27 @@ Enabling ElastiCache Multi-AZ on your Valkey or Redis OSS cluster (in the API an
+## Automatic failover and Multi-AZ
+
+Multi-AZ is the recommended configuration for high availability in ElastiCache. When you enable Multi-AZ, ElastiCache distributes your replicas across different Availability Zones and enables automatic failover. If the primary node fails, ElastiCache automatically promotes the replica with the least replication lag to primary. This provides both AZ-level resilience and rapid recovery from node failures with minimal write interruption.
+
+We recommend enabling Multi-AZ as a single configuration step for most use cases. Multi-AZ requires at least one replica in a different Availability Zone from the primary.
+
+###### How automatic failover works
+
+Automatic failover is the mechanism that Multi-AZ uses to recover from primary node failures. When Multi-AZ is enabled and the primary fails, ElastiCache promotes the replica with the least replication lag to primary. Writes can resume within seconds. ElastiCache then provisions a replacement replica in the Availability Zone where the failed primary was located.
+
+###### What happens without Multi-AZ
+
+If Multi-AZ is not enabled and a primary node fails, ElastiCache takes the primary offline and provisions a new primary node. If replicas exist, ElastiCache syncs the new primary from a replica. If no replicas exist, the new primary starts empty and data is lost unless you restore from a snapshot. This process results in a longer write interruption compared to automatic failover.
+
+###### Note
+
+While it is possible to have automatic failover enabled without Multi-AZ, replicas are not guaranteed to be in different Availability Zones in this configuration. Failover still promotes a replica to primary if one is available, but this does not protect against AZ-level failures. For most use cases, we recommend enabling Multi-AZ.
+
+Consider the following differences between cluster modes:
+
+  * **Valkey or Redis OSS (cluster mode enabled)** – Multi-AZ is enabled by default. Automatic failover is always required and cannot be disabled.
+
+  * **Valkey or Redis OSS (cluster mode disabled)** – We recommend enabling Multi-AZ and configuring at least one replica for high availability. You also have the option to enable or disable automatic failover independently.
+
+
+
+
@@ -581,0 +611,4 @@ When choosing the replica to promote to primary, ElastiCache chooses the replica
+###### Note
+
+Manual replica promotion is different from testing automatic failover. With manual promotion (described here), you must first disable Multi-AZ and automatic failover, and you can choose which replica becomes primary. To test automatic failover, use the `TestFailover` API. This requires automatic failover to be _enabled_ and simulates a primary node failure to verify your configuration. For more information, see Testing automatic failover.
+