Help Center/ Elastic IP/ FAQs/ Bandwidth/ What Are the Differences Between a Dedicated Bandwidth and a Shared Bandwidth? Can a Dedicated Bandwidth Be Changed to a Shared Bandwidth or the Other Way Around?
Updated on 2025-11-06 GMT+08:00
View PDF

What Are the Differences Between a Dedicated Bandwidth and a Shared Bandwidth? Can a Dedicated Bandwidth Be Changed to a Shared Bandwidth or the Other Way Around?

You can select a dedicated or shared bandwidth based on your requirements by referring to Table 1.

Table 1 Differences between dedicated and shared bandwidths

Item

Dedicated Bandwidth

Shared Bandwidth

Concepts

If you assign an EIP and do not add it to a shared bandwidth, the EIP uses a dedicated bandwidth by default no matter how the EIP is billed.

A dedicated bandwidth can only be used by one EIP. Each EIP can only be bound to one cloud resource, such as an ECS, a NAT gateway, or a load balancer.

A shared bandwidth can be used by multiple pay-per-use EIPs.

  • The shared bandwidth is dynamically allocated to the EIPs based on the actual usage conditions.
  • Adding an EIP to or removing an EIP from a shared bandwidth does not affect your services.

Features
  • Stable performance: The bandwidth is dedicated, so your use of the bandwidth is not affected by other resources. This type of bandwidth is ideal for applications requiring high-performance networks.
  • Quality of service (QoS): Guaranteed stable bandwidths and low latency are suitable for real-time applications.
  • Cost-effectiveness: Multiple EIPs sharing the same bandwidth can effectively reduce the overall costs. This type of bandwidth is suitable for users with limited budgets.
  • Flexibility: You can dynamically adjust the size of a shared bandwidth based on your requirements.
  • Performance fluctuation: When the bandwidth is used by multiple EIPs at the same time, the bandwidth allocated to each EIP is limited.

Applicable scenarios
  • Bandwidth preemption needs to be avoided to ensure Internet access for all EIPs at the same time.
  • High-performance and stable bandwidth is required, such as video streaming, online gaming, and financial transactions.
  • Internet access needs to be scheduled at different times to optimize bandwidth usage.
  • There are no demanding requirements on bandwidth or multiple EIPs need to be used at the same time, such as web servers and test environments.

Changes between dedicated and shared bandwidths
A dedicated bandwidth cannot be changed to a shared bandwidth and vice versa. However, you can purchase a shared bandwidth for pay-per-use EIPs.
  • Add an EIP to a shared bandwidth and then the EIP will use the shared bandwidth.
  • Remove the EIP from the shared bandwidth and then the EIP will use the dedicated bandwidth.

Bandwidth Preemption Descriptions

  • Dedicated bandwidth: Each EIP has a fixed bandwidth and is not affected by other EIPs.

    For example, both EIP-A and EIP-B are allocated 20 Mbit/s of dedicated bandwidth.

    If the bandwidth of EIP-A hits 30 Mbit/s, there will be packet loss due to the bandwidth limit, while the bandwidth of EIP-B remains idle.

  • Shared bandwidth: If the bandwidth usage of some EIPs is high, the idle bandwidth of other EIPs can be used.

    For example, two EIPs (EIP-A and EIP-B) are added to a shared bandwidth of 40 Mbit/s.

    • If EIP-A uses 30 Mbit/s and EIP-B uses 10 Mbit/s, the total bandwidth is 40 Mbit/s. EIP-A can use the idle bandwidth of EIP-B to increase its bandwidth and prevent packet loss.
    • If EIP-A uses 30 Mbit/s and EIP-B uses 15 Mbit/s, the total bandwidth reaches 45 Mbit/s and exceeds the 40 Mbit/s limit. In this case, the flexibility of the shared bandwidth fails and there will be packet loss on both EIPs.
Parent topic: Bandwidth