Table of Contents
Advertisement
S9700 Core Routing Switch
Configuration Guide - SPU
l
l
l
Issue 01 (2012-03-15)
In the least connection algorithm, the SPU makes load balancing decisions according
to the priority, weight, and number of active connections of a load balancing member.
The SPU selects load balancing members with higher priorities for providing services,
and often selects the load balancing member with the smallest number of active
connections or the smallest weight. After a load balancing member is selected, the
SPU determines whether the member is available according to the bandwidth limit,
connection quantity limit, and connection rate limit. If no load balancing member with
higher priority can be used, the SPU selects a load balancing member among those with
lower priorities according to the weight and least connection algorithm.
The least connection algorithm seamlessly distributes connection requests with different
durations to all servers or links. This algorithm is applicable when server performance
in a server group or link bandwidth in a link group varies, and users initiate connections
with different durations.
– Hash algorithm based on the IP address
In the IP address-based hash algorithm, the SPU hashes the source IP address,
destination IP address, or source and destination IP addresses and makes load balancing
decisions according to the hash value. After a load balancing member is selected, the
SPU determines whether the member is available according to the bandwidth limit,
connection quantity limit, and connection rate limit. If the load balancing member is
unavailable, the SPU selects the next load balancing member according to the hash
value.
The hash algorithm can map the following requests to the same server or link:
– Requests with the same source IP address
– Requests with the same destination IP address
– Requests with the same source and destination IP addresses
– Requests whose source IP addresses are located in the same network segment
– Requests whose destination IP addresses are located in the same network segment
– Requests whose source and destination IP addresses are located in the same network
segment
The hash algorithm is applied when requests from a user are distributed to the same
server or link, and is also applied to server load balancing. It is applied to the scenario
where all requests from a user are distributed to a server or a link. It is also applied to
firewall load balancing.
– Hash algorithm based on the HTTP URL
In the hash algorithm based on the HTTP URL, the SPU hashes the URL carried in
HTTP request packets and makes load balancing decisions according to the hash value.
After a load balancing member is selected, the SPU determines whether the member is
available according to the bandwidth limit, connection quantity limit, and connection
rate limit. If the load balancing member cannot be used, the SPU selects the next load
balancing member according to the hash value.
Health detection
Load balancing device periodically detects service status of real servers or links to collect
required information and isolate faulty servers or links. The SPU can detect whether servers
or links run normally.
Session stickiness
Connection requests of a user in a period are sent to the same server for processing.
Firewall load balancing
Huawei Proprietary and Confidential
Copyright © Huawei Technologies Co., Ltd.
6 Load Balancing Configuration
183
Advertisement
Table of Contents
