[ansible-project] Rotate groups out of load balancer

Apologies in advance here: I'm missing something fundamental. I'm trying to do something pretty simple: I'm putting new code into a QA environment, and would like to remove old machines out of a elastic load balancer and substitute in new machines. I have something that looks like: --- #

Re: [ansible-project] Re: Ansible / tower setup - problems

Just as a follow up here, we don't recommend running the very latest devel Ansible with Tower--we'd suggest sticking to the officially-released versions. There's no reason why it shouldn't work, but you may run into issues that our test matrix doesn't catch. If you're just looking for new or

[ansible-project] Re: Ansible / tower setup - problems

Thank you Sean, your suggestion worked for me and saved me time. Cheers. On Wednesday, March 4, 2015 at 12:31:19 AM UTC-5, Sean Herrala wrote: In case anyone sees this and wants an actual answer, the problem is that the 'sanity check' in the configure script that does a version check looks

Re: [ansible-project] Ansible capable of handling diskless clients?

I don't see a reason it would not. First ansible normally is run in push mode so not much for it to manage 'locally'. Managing services would be the same as with full clients, the issue here is handling the per server specific configs and general configs. Most of this can be done with ansible

Re: [ansible-project] Get Ansible Job ID inside playbook

Hi Ivo, is this a Tower question? In ansible core, the job id only exists for async tasks. Thanks! James Cammarata Director, Ansible Core Engineering github: jimi-c On Thu, May 28, 2015 at 7:54 AM, Ivo Hechmann ivo.hechm...@gmail.com wrote: Hello; How can I retrieve the current job id? Want

[ansible-project] Re: Ansible 1.9.1 SSH Error from ping module to Junos

So, after continuing to look at this it appears that SSH / paramiko are working but the ping module is failing to load the python interpreter. There are symbolic links already created from /usr/bin/python --- /usr/bin/python2.7 I tried specifying the direct path to the hosts file

[ansible-project] Re: cannot get win_template to work

Thanks for the reply, I've got a work around but do need to figure out what was going wrong: - name: create script locally local_action: template dest={{ my_dest }} src={{ my_template }} - name: run ps script on remote script: {{ my_ps_script }} On Thursday, May 28, 2015 at 7:04:42 AM

[ansible-project] Ansible 1.9.1 SSH Error from ping module to Junos

Good afternoon all, my apologies for posting but I have been stopped at this point with no result for a few days. I have installed Ansible 1.9.1 and , I believe, all required dependencies on CentOS 7 , I am looking to use Ansible to manage Junos equpment and installed the Juniper.junos module

Re: [ansible-project] Notify handler only one in the same role

Note: these tasks are in the same role. On Thursday, May 28, 2015 at 6:05:35 PM UTC-7, Steven Truong wrote: Hi Matt, Here are some tasks that call the same notify and it happened that these files might got changed during the same run. - name: creates hadoop .sh config files template:

Re: [ansible-project] Notify handler only one in the same role

Hi Matt, Here are some tasks that call the same notify and it happened that these files might got changed during the same run. - name: creates hadoop .sh config files template: src={{ item }}.j2 dest=/etc/hadoop/conf.{{ nameservice_id }}/{{ item }} owner=hdfs group=hadoop mode=0755

Re: [ansible-project] Re: Can't get the group_id value from register in ec2_group on creation

Any idea how to dig out the sg id out of this? Maybe changing the module to not be so verbose and print only the needed info would be easier? On Thursday, May 28, 2015 at 3:59:14 PM UTC+10, Igor Cicimov wrote: Thanks for replying Benno. I did exactly that with debugging and can see where the

[ansible-project] Notify handler only one in the same role

Hi all, I am using ansible 1.9.1 and in the same role I have a couple of notify to the same service to be restarted. Apparently, ansible tries to restart 3-4 times for each notify and that is undesirable. I could do without notify and create register for each task and then at the end I

Re: [ansible-project] Notify handler only one in the same role

I think you are going to need to show an example of the playbook/role and the output of this happening. On Thursday, May 28, 2015, Steven Truong djatlan...@gmail.com wrote: Hi all, I am using ansible 1.9.1 and in the same role I have a couple of notify to the same service to be restarted.

[ansible-project] How do you parameterize ansible_eth* facts and use it?

Hello, ansible experts. I'm trying to parameterize eth interface in `group_vars/all` so that I can use it in playbook easier. group_vars/all file looks like this. my_int_if: eth1 my_int_ip: {{ ansible_eth1.ipv4.address }} my_int_obj: {{ ansible_eth1 }} I could successfuly use those

[ansible-project] Re: How to pass list of arguments that contains = signs to raw command ?

raw: docker run {{ container_commands[container_name] }} registry:2.0 raw: Executes a low-down and dirty SSH command, not going through the module subsystem. This is useful and should only be done in two cases. The first case is installing python-simplejson on older (Python 2.4 and

[ansible-project] directory layout for grouping playbooks - ansible.cfg ?

Hi, we have several playbooks and roles in our project so I try to organize them in that way: \---deployment \---trunk | ansible.cfg | app1.yml | yml | app[N].yml +---files +---filter_plugins +---group_vars

[ansible-project] Re: directory layout for grouping playbooks - ansible.cfg ?

in addition, also the group_vars cannot be resolved - they are expected relative to the playbook(s) - or the inventory hosts file. = how do i define common group_vars/all.yml - when i want to structure my playbooks in (sub) directories ? Regards, Torsten -- You received this message because

Re: [ansible-project] Ansible get variable output

Hi, iirc value for register are not passed via the templating engine, maybe this should work to get the username and thier associate password. - name: Generate a random password per user shell: /usr/bin/openssl passwd -1 -in /dev/urandom | head -1 with_dict: users register: foo

[ansible-project] Get Ansible Job ID inside playbook

Hello; How can I retrieve the current job id? Want to create a Jira Issue using the jira module and add the ansible job ID as description... Thank you very much Ivo -- You received this message because you are subscribed to the Google Groups Ansible Project group. To unsubscribe from this

Re: [ansible-project] Making use of Ansible Privilege Escalation

ansible cannot chain privilege escalation methods, it can either use pbrun or su, but not both. pbrun should be able to allow you to execute all commands as a user w/o needing su. On Wed, May 27, 2015 at 1:19 PM, Aneesh Joseph aneeshkj1...@gmail.com wrote: I'm trying to find the equivalent of

Re: [ansible-project] cannot get win_template to work

Hi Jake, Seems like the win_template module was readded 6 days back , are you testing this in the latest devel branch ? - Benno On Thu, May 28, 2015 at 1:07 PM, Jake Dupuy jake.du...@gmail.com wrote: someone please help me out. I cannot get win_template or win_copy to work. I'm able to do

Re: [ansible-project] Ansible get variable output

Hey Benno, This answer is awesome! It works! (We spend so many hours with nested variables and for loops and all kinds of tricks, this is short and sweet. THANKS!) Thank you Benno. Rogier On Thursday, May 28, 2015 at 2:24:56 PM UTC+2, benno joy wrote: Hi, iirc value for register are

[ansible-project] cannot get win_template to work

someone please help me out. I cannot get win_template or win_copy to work. I'm able to do a simple ipconfig test to confirm that ansible is working with windows I've got some powershell scripts that I need to have in a template with vars so I can put them on the windows box and run them. I've

[ansible-project] Ansible capable of handling diskless clients?

Hi all, a similar question was already posted on Serverfault (http://serverfault.com/questions/693698/config-management-for-nfs-root-cluster) but got turned down as being off-topic. To summarize: Is Ansible capable of handling diskless clients that only have read-only storage mounted (besides

[ansible-project] Re: directory layout for grouping playbooks - ansible.cfg ?

Hi, I'm also in the situation to get many different projects under control and use following directory structure: - one folder per project with dependeing playbooks and roles - global inventory with one folder per stage (since the project share hosts I made only one inventory ) - group_vars and

[ansible-project] Using shell module and referencing the variable in a playbook

I am running a bash script and one of the lines is : conf_file=*x.conf* The config file simply has a variable such as: screen_url:http://www.yahoo.com The problem is that with multiple machines you would need multiple versions of the script. So the solution is to set the variable or load

[ansible-project] stdout empty

I'm new to ansible, and setting up a POC to run puppet for a specific orchestration task. However, when I run my agent runs, I'm not getting anything back in stdout, even when changes are set to occur. Here's what's in my playbook for this task. - command: /opt/puppet/bin/puppet agent -t

[ansible-project] Re: How to pass list of arguments that contains = signs to raw command ?

Have you tried command or shell? Yes I did, I guess it's a general security safeguard . But how to express/escape it correctly. I now work around through generating a script file through a template and executing that. But I guess there must be a more elegant way? -- You received this

Re: [ansible-project] Re: How to pass list of arguments that contains = signs to raw command ?

Your example seems to work with command instead of raw: $ ansible-playbook --version ansible-playbook 1.9.1 configured module search path = None $ ansible-playbook /var/tmp/equal.yml PLAY [localhost] ** GATHERING FACTS

[ansible-project] notify and handlers: in a role dont work

Hi i think this is a bug but can be that i dont understand the Docs ... i have main play ``` - hosts: all gather_facts: yes user: root vars_files: - nagios/GlobalVar/main.yml vars_prompt: - name: password prompt: MYSQL Pass that you will use Enter password