ceph storage cluster

I have previously studied how to build a mongodb environment (see). Considering that the mongodb cluster environment will be used in the future and the multi-bit 32-bit OS environment, I have constructed a mongodb cluster environment. For details about the process, see "sharding with replica set. In this experiment, four XP machines are built with VMWare 8.0, with IP addresses 192.168.111.11, 192.168.111.12

We have studied how to build a mongodb environment. Considering that we need to use the mongodb Cluster Environment and the multi-bit 32-bit OS environment in the future, we have built a mongodb cluster environment. For more information, see We have studied how to build a mongodb environment. Considering that we need to use the mongodb Cluster Environment and the

LVS-DR direct connection routing mode + NAS storage sharing 650) this. width = 650; "src =" http://www.bkjia.com/uploads/allimg/131228/053FV109-0.jpg "title =" 54.jpg" alt = "161926991.jpg"/> Tutorial steps: LVS Scheduler configuration: 1) Modify interface eth0: 0 2) Modify kernel sysctl. conf 3) install ipvsadm and Configuration Web service configuration: 1) modify the loopback interface lo: 0 2) modify the kernel 3) Install WEB I,LVS-DRDirect Connec

to adjust. starting with MySQL 5.6.3 , This parameter is no longer needed. (Error log [Warning] option ' innodb_additional_mem_pool_size ': Signed value 512000 adjusted to 524288). */In the example , the default configuration is the default configuration, which is 8M. in the Mysql manual, the recommended setting for 2G memory is 20M. In general, set up the General Assembly caused by the waste of physical memory, can be on the line for a period of time to pay close attention to the log, parame

=" float: none; "title=" 1.png "alt=" Wkiom1edeeqjzglsaaddtizr8tc704.png "/>650) this.width=650; "src=" Http://s4.51cto.com/wyfs02/M01/83/FD/wKioL1eDEEuy7_0YAAEUcSL0Fiw464.png "style=" float: none; "title=" 2.png "alt=" Wkiol1edeeuy7_0yaaeucsl0fiw464.png "/>650) this.width=650; "src=" Http://s4.51cto.com/wyfs02/M01/83/FD/wKioL1eDEEuBdlgRAAEyZTKk628570.png "style=" float: none; "title=" 3.png "alt=" Wkiol1edeeubdlgraaeyztkk628570.png "/>650) this.width=650; "src=" Http://s3.51cto.com/wyfs02/M01/8

IO path is the data channel of the storage system, the inlet and outlet should have controllable valve mechanism, sometimes the entrance is also the outlet, especially in the distributed cluster the problem of the midbrain crack needs its support. Related points in the kernel 1. The freeze function of the file system VFS layer; 2.DM especially MP This block has the queue suspend function; This block

to the DT era, our software architecture is also from standalone to cluster and distributed. However, the first step toward the distribution needs to solve the problem of multiple servers sharing login information. It is to solve the problem that multiple servers share common information. To solve this problem, we can store the public information in the state server, or store it in a database, and then store it in our "distributed cache" and so on. B

(first100elements):28384.90requests persecondlrange_300 (first300elements):9022.01requestsper secondlrange_500 (first450elements): 4810.00requestspersecondlrange_ 600 (first600elements): 3363.04requestspersecond B. through to the back-end redis view :[Email protected] nutcracker]# redis-cli 127.0.0.1:6379> KEYS) "counter:__rand_int__" 127.0.0.1:6379> [[email Protected] nutcracker]# redis-cli 127.0.0.1:6379> KEYS *) "MyList" 2) "key:__rand_int__"found our test data fragmented to two Redis insta

LVS Load Balancing cluster + NAT Mode NAT Mode 650) this. width = 650; "src =" http://www.bkjia.com/uploads/allimg/131228/05364K554-0.jpg "title =" 9.jpg" alt = "162659736.jpg"/> I,BuildLVSServer Load balancer Cluster(Address translation modeLVS-NAT) 1. Configure the Network Environment Web1: Enable route forwarding. Configure Gateway: route add default gw gateway 192.168.1.1 Web2: Enable route forwarding

IP Planning and setting node name IP Configuration Table Node name P Ublicip Vip P rivate IP Scanip 18cdbrac1 128.0.5.101 128.0.5.103 10.10.10.1 128.0.5.105 10.10.10.2 128.0.5.106 128.0.5.107 18cdbrac2 128.0.5.102 128.0.5.104 10.10.10.5 10.10.10.6 Dns 128.0.5.253

/etc/rc.local/application/bin/apachectl start/application/sersync/sersync2-d-r-o/application/sersync/confxml.xml2. Add a script under the/etc/init.d/ directoryThe Apachectl file in the Apache/bin directory is actually a startup script, copy the Apachectl file to the/etc/init.d/directory, and rename it to Apache (the name is easy to remember.)# Cp/application/apache/bin/apachectl/etc/init.d/apache2Then modify the Apache file to allow it to support service and Chkconfig commands:# Vim/etc/init.d/

Advantage Description :1. memcached is a memory cache that can read and write faster than regular files2. Can solve the problem of multiple server common sessionDrawback Description :1. Session data is stored in memory, persistence is deficient, but it is not a problem for session data2. You can also use other persistent systems to store sessions, for example: Redis,ttserver3. High-performance concurrency scenarios, the efficiency of cookies is much better than the session, so many large website