High-availability load balancing architecture (LVS + Keepalived + Nginx + Tomcat)

Scenes

business type

Software used

Large Concurrent

Features a simple, L4 forwarding

LVS+Keepalived

Large Concurrent

High functional requirements, the URL of the forwarding

A front end: the LVS + Keepalived ; rear: the Nginx + Keepalived (before L4 , the L7 load)

Concurrent little

High functional requirements, more business

Nginx+Keepalived/HAproxy+Keepalived

Note: large companies, according to the service above 3 used scenes will.

LVS three IP load balancing technology is implemented in the cluster: VS / NAT (Network Address Translation), VS / TUN (IP tunnel) and VS / DR (direct routing); typically configured DR technology. Eight kinds of scheduling algorithms (rr, wrr, lc, wlc , lblc, lblcr, dh, sh).

    Keepalived using VRRP (virtual router redundancy protocol, Virtual Router Redundancy Protocol) protocol hot backup, a software-based multi-function linux hot standby servers. It is directed to a VRRP router backup solution by a plurality of routers ----- a spare group. Through the virtual IP address shared external service provider; the same time each hot standby group, only one master server to provide services to other servers in redundant state, if the current online server fails, another server will automatically take over (priority determination to succeed order) virtual IP address, in order to continue to provide services.

    Keepalived runs on the lvs, its main function is to achieve real failure between the machine fault isolation and load balancing automatically switches to improve system availability.

Guess you like

Origin blog.51cto.com/9625010/2480480