Shulou(Shulou.com)06/03 Report--

The meaning of Cluster, cluster, cluster is composed of multiple hosts, but it is only shown as a whole in Internet applications. With the increasing requirements of the site for hardware performance, corresponding speed, service stability, data reliability and so on, a single server is unable to solve the problem: use an expensive minicomputer. Mainframe uses ordinary servers to build service cluster enterprise cluster classification according to the target difference of the cluster, it can be divided into three types: load balancing cluster, high availability cluster, high performance computing cluster, load balancing cluster

  aims to improve the responsiveness of the application system, handle as many access requests as possible, reduce latency, and achieve the overall performance of high concurrency and high load LB).

The load distribution of   LB depends on the shunting algorithm of the primary node.

High availability cluster

  aims to improve the reliability of the application system, reduce the downtime as much as possible, ensure the continuity of services, and achieve the fault-tolerant effect of high availability (HA).

  HA works in duplex and master-slave modes

High performance computing cluster

  aims to improve the CPU computing speed of application systems, expand resources and analysis capabilities, and obtain high-performance computing (HPC) capabilities equivalent to large, supercomputers.

The high performance of   high-performance computing cluster depends on "distributed computing" and "parallel computing". It integrates the CPU, memory and other resources of multiple servers through special hardware and software to realize the computing power that only large supercomputers have.

Load balancing Cluster working Mode Analysis load balancing Cluster is currently the most widely used cluster type of cluster load scheduling technology there are three working modes: address translation IP tunnel direct routing NAT mode

TUN mode

DR mode

Load balancing cluster structure load balancing structure: layer 1, load scheduler layer 2, server pool layer 3, shared services

About LVS virtual server confirming kernel support for LVS: modprobe ip_vscat / proc/net/ip_vsLVS load scheduling algorithm polling:

  distributes access requests to nodes in the cluster (real servers) in turn, treating each server equally, regardless of the actual number of connections and system load of the server

Weighted polling:

  allocates access requests in turn according to the processing capacity of the real server, and the scheduler can automatically query the load of each node and dynamically adjust its weight.

  ensures that servers with strong processing power bear more access traffic.

Minimum connections:

  allocates requests based on the number of connections established by the real server, giving priority to the node with the least number of connections

Weighted minimum connections:

  can automatically adjust weights for real servers when the performance of server nodes varies greatly.

Nodes with higher   weight will bear a greater proportion of active connection load.

Use ipvsadm tools to create and manage LVS clusters, create virtual servers 2, add and delete server nodes 3, view clusters and nodes 4, save ARP problems in load distribution policy LVS-DR in LVS-DR load balancing cluster, load balancers and node servers have to configure the same VIP address and have the same IP address in the local area network, which is bound to cause confusion in ARP communication of each server.

  when an ARP broadcast is sent to the LVS-DR cluster, because the load balancer and the node server are connected to the same network, they both receive the ARP broadcast

At this time, only the front-end load balancer responds to  , and other node servers should not respond to ARP broadcasts.

The node server is processed so that it does not respond to the ARP request for VIP and uses the virtual interface lo:0 host VIP address to set the kernel parameter arp_ignore=1: the system only responds to the ARP request that the destination IP is the local IP. After the router receives the ARP request, the MAC address of the original VIP corresponding to the ARP entry will be updated to the MAC address of VIP corresponding to RealServer keepalived implementation principle keepalived adopts vrrp hot backup protocol Realize the multi-machine hot backup function of Linux server vrrp, virtual routing redundancy protocol, is a backup solution for routers keepalivd file explains that keepalived can realize multi-machine hot standby, each hot backup group has multiple servers, the most commonly used is that the failover of dual-machine hot backup is realized by virtual IP address drift, which is suitable for all kinds of application servers.

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.