XenServer Active/Active Load Balancing

In this blog post, we will delve into the XenServer Active/Active bonding concept and how it facilitates load balancing. We will discuss its working principle and potential issues that may arise and compare it with an alternative approach called Link Aggregation Control Protocol (LACP).


XenServer Active/Active Bonding Overview:
XenServer Active/Active bonding is a mechanism that enables load balancing between two network interface cards (NICs) in XenServer. The primary objective is distributing network traffic evenly across both NICs, ensuring efficient utilization and preventing congestion. Let’s explore how this works in practice.

1. Traffic Pinning:
To initiate load balancing, the system pins traffic to one of the NICs. The selection is based on the lowest bandwidth available on the interface. This approach aims to maintain an even distribution of network load.

2. Re-balancing:
Periodically, usually every 10 seconds (or 30 minutes in newer versions like v6.1), the system verifies if the load on the two interfaces remains balanced. If an imbalance is detected, the system triggers a re-balancing process. This process involves shifting the entire traffic load to the other NIC.

3. Gratuitous ARP (GARP) and ARP Table Changes:
During re-balancing, the bond interface sends a GARP to update the Address Resolution Protocol (ARP) table on the connected Cisco switches. This update ensures the switches correctly associate the NIC’s MAC address with the appropriate IP address. However, issues can arise if the GARP is not received on the Cisco side or the Cisco switches fail to respond to the ARP Response broadcast.

4. Traffic Shifting:
Once the ARP table changes are successfully propagated, all traffic is shifted and pinned to the other NIC, providing a balanced network load distribution. This process can repeat as needed to maintain load equilibrium.


Limitations and Disadvantages of Active/Active Bonding:
Despite its benefits, XenServer Active/Active bonding has a few drawbacks worth considering:

1. Aggregation Limitation:
Active/Active bonding does not offer aggregation capabilities, limiting the bandwidth to 1G.

2. Potential Re-Balancing Issues:
The frequent re-balancing and associated GARP updates can introduce communication disruptions due to ARP changes on the Cisco side. This instability may not be suitable for larger environments and should be cautiously approached, especially when future scalability is a concern.


The Preferred Alternative: LACP
Link Aggregation Control Protocol (LACP) is the recommended choice for scenarios requiring more advanced capabilities, such as increased bandwidth and stability. LACP allows for aggregating both ports, providing a single logical port with enhanced bandwidth (e.g., 2G). It eliminates the need for frequent GARP updates and re-balancing, offering a more stable and scalable solution.


Considerations for Switch Configuration:
To implement Active/Active bonding effectively, it is important to note that the switch side must be configured to support the necessary functionality. For instance, in the case of Cisco switches, utilizing VSS/VPC or switch stacking is required to accommodate the MAC shifting between ports and maintain a cohesive understanding of the changes.



Let’s now explore the significance of BOND_STATUS_CHANGED alerts in XenServer and delve into the details of bond status changes. By deciphering these alerts, we can gain valuable insights into the network connectivity and load-balancing aspects of bonded interfaces. Additionally, we will discuss the importance of monitoring and responding to these alerts within the XenServer environment.

Effective monitoring of XenServer environments involves understanding and interpreting various alerts. One such alert is the BOND_STATUS_CHANGED alert, which provides essential information about the status of bonded network interfaces. Let’s dive deeper into these alerts and their implications.

Alert Details:
Priority: 3
Class: Host
Object UUID: a1b2c3d4-5678-90ab-cdef-1234567890ab
Timestamp: 20230622T02:27:30Z
Message UUID: 98765432-10fe-dcba-fedc-ba0987654321
Pool name: hp7k-p1 (xen-network-p1)
Body: The status of the eth0+eth2 bond changed: 2/2 up (was 1/2)

Understanding the Alert:
The BOND_STATUS_CHANGED alert indicates a change in the status of a specific bond interface, eth0+eth2, within the XenServer environment. It provides crucial information about the bond’s current status and a comparison with its previous state. The alert states that the bond status has changed from “1/2 up” to “2/2 up.”

Interpreting the Alert:
The alert’s message signifies a shift in the network connectivity of the bonded NICs. Initially, the bond had only one of the two NICs (eth0 or eth2) in an active or available state, resulting in a status of “b.” However, the recent change indicates that both eth0 and eth2 are now up and functioning correctly, leading to a status of “2/2 up.”

Importance of Monitoring Bond Status Changes:
Monitoring bond status changes is essential for the following reasons:

1. Network Health and Connectivity: Bond status changes provide valuable insights into the health and connectivity of bonded interfaces. By closely monitoring these changes, administrators can promptly identify disruptions or irregularities in network connectivity.

2. Load Balancing and Performance Optimization: Bond status changes may indicate load balancing adjustments within the XenServer environment. A change in bond status suggests that traffic is being distributed more evenly across the bonded NICs, optimizing performance and preventing congestion.

3. Troubleshooting Network Issues: Bond status changes can indicate potential network issues. Frequent or unexpected status changes warrant further investigation to identify and address underlying problems, such as faulty NICs, misconfigured bonding settings, or network connectivity disruptions.


BOND_STATUS_CHANGED alerts in XenServer provide crucial information about bond status changes, enabling administrators to monitor network connectivity, load balancing, and troubleshoot potential issues. Understanding the significance of these alerts is vital to the proactive management of XenServer environments, ensuring optimal network performance and stability.

To ensure the smooth operation of your XenServer environment, it is recommended to regularly monitor and investigate bond status changes, taking appropriate action when necessary. You can provide a reliable and efficient XenServer deployment by staying vigilant, responding promptly, and maintaining a robust network infrastructure.


XenServer Active/Active bonding offers a load balancing solution that can be suitable for certain environments, particularly smaller setups with less demanding traffic. However, for enterprises or larger-scale deployments, LACP provides a more robust and scalable approach. Understanding the specific requirements and limitations of each method is crucial when designing network configurations.

To explore the technical details further, refer to the official Citrix documentation on Active/Active bonding: [Link to CTX132559](https://support.citrix.com/article/CTX132559