Jul 31, 2017
TechnologyOn This Book
Borrowed from lab, written via a janpanese author.
This article will record the reading tips on Chapter 2(libvirtd related).
Network Configuration
Edit the netoworking definition xml:
$ cat internal.xml
<network>
<name>internal</name>
<bridge name='virbr8'/>
</network>
$ cat external.xml
<network>
<name>external</name>
<bridge name='virbr9'/>
</network>
Define the networking via following commands:
$ sudo virsh net-define external.xml
Network external defined from external.xml
$ sudo virsh net-autostart external
Network external marked as autostarted
$ sudo virsh net-start external
Network external started
$ libvirt sudo virsh net-list
Name State Autostart Persistent
----------------------------------------------------------
default active no yes
external active yes yes
internal active yes yes
kubernetes active yes yes
View the configuration in virt-manager:
Jul 27, 2017
TechnologyProblem
Reason
This is because hugo upgrade to a new version 0.25.1, while this new version
won’t give the default value of date in newly created markdown file.
Solution
Edit the themes/hyde-a/archetypes/default.md, add following items:
+++
title = ""
date = "{{ .Date }}"
description = ""
keywords = ["Linux"]
categories = ["Technology"]
+++
Now you could re-new your configuration, and then your blog will acts OK.
Jul 27, 2017
Technology目的
根据用户自定义配置，自动从ISO安装出整个系统。
准备材料
RHEL 6.6安装光盘, x86_64版本。
自定义kickstart文件，用于自定义分区/用户/密码/安装包等
红帽系列操作系统(用于制作光盘镜像，已验证Redhat7.3)
步骤
- 创建目录用于挂载安装光盘和自定义光盘,
其中
/media/bootiso用于挂载安装光盘，
/media/bootisoks用于存放自定义光盘内容:
$ mkdir -p /media/bootiso /media/bootisoks
- 拷贝安装内容到自定义光盘目录:
$ sudo mount -t iso9660 -o loop DVD.iso /media/bootiso
$ cp -r /media/bootiso/* /media/bootisoks/
$ chmdo -R u+w /media/bootisoks
$ cp /media/bootiso/.discinfo /media/bootisoks
$ cp /media/bootiso/.discinfo /media/bootisoks/isolinux
- 拷贝自定义的ks文件到isolinux目录下:
$ cp YourKickStartFile.ks /media/bootisoks/isolinux
- 配置引导选项:
$ vim /media/bootisoks/isolinux.cfg
initrd=initrd.img ks=cdrom:/isolinux/ks.cfg
- 创建ISO文件:
# mkisofs -r -T -V "MYISONAME" -b isolinux/isolinux.bin -c isolinux/boot.cat
-no-emul-boot -boot-load-size 4 -boot-info-table -o ../boot.iso .
经历此五个步骤以后，即可得到我们定制好的ISO，用此ISO即可安装出我们自定义好的系统.
kickstart示例文件:
安装了基本桌面、中文支持等。
#platform=x86, AMD64, or Intel EM64T
#version=DEVEL
# Firewall configuration
firewall --disabled
# Install OS instead of upgrade
install
# Use network installation
#url --url="http://10.7.7.2/CentOS"
cdrom
# Root password
rootpw --iscrypted xxxxxxxxxxxxxxxxxxxx
# System authorization information
auth --useshadow --passalgo=sha512
# Use graphical install
graphical
firstboot --disable
# System keyboard
keyboard us
# System language
lang en_US
# SELinux configuration
selinux --disabled
# Installation logging level
logging --level=info
# System timezone
timezone Asia/Hong_Kong
# System bootloader configuration
bootloader --location=mbr
# Clear the Master Boot Record
zerombr
# Partition clearing information
clearpart --all
# Disk partitioning information
part swap --fstype="swap" --size=1024
part / --asprimary --fstype="ext4" --grow --size=1
%packages
@basic-desktop
@chinese-support
@internet-browser
@x11
-ibus-table-cangjie
-ibus-table-erbi
-ibus-table-wubi
%end
其中rootpw以后的字段可以通过以下命令得到:
$ openssl passwd -1 "Your_Password_Here"
ks.cfg的另一种构建方法
在安装完的每一台机器上，都可以看到/root/ana…ks文件，编辑此文件即可得到我们定制化的kickstart配置。
Jul 21, 2017
TechnologyItems
Working items on one-click deployment of oracle database.
Ansible-Playbooks
Based on:
https://github.com/nkadbi/oracle-db-12c-vagrant-ansible
Refers to:
https://blog.dbi-services.com/vagrant-up-get-your-oracle-infrastructure-up-and-running/
https://blog.dbi-services.com/part2-vagrant-up-get-your-oracle-infrastructure-up-an-running/
Username/Password:
System: oracle/welcome1
Database: sys/oracle
Linux Client
Yaourt has the linux client for accessing oracle Db:
https://aur.archlinux.org/packages/oracle-sqldeveloper/
Installing method:
Download the file from oracle.com
Create Database
Create database using following command:
[vagrant@dbserver1 ~]$ su - oracle
Password:
-bash-4.2$ sqlplus "/as sysdba"
Now you got the shell like SQL>, you could input the sql in this shell:
Run `1_create_user_and_tablespace_dash.sql`
The first step will create the database user, then you could login into the
database using this user, using SQL Devloper for login and execute the
command:
Execute the following script:
msp_XXX.sql(Including 2 scripts)
Tips for getting the db config:
SQL> show parameter service_names;
.....
service_names string db1.private
Then your configuration should use the same service_names as described.
Jul 13, 2017
Technology测试环境
Docker常用的两种网络模式包括Bridge和Host模式，为测试这两种网络模式的性能，我们将创建以下的测试环境:
- 192.192.192.89 - 运行Docker容器的服务器， CentOS 7.3.
- 192.192.192.88 - 运行客户端的服务器, CentOS 7.3.
两台服务器之间的物理网络为万兆以太网络。
我们采用Iperfhttp://software.es.net/iperf/来测量网络带宽，iperf非常简单，也拥有足够多的特性用于测试基本的性能指标。
在服务器端，我们需要一个运行iperf3的Docker容器。 Docker的版本为17.05-ce.
测试将基于以下几个场景:
- 原始网络吞吐量
- 跨主机物理机到Docker(host模式)
- 跨主机物理机到Docker(Bridge模式)
- 同主机物理机到Docker(Bridge模式)
- 同主机Docker到Docker(Bridge模式-external)
- 同主机Docker到Docker(Bridge模式-internal)
原始网络吞吐量
首先，我们需要得到在没有任何Docker容器运行时的原始网络吞吐，在Server端运行:
[root@192.192.192.89 ~]# iperf3 -s -p 5202
Client端运行:
[root@192.192.192.88 ~]# iperf3 -c 192.192.192.89 -p 5202
运行测试后，服务器端和客户端都会返回诊断信息。我们暂时只关心其吞吐量:
-----------------------------------------------------------
Server listening on 5202
-----------------------------------------------------------
Accepted connection from 192.192.192.88, port 39682
[ 5] local 192.192.192.89 port 5202 connected to 192.192.192.88 port 39684
[ ID] Interval Transfer Bandwidth
[ 5] 0.00-1.00 sec 1.05 GBytes 9.05 Gbits/sec
[ 5] 1.00-2.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 2.00-3.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 3.00-4.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 4.00-5.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 5.00-6.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 6.00-7.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 7.00-8.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 8.00-9.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 9.00-10.00 sec 1.10 GBytes 9.41 Gbits/sec
[ 5] 10.00-10.04 sec 42.0 MBytes 9.39 Gbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bandwidth
[ 5] 0.00-10.04 sec 0.00 Bytes 0.00 bits/sec sender
[ 5] 0.00-10.04 sec 11.0 GBytes 9.38 Gbits/sec receiver
-----------------------------------------------------------
Server listening on 5202
-----------------------------------------------------------
可以看到，在万兆交换机的网络场景下，物理机到物理机之间的网络带宽跑满了万兆交换机的极限.
跨主机物理机到Docker(host模式)
在Docker中运行iperf3相当简单，在hub.docker.com可以找到大量的打包有iperf3的镜像，我们采用:
# sudo docker pull networkstatic/iperf3
在服务器端启动侦听5203端口的docker实例:
[root@192.192.192.89 ~]# docker run --net=host -it --rm --name=iperf3-server networkstatic/iperf3 -s -p 5203
在Client端执行对应的修改，得到的结果为:
[root@192.192.192.88 ~]# iperf3 -c 192.192.192.89 -p 5203
Connecting to host 192.192.192.89, port 5203
[ 4] local 192.192.192.88 port 40326 connected to 192.192.192.89 port 5203
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.00 sec 1.10 GBytes 9.43 Gbits/sec 20 625 KBytes
[ 4] 1.00-2.00 sec 1.10 GBytes 9.42 Gbits/sec 0 625 KBytes
//.....
结果差不多相同: 9.40 Gbits/sec
跨主机物理机到Docker(Bridge模式)
更改为5204端口，这次使用的网络模式为Bridge模式:
[root@192.192.192.89 ~]# docker run -it --rm -p 5204:5204 --name=iperf3-server networkstatic/iperf3 -s -p 5204
在客户端不作任何修改，只更换远端端口为5204:
[root@192.192.192.88 ~]# iperf3 -c 192.192.192.89 -p 5204
Connecting to host 192.192.192.89, port 5204
[ 4] local 192.192.192.88 port 53936 connected to 192.192.192.89 port 5204
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.00 sec 1.10 GBytes 9.44 Gbits/sec 15 669 KBytes
[ 4] 1.00-2.00 sec 1.10 GBytes 9.42 Gbits/sec 0 682 KBytes
[ 4] 2.00-3.00 sec 1.10 GBytes 9.42 Gbits/sec 0 691 KBytes
可以看到，在Bridge模式下，吞吐量也跑满了万兆网络的极限.
同主机物理机到Docker(Bridge模式)
在同一台主机上(192.192.192.89)上运行iperf，测试到Docker的吞吐量，沿用之前侦听5204的容器不变:
[root@192.192.192.89 ~]# iperf3 -c 192.192.192.89 -p 5204
Connecting to host 192.192.192.89, port 5204
[ 4] local 192.192.192.89 port 46720 connected to 192.192.192.89 port 5204
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.00 sec 2.77 GBytes 23.8 Gbits/sec 0 274 KBytes
[ 4] 1.00-2.00 sec 2.75 GBytes 23.6 Gbits/sec 0 274 KBytes
[ 4] 2.00-3.00 sec 2.75 GBytes 23.6 Gbits/sec 0 277 KBytes
在这种模式下，网络的吞吐量几乎三倍于万兆网络，这是因为从主机到Docker实例的网络通路会走本地的回环接口(lo-loopback)接口。
同主机Docker到Docker(Bridge模式-external)
沿用侦听5204端口的容器不变，新启动一个容器，在其中运行iperf:
# iperf3 -c 192.192.192.89 -p 5204
Connecting to host 192.192.192.89, port 5204
[ 4] local 172.17.0.5 port 59574 connected to 192.192.192.89 port 5204
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.00 sec 1.03 GBytes 8.84 Gbits/sec 91 228 KBytes
[ 4] 1.00-2.00 sec 955 MBytes 8.01 Gbits/sec 0 229 KBytes
[ 4] 2.00-3.00 sec 1.02 GBytes 8.80 Gbits/sec 0 230 KBytes
[ 4] 3.00-4.00 sec 767 MBytes 6.43 Gbits/sec 0 230 KBytes
[ 4] 4.00-5.00 sec 851 MBytes 7.14 Gbits/sec 0 230 KBytes
可以看到，如果直接使用物理机的IP地址和端口，则吞吐需要同时使用Bridge模式下物理网卡的吞吐，
此时网卡的物理性能下降明显。
同主机Docker到Docker(Bridge模式-internal)
为了避免使用物理机的IP地址带来的性能下降，直接使用容器内部的IP地址做iperf测试:
# iperf3 -c 172.17.0.4 -p 5204
Accepted connection from 172.17.0.5, port 39516
[ 5] local 172.17.0.4 port 5204 connected to 172.17.0.5 port 39518
[ ID] Interval Transfer Bandwidth
[ 5] 0.00-1.00 sec 2.39 GBytes 20.5 Gbits/sec
[ 5] 1.00-2.00 sec 2.50 GBytes 21.5 Gbits/sec
[ 5] 2.00-3.00 sec 2.50 GBytes 21.5 Gbits/sec
可以看到，在这种模式下，容器之间的通信还是基于lo(loopback)接口来做的，几乎三倍于万兆交换机的峰值速度。
结论
各次测试的对比数据整理如下:
| 物理机-物理机 | 9.40 Gbit/sec | 100% |
| 跨物理机到Docker(host模式网络) | 9.40 Gbit/sec | 100% |
| 跨物理机到Docker(Bridge模式网络) | 9.40 Gbit/sec | 100% |
| 同主机内到Docker(Bridge模式网络) | 23.8 Gbit/sec | 250% |
| 同主机Docker到Docker(Bridge模式-ex) | 8.00 Gbit/sec | 85% |
| 同主机Docker到Docker(Bridge模式-int) | 21.00 Gbit/sec | 220% |
结论： 在Docker运行环境中，网络的吞吐量近似于本地网络IO，基本上不会有性能损耗。需要特别注意的是，一定要避免同主机中的Docker实例彼此使用物理机IP/端口进行通信，那样会带来性能的明显下降。
