cf push failed :execution expired (Timeout::Error)

[Ward Lee]

Hi, I'm try to deploy a war application into API server in local cf(10.244.0.34). Before I can deploy it. Now I modify director's IP from 192.168.50.4 to another IP as public network, from bosh-lite or bosh  VMs, they can ping all outside IP including d2vm4m9hl67ira.cloudfront.net, but the downloading failed. What's wrong with it?Any help appreciated!

icity@ubuntu-host-131:~/download$ cf push hw -p helloworld.war

Updating app hw in org org1 / space space1 as admin...

OK

Uploading hw...

Uploading from: /home/icity/download/helloworld.war

165.6K, 17 files

OK

Stopping app hw in org org1 / space space1 as admin...

OK

Starting app hw in org org1 / space space1 as admin...

OK

-----> Downloaded app package (5.8M)

-----> Java Buildpack source: system

-----> Downloading Open JDK 1.7.0_51 from http://d2vm4m9hl67ira.cloudfront.net/openjdk/lucid/x86_64/openjdk-1.7.0_51.tar.gz 

/var/vcap/packages/dea_next/buildpacks/lib/installer.rb:18:in `system': execution expired (Timeout::Error)

        from /var/vcap/packages/dea_next/buildpacks/lib/installer.rb:18:in `compile'

        from /var/vcap/packages/dea_next/buildpacks/lib/buildpack.rb:73:in `block in compile_with_timeout'

        from /var/vcap/packages/dea_next/buildpacks/lib/buildpack.rb:72:in `compile_with_timeout'

        from /var/vcap/packages/dea_next/buildpacks/lib/buildpack.rb:54:in `block in stage_application'

        from /var/vcap/packages/dea_next/buildpacks/lib/buildpack.rb:50:in `chdir'

        from /var/vcap/packages/dea_next/buildpacks/lib/buildpack.rb:50:in `stage_application'

        from /var/vcap/packages/dea_next/buildpacks/bin/run:10:in `<main>'

FAILED

Start app timeout

TIP: use 'cf logs hw --recent' for more information

[Stephan Oudmaijer]

Hi,

Try ssh-ing into a dea and wget the tar file. I think the download is either REALLY REALLY slow on your side or you have network/firewall issues and you cannot download the JDK package.

I dont really know if the healthCheckTimeout is aborting your download. You might want to try increasing it.

[Ward Lee]

Hi,Stephan.Thanks a lot! 

I tried to ssh to one DEA, there is no job named DEA so I ssh to a job named runner_z1. I can use wget to download the pacakge file. But it's very slow because of the poor network. So I think it's the Timeout parameter aborting . How to check and set HealthCheckTimeout?

[Stephan Oudmaijer]

Hi,

when using the go client you can push with minus t. 

cf push -t 180

With maven you can add the <healthCheckTimeout> element to the configuration.

HTH.

[Ward Lee]

Thanks ,Stephan.  

The command is enough for resolving this issue. could you give me some tutorials link for how to use maven to set this parameter ? Appreciated for it!

在 2014年3月11日星期二UTC+8下午3时14分38秒，Stephan Oudmaijer写道：

[Ward Lee]

Thanks ,Stephan.  

The command is enough for resolving this issue. could you give me some tutorials link for how to use maven to set this parameter ? Appreciated for it!

在 2014年3月11日星期二UTC+8下午3时14分38秒，Stephan Oudmaijer写道：

Hi,

[Ward Lee]

Hi, Stephan, I tried to use -t ,but it failed:

$ cf push hw -p helloworld.war -t 3600

Updating app hw in org org1 / space space1 as admin...

FAILED

Server error, status code: 400, error code: 100001, message: The app is invalid: health_check_timeout maximum_exceeded

So maybe I have to modify this parameter. Any advise?

[Stephan Oudmaijer]

Hi,

in the ccng job there is a setting for the maximum health check timeout:

 ccng.maximum_health_check_timeout:
    default: 180
    description: "Maximum health check timeout that can be set for the app"

You can try to increase it.

HTH

[Stephan Oudmaijer]

There is also another setting, I think you will have to increase this one:

  ccng.max_staging_runtime:
    default: 900
    description: "Timeout for staging a droplet"

Downloading the jdk etc is still in the staging phase. I think this setting is actually causing your deployment to fail. The health_check_timeout is probably the timeout for the application startup. But your app does not even finish staging yet.

[Ward Lee]

Thanks very much, Stephan. I found an useful plugin for cloud foundry :Cloud Foundry Maven Plugin

https://github.com/cloudfoundry/cf-java-client/blob/master/cloudfoundry-maven-plugin/README.md

In this plugin ,it can change the health_check_timeout, but there is no maximum_health_check_timeout and max_staging_runtime.Could you tell me how to change these two parameters?

<build> <plugins> <plugin> <groupId>org.cloudfoundry</groupId> <artifactId>maven-cf-plugin</artifactId> <version>1.0.2</version> <configuration> <server>10.244.0.34</server> <target>https://api.10.244.0.34.xip.io</target> <org>org1</org> <space>space1</space> <appname>hw</appname> <url>hw.local.cfapps.io</url> <memory>128</memory> <diskQuota>1024</diskQuota> <healthCheckTimeout>3600</healthCheckTimeout> <env> <ENV-VAR-NAME>env-var-value</ENV-VAR-NAME> </env> <services> <service> <name>postgres-test</name> <label>elephantsql</label> <plan>turtle</plan> </service> <service> <name>web-service</name> <label>user-provided</label> <userProvidedCredentials> <url>http://example.com/service</url> <accessKey>abc123</accessKey> </userProvidedCredentials> </service> </services> </configuration> </plugin> </plugins> </build>

[Stephan Oudmaijer]

The settings I am talking about are cf deployment properties for the cloudcontroller.