之前寫過一個Guide for Cloud Foundry New Teamer。不過似乎已經有些過時，那會實驗室主要是針對的CF v1進行的研究，現在已經全面進入V2時代了。所以更新一下關於CloudFoundry運維的一些內容。如果有時間也可再回頭看看V1的那個帖子，希望能有所幫助。
部署
關於部署，目前使用的工具一般有兩種，BOSH和cf_nise_installer。BOSH適用於集群安裝，cf_nise_installer適用於單節點安裝。下面主要以cf_nise_installer的安裝方法為主描述部署的運維流程。
cf_nise_installer實際上就是一大堆shell腳本建立起來的項目，對shell腳本熟悉的人打開上面的鏈接就可以看到整個部署的流程。
第一步就是安裝運行環境。
通過cf_nise_installer中的這段install.sh腳本的代碼我們可以看到它的安裝流程如下：
./scripts/install_ruby.sh
source ~/.profile
./scripts/clone_nise_bosh.sh
./scripts/clone_cf_release.sh
./scripts/install_environemnt.sh
./scripts/install_cf_release.sh
- 安裝ruby：cf_nise_installer使用的是rbenv這個ruby安裝工具。同樣比較有名的ruby安裝工具還有rvm，這兩個工具任選一個即可，如果是使用某個固定的ruby版本的話，建議使用源碼安裝。
- 下載nise_bosh項目：實際上cf_nise_installer是基於nise_bosh的一個腳本，真正執行安裝的就是nise_bosh這個項目。nise_bosh是一個基於bosh的項目，把bosh關於IaaS層的內容去除，保留了虛擬機上組建安裝的內容。所以使用nisebosh無需IaaS層的API支持，只需要虛擬機即可安裝。
- 制作cf_release：cf_release是cf源碼經過編譯后的內容。從github上clone下來的cf_release，checkout到指定版本git使用簡易入門，然后執行
git submodule update --init --recursive把子模塊submodule下載下來，再執行bosh create release命令，就可以得到一個完整的cf_release，當然，這里面又是漫長的下載。實驗室已經制作了幾個常用的cf_release版本，在內網可以下載。 - 安裝環境，這里執行的其實就是nise_bosh/bin/init這個腳本，打開就可以看到下載了很多基礎的運行時環境，以及監控使用的monit工具。
- 安裝cf組件。到這里就是真正的安裝cf組件了。安裝的命令很短，東西都在配置文件里面了。可以打開manifests/template查看一下。默認的域名都設置成了
192.168.10.10.xip.io，執行 generate_deploy_manifest.sh腳本可以更改域名和密碼，不過需要在環境變量中導入export NISE_DOMAIN=[你的域名]export NISE_PASSWORD=[你的密碼]
關於域名 架設本地域名解析服務器，如bind9之類的，然后在/etc/resolv.conf下面的nameserver加上本地域名解析服務器IP即可。在域名解析服務器上加上一條域名對應組件機器的IP就可以順利用域名訪問集群了。
關於一次正常的連接 [用戶訪問域名]->[域名解析服務器解析出IP]->[Haproxy組件收到請求轉發]->[Gorouter接到請求]->[DEA/CC最終處理]
集群使用
安裝CF命令行工具
下載並安裝cf_clidpkg -i ***.deb。新版cf命令行的命令有比較大的變化，可以使用cf --help看一下。
制作離線java_buildpack
cloudfoundry從V168以后，就不在cf-release里面放入buildpack了，也就是雲應用的運行時環境。每次應用上傳都要去pivotal的網站上下buildpack，速度比較慢，所以就涉及到了制作離線的buildpack。官方的java_buildpack就提供了制作離線包的功能。Clone下來以后執行如下步驟就制作成功了一個zip包。
bundle install
bundle exec rake package OFFLINE=true
...
Creating build/java-buildpack-offline-cfd6b17.zip
然后使用cf命令上傳離線的buildpack，參數的意思可以從help中查看。
root@cf-one:~/java-buildpack# cf create-buildpack test_pack ./build/java-buildpack-offline-abe37f7.zip 0
Creating buildpack test_pack...
OK
Uploading buildpack test_pack...
OK
root@cf-one:~/java-buildpack# cf buildpacks
Getting buildpacks...
buildpack position enabled locked filename
java_buildpack 2 true false java-buildpack-v2.1.2.zip
ruby_buildpack 3 true false buildpack_ruby_v46-245-g2fc4ad8.zip
nodejs_buildpack 4 true false buildpack_nodejs_v8-177-g2b0a5cf.zip
test_pack 1 true false java-buildpack-offline-abe37f7.zip
使用CF-CLI命令行上傳應用
如果用戶名密碼都是默認的話，使用流程基本如下：
target
root@cf-one:~/yd/cf_nise_installer# cf api --skip-ssl-validation api.test4.sel
Setting api endpoint to api.test4.sel...
OK
API endpoint: https://api.test4.sel (API version: 2.2.0)
Not logged in. Use 'cf login' to log in.
login
root@cf-one:~/yd/cf_nise_installer# cf login
API endpoint: https://api.test4.sel
Email> admin
Password>
Authenticating...
OK
Targeted org DevBox
Select a space (or press enter to skip):
Space>
API endpoint: https://api.test4.sel (API version: 2.2.0)
User: admin
Org: DevBox
Space: No space targeted, use 'cf target -s SPACE'
org
root@cf-one:~/yd/cf_nise_installer# cf create-org sun
Creating org sun as admin...
OK
TIP: Use 'cf target -o sun' to target new org
root@cf-one:~/yd/cf_nise_installer# cf target -o sun
API endpoint: https://api.test4.sel (API version: 2.2.0)
User: admin
Org: sun
Space: No space targeted, use 'cf target -s SPACE'
space
root@cf-one:~/yd/cf_nise_installer# cf create-space test
Creating space test in org sun as admin...
OK
Assigning role SpaceManager to user admin in org sun / space test as admin...
OK
Assigning role SpaceDeveloper to user admin in org sun / space test as admin...
OK
TIP: Use 'cf target -o sun -s test' to target new space
root@cf-one:~/yd/cf_nise_installer# cf create-space test
Creating space test in org sun as admin...
OK
Assigning role SpaceManager to user admin in org sun / space test as admin...
OK
Assigning role SpaceDeveloper to user admin in org sun / space test as admin...
OK
TIP: Use 'cf target -o sun -s test' to target new space
root@cf-one:~/yd/cf_nise_installer# cf target -o sun -s test
API endpoint: https://api.test4.sel (API version: 2.2.0)
User: admin
Org: sun
Space: test
Push APP
root@cf-one:~# cf push test -p helloworldWeb.war -b test_pack
Creating app test in org sun / space test as admin...
OK
Creating route test.test4.sel...
OK
Binding test.test4.sel to test...
OK
Uploading test...
Uploading app files from: helloworldWeb.war
Uploading 2.6K, 10 files
OK
Starting app test in org sun / space test as admin...
OK
-----> Downloaded app package (4.0K)
-----> Java Buildpack Version: abe37f7 (offline) | https://github.com/cloudfoundry/java-buildpack.git#abe37f7
-----> Downloading Open Jdk JRE 1.7.0_60 from http://download.run.pivotal.io/openjdk/lucid/x86_64/openjdk-1.7.0_60.tar.gz (found in cache)
Expanding Open Jdk JRE to .java-buildpack/open_jdk_jre (2.1s)
-----> Downloading Tomcat Instance 7.0.54 from http://download.run.pivotal.io/tomcat/tomcat-7.0.54.tar.gz (found in cache)
Expanding Tomcat to .java-buildpack/tomcat (0.1s)
-----> Downloading Tomcat Lifecycle Support 2.1.0_RELEASE from http://download.run.pivotal.io/tomcat-lifecycle-support/tomcat-lifecycle-support-2.1.0_RELEASE.jar (found in cache)
-----> Downloading Tomcat Logging Support 2.1.0_RELEASE from http://download.run.pivotal.io/tomcat-logging-support/tomcat-logging-support-2.1.0_RELEASE.jar (found in cache)
-----> Uploading droplet (38M)
0 of 1 instances running, 1 starting
1 of 1 instances running
App started
Showing health and status for app test in org sun / space test as admin...
OK
requested state: started
instances: 1/1
usage: 1G x 1 instances
urls: test.test4.sel
state since cpu memory disk
#0 running 2014-06-19 08:10:08 AM 0.0% 170.5M of 1G 96M of 1G
root@cf-one:~# curl test.test4.sel
<html>
<body>
<h2>Hello World!</h2>
</body>
</html>
至此，就成功上傳了一個應用了。
查看應用狀態
root@cf-one:~# cf apps
Getting apps in org sun / space test as admin...
OK
name requested state instances memory disk urls
test started 1/1 1G 1G test.test4.sel
重啟一個應用
oot@cf-one:~# cf restart test
Stopping app test in org sun / space test as admin...
OK
Starting app test in org sun / space test as admin...
OK
0 of 1 instances running, 1 starting
0 of 1 instances running, 1 starting
1 of 1 instances running
App started
Showing health and status for app test in org sun / space test as admin...
OK
requested state: started
instances: 1/1
usage: 1G x 1 instances
urls: test.test4.sel
state since cpu memory disk
#0 running 2014-06-19 08:19:38 AM 0.0% 170M of 1G 96M of 1G
集群維護
目前集群的維護都是使用的monit工具進行簡單的組件監控。更多的維護還需要查看組件的日志。組件日志一般在/var/vcap/sys/log中
查看集群組件狀態
root@cf-one:~# monit summary
The program 'monit' is currently not installed. You can install it by typing:
apt-get install monit
You will have to enable the component called 'universe'
root@cf-one:~# /var/vcap/bosh/bin/monit summary
The Monit daemon 5.2.4 uptime: 44m
Process 'nats' running
Process 'nats_stream_forwarder' running
Process 'cloud_controller_ng' running
Process 'cloud_controller_worker_local_1' running
Process 'cloud_controller_worker_local_2' running
Process 'nginx_ccng' running
Process 'cloud_controller_worker_1' running
Process 'cloud_controller_clock' running
Process 'uaa' running
Process 'uaa_cf-registrar' running
Process 'haproxy' running
Process 'gorouter' running
Process 'warden' running
Process 'dea_next' running
Process 'dir_server' running
Process 'dea_logging_agent' running
Process 'loggregator' running
Process 'loggregator_trafficcontroller' running
Process 'etcd' running
Process 'hm9000_listener' running
Process 'hm9000_fetcher' running
Process 'hm9000_analyzer' running
Process 'hm9000_sender' running
Process 'hm9000_metrics_server' running
Process 'hm9000_api_server' running
Process 'hm9000_evacuator' running
Process 'hm9000_shredder' running
Process 'postgres' running
System 'system_cf-one' running
重啟某個組件
root@cf-one:~# monit restart postgres
重啟所有組件
root@cf-one:~# monit restart all