Unresponsive agents

Oleg Shaldibin

unread,

Apr 15, 2012, 9:26:42 PM4/15/12

to bosh-...@cloudfoundry.org

This means your VMs have never been created in the first place, so there might be a problem with your infrastructure or CPI. VM references are in the BOSH DIrector database but they don't have cloud ids, so BOSH is unable to do anything meaningful. We'll need more context: can you provide us with either your original 'create deployment' debug task log or your director hostname (you can get debug task log with bosh tasks recent <n>, followed by 'bosh task <create-deployment-task-id> --debug').

--
Best,

Oleg

Dr Nic Williams

unread,

Apr 15, 2012, 9:51:02 PM4/15/12

to bosh-...@cloudfoundry.org

I've added debug output to my gist.

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Dr Nic Williams

unread,

Apr 15, 2012, 9:51:32 PM4/15/12

to bosh-...@cloudfoundry.org

Yes, it doesn't look like the VMs got created.

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Oleg Shaldybin

unread,

Apr 15, 2012, 10:07:10 PM4/15/12

to bosh-...@cloudfoundry.org

OK, looks like some compilation VMs got created in the DB but not in the infrastructure. Right now it's easiest just to delete this deployment and try deploying from scratch (you might need to use --force to delete it). We'll add cloudcheck handler for this particular situation soon.

--
Best,

Oleg

Dr Nic Williams

unread,

Apr 15, 2012, 10:11:03 PM4/15/12

to bosh-...@cloudfoundry.org

Ahh, I can imagine that happening as I changed the compilation settings along the way (to make them AWS specific). Deployment deleted and recreated and its underway. Thanks.

Nic

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Dr Nic Williams

unread,

Apr 15, 2012, 10:25:28 PM4/15/12

to bosh-...@cloudfoundry.org

Things are looking good:

Compiling packages

wordpress/0.1-dev (00:02:52)

mysql/0.1-dev (00:03:10)

mysqlclient/0.1-dev (00:02:50)

nginx/0.1-dev (00:05:08)

apache2/0.1-dev |oooooooooooooooo | 4/6 00:14:55 ETA: --:--:--

Oooh, I forgot to add the "reuse" flag for compilation :)

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Dr Nic Williams

unread,

Apr 15, 2012, 11:12:54 PM4/15/12

to bosh-...@cloudfoundry.org

Next error (getting close!) - "Can't provide any defaults since this is a VIP network"

$ bosh deploy

Getting deployment properties from director...

Compiling deployment manifest...

Please review all changes carefully

Deploying `wordpress-aws.yml' to `yourboshname' (type 'yes' to continue): yes

Tracking task output for task#20...

Preparing deployment

binding deployment (00:00:00)

binding release (00:00:00)

binding existing deployment (00:00:00)

binding resource pools (00:00:00)

binding stemcells (00:00:00)

binding templates (00:00:00)

binding unallocated VMs (00:00:00)

binding instance networks (00:00:00)

Done 7/7 00:00:00

Preparing DNS

binding DNS (00:00:00)

Done 1/1 00:00:00

Creating bound missing VMs

common/0: Can't provide any defaults since this is a VIP network (00:00:00)

Error 1/3 00:00:00

Debug is at https://gist.github.com/87061a9b550cefbe7d16

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Vadim Spivak

unread,

Apr 15, 2012, 11:16:05 PM4/15/12

to bosh-...@cloudfoundry.org

You have to provide a dynamic network as well, since a VIP network is not known to the guest.

-Vadim

Dr Nic Williams

unread,

Apr 15, 2012, 11:22:39 PM4/15/12

to bosh-...@cloudfoundry.org

Currently I am combining:

* code reading - though its still hard to figure out how the manifest files maps into the various Bosh::Director::XXX models; and

* wild guessing of what schema to go with for an AWS configuration

Do you have any suggestions for changes to the manifest file (in the gist debug output)?

Nic

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Dr Nic Williams

unread,

Apr 15, 2012, 11:23:14 PM4/15/12

to bosh-...@cloudfoundry.org

BTW, all the validation errors that it does give you are very helpful. Thanks for putting those in.

Nic

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Vadim Spivak

unread,

Apr 15, 2012, 11:41:22 PM4/15/12

to bosh-...@cloudfoundry.org

Add the following to the network section for each job.

- name: default

default: [dns, gateway]

-Vadim

Dr Nic Williams

unread,

Apr 16, 2012, 12:47:27 AM4/16/12

to bosh-...@cloudfoundry.org

Thanks Vadim, I now have 3 lovely VMs booted.

Though, the system doesn't currently work - none currently have elastic IPs, and if I point my browser at each AWS-generated public dns, one of them shows a 504 nginx error (yep, the random nginx VM).

From chatting with Oleg on Friday, I think each VM needs a well-known elastic IP. How do I do this?

My attempts to give each job a specific vip network that had a static IP end me back with the error from before: "Can't provide any defaults since this is a VIP network"

The two manifests are at https://gist.github.com/75ced173309bba0d8b23. An example of a failing job is below:

jobs:
  - name: nginx
    template: nginx
    instances: 1
    resource_pool: common
    networks:
    - name: nginx
      default: [dns, gateway]
      static_ips:
        - 23.23.245.179
    cloud_properties:
      instance_type: m1.small

Thanks again for all help. Soon, we will have a lovely AWS tutorial :)

Nic

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism

Dr Nic Williams

unread,

Apr 16, 2012, 12:48:09 AM4/16/12

to bosh-...@cloudfoundry.org

I applied my first changes to a deployment manifest and it is very cool to be shown the changes before they are applied.

--

Dr Nic Williams

Engine Yard, VP Developer Evangelism