Welcome toChina Server Net
About us | Add to Favorites | Contact us | 中文版
Position:home>Solutions >
The solution of system of server of IBM HACMP double machine
From;    Author:Stand originally

System of server of IBM HACMP double machine solution

[get together outstanding net solution] the solution of system of server of IBM HACMP double machine

Hacmp (High Availability Cluster Multi-Processing) the main function that double machine heats up backup software is to raise all of client computer science department and the dependability that its apply, is not the dependability of only station lead plane.

One, the working principle of system of Hacmp double aircraft

The working principle of HACMP is to use LAN to come the condition that monitoring lead plane and network, net blocks. There are TCP/IP network and blame TCP/IP network in environment of a HACMP. TCP/IP network applies a client to carry the public network of the visit namely, this net can be the network that most AIX place supports, be like Ethernet, t.R. , FDDI, ATM, SOCC, SLIP, etc. Blame TCP/IP network is used be opposite for HACMP HA environment (Cluster) a when medium each node undertakes monitoring and offer communication method that replaces TCP/IP, it can be to use RS232 to string together buccal line to join each node, also can be the SCSI each node gets stuck or SSA card setting becomes Target Mode kind.

1, two servers that regard double machine as the system (lead plane A and B) run Hacmp software at the same time;

2, the server is divided outside moving from the application of machine normally, regard the backup of the other side as lead plane again at the same time;

Lead plane A (moving application) : Service_ip: 172.16.1.1 Standby_ip: 172.16.2.1 Boot_ip: B of 172.16.1.3 lead plane (Standby) : Service_ip: 172.16.1.2 Standby_ip: 172.16.2.2 Boot_ip: 172.16.1.4

3, system of two lead plane (A and B) in whole moving process, through " palpitant line " the moving circumstance of the mutual the other side that monitor (the soft hardware that includes a system moves, moving circumstance of network news report and application) ;

4, once discover lead plane of the other side move abnormal (give trouble) when, the application on breakdown machine can stop to move instantly, this machine (the backup machine of breakdown machine) the application that breakdown machine can start to go up on the machine in oneself instantly, the application breakdown machine and its resource (the IP address that includes to be used and disk space) take-over come over, make the application on breakdown machine continues to move in this machine;

5, of application and resource take-over the process is finished automatically by Ha software, need not intervene artificially;

6, when two lead plane work normally, OK also according to need general among them on a machine use factitious switch to another machine (backup machine) on move.

Structural sketch map

HACMP double machine graph of construction of system
Previous12 Next