Recently found in the project, the amount of data generated large, the database is SQL2008R2, we found Ah, server memory footprint, CPU consumption is not high, but the whole server is very slow.
The discovery was a disk IO problem due to too many writes. Recently in the thinking, the data staged memory, timed batch storage database such an idea has not been specifically tested and verified, should theoretically reduce the consumption of IO and connection consumption.
Because my notebook is older, installed a win8.1,4g memory, running in Hyper-V virtual machine, has been prompted to run out of memory, then applied for an AWS Cloud, a great effort to succeed, but also deducted my four dollars, I do not know will not return to me.
Apply for a good try my idea, the real effect and the feasibility of the program also need to be verified again.
The main reason for using redis3.0 is that they can be clustered and can be persisted. These characteristics can not be used, it should be the time to use the upgrade better. Also want to use, accumulate points of experience, in other business more use.
Because Linux is a rookie, the installation of Redis is also not very skilled. Be afraid to forget later, then record it. (on Windows with the Linux transplant in short, a bit of a pit, the original use squid when there is a problem with the handle, and later had to change Linux)
Redis Specific installation steps:
1, wget http://download.redis.io/releases/redis-3.0.0.tar.gz
2. Tar xzvf redis-3.0.0.tar.gz
3. CD redis-3.0.0
4. Make
In the fourth step may error, should be not installed C + + compilation environment, installation environment on the line, I installed on the centos6.5, the problem is not C + + environment, installation on the line
Yum Install gcc-c++
Once the installation is complete, make again, it should still pass, execute make Distclean and then do
Then, when you are done, execute make test
In this step, it could be wrong. Tcl is not installed, install TCL
Yum Install-y tcl
Execute make test again.
Then it should be OK.
Here's a little bit of aftercare work:
1. CP redis-server/usr/local/bin/
2. CP Redis-cli/usr/local/bin
Enter SRC directory
3, mkdir/etc/redis3.0
4, mkdir/var/redis3.0
5, Mkdir/var/redis3.0/log
6, Mkdir/var/redis3.0/log
7, mkdir/var/redis3.0/6379
8. CP redis.conf/etc/redis3.0/6379.conf
modifying configuration Items
Daemonize Yes
Pidfile/var/redis3.0/run/redis_6379.pid
Logfile/var/redis3.0/log/redis_6379.log
dir/var/redis3.0/6379
Run
redis-server /etc/redis/6379.conf
至于具体的配置细节,我还没有进行具体的配置、在接下来会配置两台先试试我的想法,具体测试效果也待验证。
第一次写,算是篇日志吧,原来一直看大家写,今天也把写一下,发现写篇真的很不容易,以上的安装过程我还是参考园友(基本上是照抄)的。账号已经5年多点了,时间啊真正的快啊。
安装步骤参考:
http://www.cnblogs.com/haoxinyue/p/3620648.html
CentOS installation redis3.0 Data IO performance issues to address database frequent insertion