In addition to Weibo, there is also WeChat
Please pay attention
WeChat public account
Shulou
2025-01-16 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Servers >
Share
Shulou(Shulou.com)06/03 Report--
The following brings you what is the essence of load balancing technology. I hope it can give you some help in practical application. There are many things involved in load balancing, not many theories, and there are many books on the Internet. Today, we will use the accumulated experience in the industry to make an answer.
What is load balancing, the essence of load balancing?
Components required to achieve load balancing:
Loaded traffic or request
Entrances that accept traffic or requests
Loaded object
Mode of load
Conditions of load
Full Agent Architecture (full-proxy) F5
When visitors request a cloud server, they will first request the load balancer (set up three times to cover their hands, F5 will detect the requested content, and then F5 will request the server (three-way handshake)
Non-full agent architecture model
LVS: when the visitor sends the request packet, the load balancer will use the algorithm to communicate the request to the server (the load balancing function is to reverse proxy for NAT forwarding)
Access rules for VIP or VS (general network technology)
Destination address:
10.10.1.100:443
10.10.1.100Ru *
10.10.1.It is not easy to see 443
10.10.1. Mistleau *
*: 443
*: *
Note: to distinguish the relationship between VIP and its own IP on load balancers
VIP is used to receive requests. There can be more than one load balancer, or even thousands of ip itself is used to communicate with other devices, and several should be provided according to the actual network conditions.
Finally, for the sake of security, a fully matched type of VIP should set the traffic direction, otherwise all traffic will be processed, resulting in bandwidth consumption and performance waste.
Object of load:
Local load
In the local load balancer, the object of the load is the application server, and the same kind of application server that is loaded is called the load balancing pool (group). The members of the group, that is, each server (node), can dynamically adjust their status through policies to achieve different degrees of request allocation.
The main technologies involved are high and low priority groups, disable smooth downline, connection limit, health check of specific nodes, algorithms and so on.
Global load
In the global load balancer, the object of the load is the DNS request, which is mainly used to resolve the DNS request to the ip address (public network or private network) of different data centers or different lines (operators) of a data center. Complete DNS system related technology.
The main technologies involved are address library, BIND, performance optimization, recursion, iteration and forwarding.
Link load
In the link load balancer, the object of the load is the link (that is, the gateway). This technology combines the local load and the global load to derive a link load pattern, which functionally combines the first two, but can not replace the former two.
The main technologies involved are four layers of load, partial DNS parsing, and NAT related technologies.
Mode of load (algorithm)
Static state
Polling
Ratio
Dynamic
Minimum number of connections
The fastest response speed
Ip hash
Dynamic ratio
Special
High and low priority group
Failure algorithm
Session persistence:
Session persistence is a special strategy, which aims to load the "same" request on the same object under certain conditions, so as to achieve the consistency of the session before and after. Such as: shopping checkout
Species:
Source address session persistence
Destination address session persistence
Cookie session persistence
Special session persistence
Tip: health examination technology can be simply understood as a detection mechanism, which is essentially a technology in which someone checks an object in a certain way and at a certain time interval, and judges the state of the object according to the expected response value.
There are five steps in a health check:
Initiated sourc
Check object
Inspection mode
Check time
Expect a response
After reading the above about the essence of load balancing technology, if there is anything else you need to know, you can find out what you are interested in in the industry information or seek answers from our professional and technical engineers. Technical engineers have more than ten years of experience in the industry.
Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.
Views: 0
*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.
Continue with the installation of the previous hadoop.First, install zookooper1. Decompress zookoope
"Every 5-10 years, there's a rare product, a really special, very unusual product that's the most un
© 2024 shulou.com SLNews company. All rights reserved.