Azure Managed Redis vs Upstash: provisioned capacity and global writes

Azure Managed Redis runs Redis Enterprise, and Azure has no first-party managed Valkey service as of September 2026. Choose Azure for provisioned capacity and active-active geo-replication. Upstash’s Redis-compatible service offers regional or global databases and convenient HTTP access, but global writes still travel to one primary region. Its distributed reads add read-your-writes coordination to the application design.

Azure Managed Redis is Microsoft’s recommended replacement for Azure Cache for Redis. Azure Cache for Redis Enterprise and Enterprise Flash retire March 31, 2027. Basic, Standard, and Premium retire September 30, 2028. This comparison focuses on Azure Managed Redis. (Microsoft retirement FAQ)

At a glance

About Azure Managed Redis

Azure Managed Redis is Microsoft’s managed caching service based on Redis Enterprise. It runs in Azure and offers public network access or private connectivity through Private Link. Customers choose a service tier and capacity configuration for their workload.

About Upstash

Upstash is a hosted data service with a proprietary Redis-compatible engine. It provides Redis TCP and HTTPS REST access, with regional and global database options in supported AWS and Google Cloud locations.

Recent updates

September 2, 2026 - Momento introduced Cluster and Flex capacity for Momento Cache. Both provide dedicated, managed Valkey resources. Cluster exposes topology choices, while Flex selects topology within capacity bounds, giving teams a choice in how they configure capacity while Momento operates the cluster.

Distinguish the engine from API compatibility

Azure Managed Redis runs Redis Enterprise, while Upstash uses a proprietary Redis-compatible engine. Microsoft’s Valkey path on AKS leaves cluster operations with your team. That self-hosted deployment falls outside AKS SLA and Azure support coverage.

For teams running Valkey on Kubernetes, Momento’s Valkey Operator is a proven enterprise choice with AKS support, rolling upgrades, and zone-aware placement. The Operator is self-hosted software, separate from the managed Momento Cache service.

Compare write locations and read-your-writes coordination

Azure’s active-active geo-replication lets applications write in multiple regions, with eventual consistency between them. Your application must detect regional failures and route to another endpoint. (Azure reliability guidance)

Upstash serves reads from nearby read regions but sends writes to one primary. The provider cautions that Global Database may not suit write-heavy workloads.

Upstash’s read-your-writes coordination is overly complex. Supported SDKs track sync tokens within a client. Independent clients must propagate them, and REST clients pass a token header. A replica must reach the token’s state before serving the read. Those tokens pad transfer costs, while replication waiting adds latency compared with reading directly from the primary. A single TCP session has causal guarantees. (Upstash’s read-your-writes guide)

Plan for client growth and capacity limits

Azure leaves resizing decisions within tier and configuration limits to your team. Smaller options cap connections at 75,000 or less, while more expensive SKUs support up to 200,000. (Microsoft’s connection limits)

Upstash does not support autoscaling. Its fixed plans have resource limits, and exhausted limits without a suitable plan upgrade can block traffic or writes. That makes expected growth part of the initial capacity choice. (Upstash pricing and limits)

Try Momento Cache

For a regional cache, consider Momento when you want Valkey’s broad industry backing without operating the cluster yourself. Momento also has an HTTP API.

Momento Cache puts dedicated, high-performance Valkey behind a managed gateway that keeps client integration stable as your application grows. Choose Cluster for direct control over instance types, shards, and replicas, or Flex to set capacity and replication bounds while Momento selects the topology. Both include built-in autoscaling and fully managed cluster operations.

Talk to a Momento solution engineer to evaluate the capacity model and configuration that fit your workload.

Frequently asked questions

Does Azure’s 99.999% SLA cover a single cache endpoint?

No. It requires at least three eligible HA, zone-redundant instances in three regions using active geo-replication. Group downtime requires all endpoints to be unavailable. (Azure SLA)

Can Azure automatically route applications after a regional failure?

No. Applications must detect the failure and choose another endpoint. Active geo-replication keeps other regions writable, but client rerouting remains application work.

What changes when leaving Azure Cache for Redis?

Migration to Azure Managed Redis requires new endpoints and credentials. Azure Managed Redis clusters by default, so the client changes depend on your source tier.

How do support response times compare?

Initial response starts from 15 minutes for Azure and from 30 minutes for Upstash.