Probably add some extra scripting to check if task1 and task2 are started before kicking off other tasks by doing a sleep-wakeup-check-sleep loop till task1 and task2 are really running as they are

Oh nice trick, but that's not a bosh default behavior, just use for

the docker bosh release supports "depends on" to start containers in a particular order: https://github.com/cf-platform-eng/docker-boshrelease/blob/master/examples/docker-aws.yml#L88-L89 -- Thank

Indeed :) allan.espinosa(a)rakuten.com> wrote:

Argh sorry. Read the cpi job spec again Apparently I have the hash at the wrong level. s/cpi.env.http_proxy/env.http_proxy/

Hi, We need specify https_proxy to connected to our avcenters. We're able to do this when we deploy Bosh directors using bosh micro and multi-bosh setups. But I cannot get it running under bosh

In my case, I create job for each container and all are run within same vm. So the result is out of control. Except run all containers in one job. Is it possible that change some job to package,

With YAML it's very much different to have *--- !ruby/object:DeploymentManifest * *networks: * *- !ruby/object:AWSDynamicNetworkBean * * type: dynamic* vs *---* *networks: * *- **type:

I am not sure why it fails now, but my two cents are why not just serialized the object? In the case of python, i would see that if I try to save an object like in your case:

IIRC, Bosh would kick off the jobs in the order you have declared (serial). So, you can have jobs 1 and 2 go first, then the others... If there are multiple job templates all bundled to run within the

Hi, I am aware that this does not have anything to do with BOSH but when a ruby object is converted to yaml, this format is appended automatically. Couple of months ago, I was able to deploy with a

You can not control the sequence when there are several compilation workers but you can set the workers number to 1. In the manifest: compilation: workers: 1 Also, you can run all of the

But How to control these jobs run sequence? I tried these jobs will run concurrently.

Your manifest is wrong "*- !ruby/object:AWSDynamicNetworkBean *" are not a yaml field. You script using to_yaml is just wrong, nothing related to bosh.

Right there is no dependency between job, but I think in the manifest you can put an order job: -name: docker1 ... -name:docker2 .... etc... But dependency don't really exist in bosh.

Hi, would be great if someone could please help me with this matter. Regards, Kinjal

I have five Docker image which have following runtime dependency: The Docker image 1 and Docker image 2 must be run before Docker image 3, 4 ,5 I create five bosh Jobs for them, but bosh document said

Example : Get the file here for example , http://www.oracle.com/technetwork/java/javase/downloads/jre8-downloads-2133155.html You need to add to the blobstore to be able to use it, for dev local

i understand somehow bosh works not of al of it i created this a 2 months ago https://github.com/cloudfoundry-community/bosh-scaler-boshrelease and i understand how to extract it and use the

Up to you but nornally you just need to create a package to install openJdk, for example https://github.com/cloudfoundry-community/jenkins-boshrelease/tree/master/packages we install the jre, should