All Products
Search

This Product

    Microservices Engine:Nacos Enterprise Edition capacity specifications

    Document Center

    Microservices Engine:Nacos Enterprise Edition capacity specifications

    Last Updated:Jul 09, 2026

    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.

    • QPS (queries per second) -- In the context of Nacos capacity planning, QPS counts the total number of Nacos protocol-layer requests that clients send to the Nacos server per second:

      QPS = service discovery requests (get instance list) + configuration pull requests (fetch configuration)

      This metric reflects only Nacos protocol traffic. It is independent of the business traffic your application routes through Server Load Balancer (SLB) or the HTTP/RPC calls between your microservices.

    Note: These benchmarks are based on Nacos client version 1.x. If you use Nacos client version 2.x, the actual performance ceilings are higher. For example, with a 16C32G 3-node cluster and client 2.x, the maximum number of configurations can reach 20,000 (requires a quota increase request) and the connection limit can reach 11,520. (Recommended) Upgrade to client version 2.x for better performance. To obtain 2.x-specific capacity benchmarks for your target specification, contact your account manager or submit a ticket.

    Specification

    Nodes

    Connections

    QPS

    Transactions per second (TPS)

    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.

    FAQ

    Q: How do I identify which type of rate limiting my MSE Nacos instance triggered, and what is the corresponding threshold?

    A:

    1. Identify the rate-limiting type. Check the metric name in your monitoring dashboard or alert notification. MSE Nacos antifragility throttling covers the following types:

      • Config publications (TPS) -- Triggered when configuration publish operations exceed the per-second transaction limit.

      • Config queries (QPS/TPS) -- Triggered when configuration pull requests exceed the per-second query or transaction limit.

      • Service pushes (TPS) -- Triggered when service push operations exceed the per-second transaction limit.

    2. Look up the threshold for your specification. Find your instance specification in the antifragility throttling table above. The values are per-node limits. For a multi-node cluster, multiply the per-node value by the number of nodes to get the cluster-level threshold.

      Example: for a Medium.Platinum.x1 3-node cluster, the per-node config query QPS limit is 400, so the cluster-level limit is 400 × 3 = 1,200 QPS.