Overview of Spillover Load Balancing - NCOS-Load-Balancing-Modems-on-a-capable-Cradlepoint-Router/Overview-of-Spillover-Load-Balancing

Load Balancing Modems on a Capable Ericsson Cradlepoint Endpoint

ft:locale
en-US
ft:sourceName
Paligo_Prod
Document Type
Article

Ericsson Cradlepoint endpoints use spillover load balancing, which balances modem usage based on available bandwidth. These endpoints do not perform aggregate or channel-bonded load balancing, meaning they do not sum the bandwidth of attached modems, and combined bandwidth will not appear in online speed tests.

How Spillover Load Balancing Works

  1. Bandwidth-Based Allocation

    Each modem connected to the endpoint is assigned a bandwidth value. For example:

    • Modem A: 1 Mbps available bandwidth

    • Modem B: 2 Mbps available bandwidth

    When a client device initiates an IP connection:

    • The endpoint checks the available bandwidth of all modems.

    • The connection is assigned to the modem with the most available bandwidth (e.g., Modem B).

  2. Dynamic Bandwidth Updates After a connection is made, the remaining bandwidth of the selected modem is recalculated. New connections are similarly assigned to the modem with the most available bandwidth at the time. Modem A will only be used if its available bandwidth exceeds Modem B’s.

  3. Uneven Modem Usage Modems may not be utilized evenly, especially when there is a large difference in available bandwidth. For example:

    • If Modem B is a 4G modem and Modem A is a 3G modem, Modem A may see minimal usage.

    • Protocols like BitTorrent, which generate numerous IP connections, can utilize both modems effectively.

Limitations

  • A modem with two SIM slots can only connect to one SIM card at a time, preventing load balancing between SIM cards within the same modem. Load balancing between two SIM cards requires two separate cellular modems.

For details on all available load balancing algorithms, see Load Balancing Algorithms.Overview of Load Balancing