Nacos Enterprise Edition capacity specifications and billing guide - Microservices Engine

MSE Nacos Enterprise Edition runs on dedicated resources, guarantees a 99.99% Service-Level Agreement (SLA), and delivers 300% faster service push performance compared to self-managed open source Nacos. The following sections describe the capacity thresholds and queries per second (QPS) benchmarks for each Enterprise Edition specification.

What's included in Enterprise Edition

Dynamic MCP registration

Enterprise Edition is compatible with Mesh Configuration Protocol (MCP) in open source Nacos 3.0. It provides dynamic service registration, protocol migration, and real-time optimization to support the upgrade requirements of AI agents and microservices models.

Dynamic service registration -- Register MCP services through an SDK or the console. Standardize service discovery and manage services visually.
One-click protocol migration -- Convert existing HTTP services to MCP services with a single click. Define protocol mapping rules with JSON templates to integrate with the MCP ecosystem.
Live tool description updates -- Update MCP tool descriptions in the console at runtime to improve AI agent call performance and developer efficiency without modifying code.

99.99% SLA with dedicated resources

Enterprise Edition guarantees 99.99% availability through dedicated underlying core resources.

High availability (HA) architecture -- Supports year-round uninterrupted operation for industries with strict uptime requirements, such as finance, energy, and state-owned enterprises.
Resilience under load -- Maintains efficiency and stability during traffic peaks and complex operational scenarios, supporting the continuous operation of critical business applications.
Resource isolation -- Eliminates shared-resource contention, significantly reducing the risk of failures.

Encrypted data source management

Enterprise Edition provides end-to-end encrypted data source management. It integrates with Key Management Service (KMS) to enable dynamic key rotation and zero-configuration access for mainstream databases, including ApsaraDB RDS, ApsaraDB for Redis, and PolarDB.

Automatic key rotation -- KMS encrypts and stores credentials, then rotates them on schedule, eliminating the threat of plaintext credential exposure. When a key changes, the system synchronizes the updated configuration automatically -- no manual steps required.
Seamless application access -- Connect to data sources through configuration without modifying code. Key rotation happens transparently at runtime.
End-to-end encryption -- Data source configuration is encrypted both in transit and at rest to ensure the security of sensitive information.

Capacity thresholds

Use these thresholds to size your cluster. All values are for a 3-node cluster.

Key terms:

Connection -- Each Nacos client establishes a connection to the Nacos server. Typically, one pod uses one connection. If a pod uses both the service registry and the configuration center, it requires two connections.
Safe level -- The recommended operating ceiling for core business. Staying below this level provides stability during unexpected traffic bursts, cluster restarts, or upgrades.
Alert level -- The upper boundary. Sustained traffic above this level may compromise cluster stability.

Note: These benchmarks are based on client version 1.x. For better performance, we recommend that you upgrade your client to version 2.x.

Specification	Nodes	Connections		QPS		Transactions per second (TPS)
Specification	Nodes	Safe level	Alert level	Safe level	Alert level	Safe level	Alert level
Small.Platinum.x2	3	1,200	1,800	1,200	1,800	600	900
Medium.Platinum.x1	3	2,400	3,600	2,400	3,600	1,200	1,800
Medium.Platinum.x2	3	4,800	7,200	4,800	7,200	2,400	3,600
Large.Platinum.x1	3	9,600	14,400	9,600	14,400	4,800	7,200

Scaling for different node counts

To estimate capacity for a cluster with more or fewer nodes, scale these values proportionally.

For example, a 5-node Medium.Platinum.x1 cluster supports approximately:

Connections (safe level): 2,400 / 3 x 5 = 4,000
QPS (safe level): 2,400 / 3 x 5 = 4,000
TPS (safe level): 1,200 / 3 x 5 = 2,000

Antifragility throttling

MSE Nacos Enterprise Edition enables antifragility mode by default. When resource usage or API traffic reaches a threshold, the system automatically applies throttling to protect cluster stability.

The following limits apply per node. To calculate limits for a multi-node cluster, multiply the per-node value by the number of nodes.

Specification	Connections	Same-config publications (times/min)	Config publications (TPS)	Config queries (QPS)	Service pushes (TPS)
Small.Platinum.x2	800	20	100	200	800
Medium.Platinum.x1	1,600	20	100	400	1,600
Medium.Platinum.x2	3,200	20	100	800	3,200
Large.Platinum.x1	6,400	20	100	1,600	6,400

Note: The capacity thresholds table shows limits for the entire 3-node cluster. The antifragility throttling table shows limits per individual node. When you compare the two tables, multiply the per-node throttling values by 3 to get cluster-level equivalents.