Yes Packer is probably not the tool you are looking for. It's more for
creating immutable cloud and VM images.
On Feb 1, 2018 22:32, "Larry Velez" wrote:
>
> I might have googled my way to one possible answer to what I am looking
> for which might be more in line with our laptop needs than Packe
Hmm check the plugin and it looks correct. Can you share your template and
command line when you get this error?
On Jan 31, 2018 20:00, "Rickard von Essen"
wrote:
> I think you are reading the error message backwards. Packer converts the
> value to a number, but the plugin (whi
I think you are reading the error message backwards. Packer converts the
value to a number, but the plugin (which is unknown to me) expects a
string. I think this is probably a bug in the plugin.
On Jan 31, 2018 19:08, "Kevin Milner" wrote:
> Hello,
> In my variables portion I want to be able to
That is how it works out of the box, see
https://www.packer.io/docs/provisioners/powershell.html#valid_exit_codes
On 30 January 2018 at 13:19, shailender chauhan <
shailendersinghchau...@gmail.com> wrote:
> Hi, I would like to cancel Packer AMI creation if one of my provisioner
> powershell scrip
if [[ $# != 0 ]];then
>>
>>
>> while getopts ":i:n:g:s:t:f:dvh" Option
>> do
>> case $Option in
>> i ) _IP=$OPTARG;;
>> n ) _NM=$OPTARG;;
>> g ) _GW=$OPTARG;;
>> * s ) _HN=$OPTARG;;*
>>
>> On S
See
https://www.packer.io/docs/provisioners/shell.html#default-environmental-variables
On Jan 26, 2018 17:57, "'Andreas Sommer' via Packer" <
packer-tool@googlegroups.com> wrote:
> {
>> "type": "shell-local",
>> "command": "env vm_name={{user `vm_name`}} ./packer/scripts/do-something-
>> with-nam
t;
>
>
> On Thursday, January 25, 2018 at 1:33:39 PM UTC+8, Rickard von Essen wrote:
>>
>> You are missing on pair of [ ] in your vboxmanage_post command, it should
>> be:
>>
>> "vboxmanage_post": [
>> ["modifyvm", "{{.Name}}",
[--ioapic on|off]
> 2018/01/25 10:26:44 packer: [--hpet on|off]
> 2018/01/25 10:26:44 packer:
> [--triplefaultreset on|off]
> 2018/01/25 10:26:44 packer: [--apic on|off]
> 2018/01/25 10:26:44 packer: [--x2apic on|off]
"export_opts":
> [
> "--manifest",
> "--vsys", "0",
> "--description", "create vm with nat and then change to bridged",
> "--version", "2.0"
> ],
> "source_path"
1) You really want >> to add to the /etc/sshd_config, your single >
truncates the file.
2) Don't sudo if you run as root.
Instead of reinventing the wheel start with a well-known template such as
https://github.com/chef/bento and adapt in small iterations.
My general advice is when you run into p
See https://www.packer.io/docs/builders/virtualbox-iso.html#vboxmanage_post
On Jan 24, 2018 14:11, "JerryWang" wrote:
> this steps should be done after provioner ( and vm was poweroff ), where
> to put this operation ? builder ? or post-processor ?
> thanks.
>
> --
> This mailing list is gov
Bridge mode works, but as Alvaro said packer can't know the IP of the
guest. So either you run with a fixed IP or you add the bridge network but
also keep the nat if.
On Jan 22, 2018 08:26, "Alvaro Miranda Aguilera" wrote:
> the nat mode will use the host networking to reach the network, why you
We just include the git short hash in the name of the image (AMI) and tag
with source AMI. But we only build cloud images and only for internal use.
On Jan 19, 2018 22:55, "Nate Dobbs" wrote:
> This isn't an issue for packer per se but this may be the best place to
> discuss the topic. Those of
Can you share your packer template?
On Jan 20, 2018 06:01, "HSB" wrote:
> Hi, Created new VM Box using Vagrant. I can SSH and also connect to
> internet from that host.
>
> But when I execute packer build (type: lxd) with shell commands for ubuntu
> updates, I get following error. I can CURL all
How does your template look?
On Jan 15, 2018 21:35, "Louis Mayorga" wrote:
> found error: Failed to upload packer-win2016gui-vmware/packer-vmware-iso.ova:
> InvalidParameter: 1 validation error(s) found. minimum field size of 1,
> CreateMultipartUploadInput.Key.
>
> I thought that i was probably
Thanks for sharing the solution!
On Jan 15, 2018 19:52, "Patrick Slattery" wrote:
> I was able to resolve my issues by the following steps:
>
> 1) I used a HTTP proxy (Fiddler on Windows) to determine exactly what
> errors were being returned by the Rackspace API, this lead me to believe
> that
>
> Lucky
>
>
> On Friday, January 5, 2018 at 3:06:01 PM UTC-8, Rickard von Essen wrote:
>>
>> To my knowledge it's not possible. Do you have an imported boot disk?
>>
>> On Jan 4, 2018 23:33, "Lucky" wrote:
>>
>>> Hi
>
Most likely this is because you are building against a newer kernel version
that you just installed (with apt-get update etc) but is not yet running
since you haven't rebooted.
See how bento does it:
https://github.com/chef/bento/blob/master/_common/virtualbox.sh
https://github.com/chef/bento/blob
, ask(retry)
>
> sometimes I know the error is not critical and cannot avoid it for some
> reason, seems no way to ignore the error to continue build box?
>
> BR/
> Robert
>
>
> On Saturday, January 6, 2018 at 2:32:09 AM UTC-5, Rickard von Essen wrote:
>>
>> Pac
Packer always stops when script exits with a non zero exit code. Try adding
"set -x" at the top of your script (after the shebang) and se why it fails.
You can also run packer build with -on-error=ask to get time to ssh into
the VM and research the state it is in when it fails.
On 6 January 2018
turday, January 6, 2018 at 4:22:18 AM UTC+1, Peter Goodall wrote:
>
> Thanks Rickard,
>
> "extra_arguments": ["-v"]
>
> --Peter G
>
> On Saturday, January 6, 2018 at 10:07:49 AM UTC+11, Rickard von Essen
> wrote:
>>
>> I think
I think I have that somewhere, let me get back with it tomorrow.
What's your extra_arguments to ansible?
On Jan 5, 2018 09:09, "Peter Goodall" wrote:
> also... the lxd builder runs shell commands fine remotely on the guest.
>
> --
> This mailing list is governed under the HashiCorp Community G
To my knowledge it's not possible. Do you have an imported boot disk?
On Jan 4, 2018 23:33, "Lucky" wrote:
> Hi
>
> I was wondering, if there's equivalent in packer to create an image out of
> GCE's instance bootable disk?.
>
> Something what gcloud cli suggests, such as
> gcloud compute --proje
Run packer build -on-error=ask template.json
And debug your network setup from the console.
On Jan 5, 2018 10:46 PM, wrote:
> That's a typo and has been fixed. Somehow, only one build passed the
> waiting for ssh stage and almost succeeded. But when I repeated the build,
> it started to fail ag
Seems that there is missing some alignment there. When you have done some
research feel free to open a new issue either on Packer or Terraform to get
them aligned.
On Jan 1, 2018 23:38, "Dave Woodward" wrote:
> In a nutshell, the Packer vsphere-template post processor is clearing the
> UUID whil
I think as Alvaro said the simplest solution is to build a small wrapper
script and pass the source_image in as a variable. Another better
alternative is to implement something similar to amazon's source_ami_filter
1) and open a PR to include this in the openstack builder.
1) https://www.packer.io
lp. Wish there was an easy way for me to buy you a cup of
> coffee
>
> On Thu, Dec 21, 2017 at 9:17 AM, Rickard von Essen <
> rickard.von.es...@gmail.com> wrote:
>
>> It looks like you copied this from a PDF or something similar because if
>> I look really closely i
e build: false
> 2017/12/21 08:47:25 On error:
> 2017/12/21 08:47:25 Waiting on builds to complete...
> 2017/12/21 08:47:25 Builds completed. Waiting on interrupt barrier...
> 2017/12/21 08:47:25 ui:
> ==> Builds finished but no artifacts were created.
> 2017/12/21 08:47:25 [INF
There should only be one binary nowdays. If you run: PACKER_LOG=1 packer
build template.json
You will get a debug log. Paste it here and I'll think it will be easy to
help.
/ Rickard
PS. Awesome book ;-)
On Dec 21, 2017 13:30, "Gray Jones" wrote:
> I have tried to install on a mac os x laptop.
"Primary": true,
>>> "PrivateIpAddress": "172.31.89.61",
>>> "Association": {
>>> "IpOwnerId": "amazon",
>>>
You probably just have to set the IP in ssh_host, see
https://www.packer.io/docs/templates/communicator.html#ssh_host
On Dec 16, 2017 14:29, wrote:
Hi Alvaro,
Thanks for quick response, I've attached the 'myproduct.json'. VM comes up
with IP 172.15.0.40 (private IP) and I can ping other hosts i
Doesn't
services --disabled=NetworkManager
make dhcp not run automatically. I think you should get everything running
before changing the default network setup.
On 15 December 2017 at 06:54, Shankar Bala wrote:
> hi
>
> below is my kickstart file
>
> ===
>
>
> # KICKSTART
>
> ##key --s
Please supply your Kickstart file.
On Dec 15, 2017 05:28, "Ian W" wrote:
> Do you have something like the following in your kickstart file:
>
> network --bootproto=dhcp --device= --activate
>
>
>
> On Thursday, December 14, 2017 at 11:23:02 PM UTC-5, Shankar Bala wrote:
>>
>> Hi All
>>
>> I hav
> Or perhaps I have some Amazon IAM issue? I used the suggested IAM template
> out of the documentation.
>
> On Dec 14, 2017 3:29 AM, "Rickard von Essen"
> wrote:
>
>> That exact template worked for me.
>>
>> On 14 December 2017 at 06:21, Ian W wrote:
&g
That exact template worked for me.
On 14 December 2017 at 06:21, Ian W wrote:
> It always fails the 'waiting for instance' stage.
> I have tried various sizes of instance_type
>
> ==> amazon-ebs: Adding tags to source instance
> amazon-ebs: Adding tag: "Name": "Packer Builder"
> 2017/12/13 2
Packer doesn't manage the content of the ovf.
On Dec 13, 2017 23:44, wrote:
> Not sure if this is a packer issue, Virtualbox issue, or a VMware issue.
> Starting here with packer.
>
> Creating images with packer 1.1.3 and have "keep_input_artifact": true.
>
> I assumed the output-virtualbox-iso/
This is probably better directed to the terraform list instead of the
packer user list.
On Dec 11, 2017 16:32, "Hyeongseob Kim" wrote:
> Hi everyone!
>
> I'm using terraform for my project.
> and I met critical problem with terraform.
>
>
> First, I created redis using following terraform module
],
>
> "post-processors": [
>
> [
>
> {
>
> "type": "docker-tag",
>
> "repository": "hello/cassandra-demo",
>
> "tag": "0.1"
>
> }
>
>
I can use
> to spawn a container.
>
> when I run packer build packer.json I get the following
>
> *==> docker: Exporting the container*
>
> *==> docker: Killing the container:
> b33b07f7f0ecf5026574d29ee28aed0f71659ae11955994ac27fca060c7e7c22*
>
> *Build &#
See my answer on SO https://stackoverflow.com/a/47747555/226174
On 10 December 2017 at 22:12, kant kodali wrote:
> Can anyone here provide a simple working example using my packer.json file
> in my previous email ? That would be super helpful since I can try running
> things myself and understan
Compare your network settings with the one in:
https://github.com/hashicorp/packer/issues/5303
And this suggest that you need "ethernet0.networkName", see
https://github.com/hashicorp/packer/issues/4540
And packer will not work if there is a vmotion event, see
https://github.com/hashicorp/packer/
>
> So my questions
> 1) which would mac users recommend specifically for working with packer
> Vmware or Virtualbox
>
Parallels Pro is also a good option, even it's a bitt more expensive, but
you get more features, especially if you want to run Windows on your mac.
> 2) Does the basic version of
not very convenient).
On 6 December 2017 at 06:30, Justin Keller wrote:
> Sorry, builder is amazon-ebs.
>
> On Tue, Dec 5, 2017 at 9:26 PM, Rickard von Essen <
> rickard.von.es...@gmail.com> wrote:
>
>> Which builder are you using?
>>
>> On Dec 6, 2017 04:0
Which builder are you using?
On Dec 6, 2017 04:06, "Justin" wrote:
> What is the recommend way to add SSH public keys to a remote user's
> authorized_keys? Should I just use the file provisioner and transfer a
> complete authorized_keys into ~/.ssh? Is there a specific module for this?
> Thanks.
Pretty sure that is VRDP which is a superset of RDP. Do you have anything
that backs up that it's actually VNC?
// Rickard
On 3 December 2017 at 07:19, wrote:
> Building a Windows 10 virtualbox image from https://github.com/
> StefanScherer/packer-windows.git
>
> $ packer build -only=virtualbox
Can you share your playbook or something that reproduces the problem?
On Dec 3, 2017 02:32, "Rajinder Singh" wrote:
> I am baking an image in Azure using packer.
> Base image is Centos 7.4
>
> If I run packer from MacOS image bakes successfully.
> If I try to bake the exact same image from Ubunt
Maybe you have a firewall on your host intercepting the traffic. Try adding
-on-error=ask and do some troubleshooting of your network set up.
On Nov 28, 2017 12:15 PM, wrote:
> FYI: Goal is to provision to vSphere.
>
> --
> This mailing list is governed under the HashiCorp Community Guidelines -
lect enp0s3
>
> # To set a different link detection timeout (default is 3 seconds).
> # Values are interpreted as seconds.
> #d-i netcfg/link_detection_timeout string 10
>
> # If you have a slow dhcp server and the installer times out waiting for
> # it, this might be useful.
&g
may be because of that
> the pressed.file is not accessible for the VM
>
> Correct me if i'm wrong and what could be the problem to retrieve the
> preseed file.
>
> Regards
> Valli
>
> On Sunday, 26 November 2017 01:03:09 UTC+5:30, Rickard von Essen wrote:
>>
&g
Just to clearify, why do you need a bridged network?
On Nov 25, 2017 20:04, "valli nayagam" wrote:
Hello everyone,
I'm newbie to packer, I'm trying to install debian os in virtualbox. But it
looks network is not connected with host machine and failed to retrieve the
preseed file.
Json file:
{
You can use a shell and file provisioner to pass the ip to the host.
Similar to this example (but you need to change hostname to ip etc.)
"provisioners": [
{
"type": "shell",
"inline": [ "hostname > /tmp/hostname" ]
},
{
"type": "file",
"direction": "download",
"source":
Here you go
https://github.com/Diabol/packer-demo/blob/master/windows-cloud-ansible.json
On Nov 23, 2017 17:57, "SurferL" wrote:
> Another point to add, is that it pretty much always fails on long
> playbooks / roles (with multiple installations via win_chocolatey for
> example), but works fine
ws-shell...
>
> On Thursday, 23 November 2017 14:23:28 UTC, Rickard von Essen wrote:
>>
>> Shell-local would run on your host, ie your macOS machine.
>>
>> On 23 November 2017 at 15:13, SurferL wrote:
>>
>>> Running Packer on MacOS -> to setup a Windo
Shell-local would run on your host, ie your macOS machine.
On 23 November 2017 at 15:13, SurferL wrote:
> Running Packer on MacOS -> to setup a Windows machine
>
> I have Cygwin installed on my Windows machine and would like to use that
> for a "shell-local" command, rather than the default comm
7b1aaf7afb22046) to become
> ready: unexpected state 'shutting-down', wanted target 'running'
>
> ==> Builds finished but no artifacts were created.
> Finished: FAILURE
> Build step 'Packer' marked build as failure
> Warning: you have no plugins pro
ome ready: unexpected state 'shutting-down',
> wanted target 'running' [0m
>
> ==> Some builds didn't complete successfully and had errors:
> --> amazon-ebs: Error waiting for instance (i-xxx) to become
> ready: unexpected state 'shutting-down&
Unfortunately that doesn't seem to be supported. See
http://docs.aws.amazon.com/AWSEC2/latest/APIReference/API_EbsBlockDevice.html
and
http://docs.aws.amazon.com/AWSEC2/latest/APIReference/API_RunInstances.html
As a workaround you can prepare a CMK encrypted (empty) snapshot and attach
that in you
To my knowledge that doesn't work on openstack. Use qemu and create a raw
disk locally and then import it to OS.
On Nov 16, 2017 14:18, "MartinKenneth Michalsky"
wrote:
> Hi, I was reviewing the posts and the documentation, I did not find a way
> to create an openstack image starting from a loca
Rubber ducking ;)
-only=virtualbox-iso
On Nov 16, 2017 06:52, wrote:
> I don't think this is a bug, feels like a knowledge problem on my part,
> post here vs opening an issue.
>
> I use packer on macOS all the time, I've been tasked with setting up
> packer on a Xenial (linux) machine.
>
> Foll
See my answer on SO https://stackoverflow.com/a/47301671/226174
On 15 November 2017 at 00:02, noisy noisyBits
wrote:
> Hi all,
>
> I am struggling to run serverspec test on GCP packer instance from local
> machine.
> I have created a packer json config file as follows to create an image on
> GC
I think you are correct on all points.
I think the best way to solve this is to open an issue on github and gather
relevant facts (as you already started to do). Most important is to
reference in which OS version this was dropped and figuring out if there is
any non EOL version of OS that needs th
Could you provide your template and playbook?
On 10 November 2017 at 15:23, Travis Stevens wrote:
> This question mixes Ansible and Packer as I am having a lot of problems
> locating the issue.
> We have a build server that currently executes an Ansible playbook that
> calls packer to create new
Hard to say with just fragments of your template etc.
But verify that you have NOPASSWD set in /etc/sudoers.conf for the user
connecting to the VM and that it don't require tty (requiretty).
If those doesn't help you supply a full template and scripts/playbook.
On Nov 7, 2017 07:03, "Rajinder Si
See https://www.packer.io/docs/builders/amazon-ebs.html#tags
On Nov 6, 2017 18:33, wrote:
> I am using source_filter_ami to select the latest AMI and it is working
> great. Is there a way I can reference the AMI the source filter selected so
> I can Tag the new AMI with what source AMI was used
You can't interpolate type and it wouldn't help much since they generally
will have different options. If you really want to template type you can
preprocess your template (with sed etc.).
On Nov 2, 2017 22:04, "Sabitha M" wrote:
> Hi ,
>
> Here is my packer template. I am trying to pass command
VirtualBox network is a bit quirky. By default packer sets up nat for ssh.
So by default it ssh to 127.0.0.1: where this is NAT'ed to the
guest IP port 22. If you want to suppress this see
https://www.packer.io/docs/builders/virtualbox-iso.html#ssh_skip_nat_mapping
But if you are just learning Pac
It was some time since I worked with OS but I think they actually
deprecated extentions in favor of micro versions. Maybe the hole dependency
on extensions can be removed.
It would be great to get some up-to-date facts on the current situation.
On Oct 31, 2017 16:34, wrote:
> Would it be possib
There is not much magic here, packer just executes "ansible-playbook" (see
https://www.packer.io/docs/provisioners/ansible-local.html#command).
Di you place the ansible.cfg in the correct directory?
On Oct 31, 2017 14:19, "Tim Coote" wrote:
> Is there a way to get the packer ansible-local provi
can you get the service catalog with the cli with the same creds?
On 26 October 2017 at 15:57, Holger Protzek
wrote:
> Debugging this with a test case:
>
> catalog, err := result.ExtractServiceCatalog()
> if err != nil {
> return err
> }
>
> returns for catalog
>
> len:0, cap:0
>
> without an er
What's the structure of your post-processor section in your packer
template? Is it a sequence?
On Oct 25, 2017 18:20, "John C" wrote:
> Currently I'm using Packer to build a VMWare template and then have
> configured the post-processor section of the JSON with multiple vsphere
> tasks, to copy t
It is executed but in it's own shell since you use the shell form not the
exec form.
See
https://www.ctl.io/developers/blog/post/dockerfile-entrypoint-vs-cmd/
On Oct 20, 2017 23:09, "Raghava Karthik" wrote:
> I can see the command " /bin/sh -c source /etc/environments" in docker
>> inspect , bu
Could you justify "doesn't work", for example what's the output of "docker
inspect "?
On 20 October 2017 at 16:10, Raghava Karthik
wrote:
> I was trying to run a command using CMD along with ENTRYPOINT (entrypoint
> works perfectly).
> All i want to do is load environment variables "source /etc/
d on github? Thanks!
>
> On Friday, 20 October 2017 17:38:21 UTC+13, Rickard von Essen wrote:
>>
>> you see no value whatsoever in adding an ability to preserve the VM after
>>> post-processor starts running and then fails in order to investigate the VM
>>> state?
t consistently work all the
> way through?
>
> Cheers,
> Andrew
>
> On Thursday, 19 October 2017 17:37:42 UTC+13, Rickard von Essen wrote:
>>
>> 1) the -on-error only influence the execution of builders (which
>> provisioning is a part of). That should be made clear
It should be pretty straight forward. Start with something that works, say
a bento box 1). When you got a hang of that just add your customisations
with puppet-server provisioner 2).
If you have any problems on the way just send a question here.
// Rickard
1) https://github.com/chef/bento/
2) ht
1) the -on-error only influence the execution of builders (which
provisioning is a part of). That should be made clear in the docs.
2) unfortunately stdout/stderr gets swallowed, it looks like someone
intended it to be passed to packer but did it wrong. (in the meantime the
best way to troubleshoo
2300218 is a canary value set when SSH is disconnected without the server
closing the channel. Usually this happens when you restart/upgrade sshd or
reboots in a script. See
https://www.packer.io/docs/provisioners/shell.html#expect_disconnect
expect_disconnect should only be set to true if you un
A snapshot (AMI) is much more predictable than launching an EC2 with a
gigantic user-script that installs a big system. The draw back is that you
have to re-spin the image regularly to update. I immutable servers have the
same pros and cons regardless if they are the core of your infrastructure
man
Can you open an issue for this. Quickly checking the code it seems that we
just should add a .Error() whenever we get an error in OpenStack.
// Rickard
On 16 October 2017 at 10:08, Jean-Gert Nesselbosch
wrote:
>
> hello everybody,
>
> I'm evaluating Packer (in an openstack-context) for my compa
Haven't done that my self. One way of accomplice this is to use as a file
provisioner to download a status file and wrap packer with a script that
determines if the build was failed or aborted due to no updates.
On 29 September 2017 at 22:09, 'Sean O'Reilly' via Packer <
packer-tool@googlegroups
You need to wait for cloud-init to finish, see
https://www.digitalocean.com/community/questions/how-to-make-sure-that-cloud-init-finished-running
Cloud-init is best suited for setting up access related things (which might
be impossible with a provisioner). Cloud-init only reports errors to a log
f
upposed to
> be used as part of CI/CD please set me straight... but we'd really hate to
> stop using it because Packer really is such a great tool that solves all of
> our problems save this one minor disconnect.
>
> Thanks,
> Jon A
>
>
>
> On Tuesday, October 10,
It's really hard for me to help much more without a full repro case. The
general advice I can give is to use a existing ssh_keypair that you have
access to and run with --on-error=ask and wait for ansible to fail, jump
into the instance and troubleshoot (check syslog and auth logs).
Just to give y
Yes it works nicely. If you provide some details and explain how you
troubleshooted this we might be able to help you out.
On Oct 11, 2017 18:18, "Paul Chen" wrote:
> I try to run Packer on a CentOS VM spawn from OpenStack but it hangs w/o
> error messages. Is it ever possible to use Packer to b
Then add - as extra args to ansible and rerun packer with PACKER_LOG=1.
If you don't find the problem try posting the log here.
On Oct 11, 2017 07:54, "rick turner" wrote:
> But still not working with packer
>
> On Wednesday, October 11, 2017 at 10:57:31 AM UTC+5:30, rick turner wrote:
>>
>>
Seems like your user is not allowed to do sudo without password.
On 11 October 2017 at 06:00, rick turner
wrote:
> Description:
> -
> I am trying to execute a ansible playbook in my base.json but it's
> throwing me permission related errors:
>
> Error:
> ---
> amazon-ebs: fat
The situation is the same. It's not possible and it steps away a bit from
the purpose of Packer.
Regarding the null builder it's main purpose is to speed up development of
provisioners and their script, since you can quickly iterate over them
without the need of launching and destroying instances.
Hi,
A LXC container doesn't run sshd. To run ansible against one you need to
use a custom connection plugin like this 1). You can compare with the
custom connection plugin used for WinRM which is in the docs.
When you have a working example it would be awesome if you could post it
here.
1) http
I can confirm that that is how it works.
On Oct 6, 2017 14:08, "Alvaro Miranda Aguilera" wrote:
> Hello
>
> I am not 100% sure, but seems is the url of the image you already have in
> the openstack.
>
> Checking this, looks like it use the id to find the url, or you can just
> use the url.
>
> h
You have to define *all* user variables is the "variables" block.
"output_directory": "{{user `output_directory`}}",
but:
"variables" : {
"ssh_username": "root",
"ssh_password": "solidfire",
"headless": "true"
},
On 27 September 2017 at 23:06, wrote:
Hi,
That totally makes sense, or at least to create VM templates for those VMs.
I don't have any ESXi templates at hand but a quick search on github found
me these that looks right:
https://github.com/sdorsett/packer-centos7-esxi
If you start by getting them to run on your system, and if you have
buntu builds but not for CentOS
>
> I got the previous log messages from /var/log/messages but that's all
> there was relating to "cloud"
>
> On Tuesday, September 19, 2017 at 2:45:10 PM UTC+1, Rickard von Essen
> wrote:
>>
>> So then the problem is that the
e:
> Yes, that was the link I used to get the public key into my
> authorized_keys file
>
> On Tuesday, September 19, 2017 at 2:31:09 PM UTC+1, Rickard von Essen
> wrote:
>>
>> Does curl http://192.168.1.1/latest/meta-data/public-keys give you the
>> pub key for
2.168.1.1
> Sep 19 13:15:10 localhost cloud: Changing password ...
> Sep 19 13:15:10 localhost cloud: Sending acknowledgment to password
> server at 192.168.1.1
>
> Don't see anything in our external logging either that sheds any light
>
>
> On Tuesday, September 19, 20
I would start by creating a template where you have some fixed credentials
that you can use for debugging (or use the console). Run packer build
--on-error=ask, so when it fails you have time to troubleshoot.
Jump into the node to debug if cloud-init is functioning as it should, see
http://cloudin
Maybe https://github.com/hashicorp/packer/issues/5347
On 19 September 2017 at 10:52, 'Chris Nappin' via Packer <
packer-tool@googlegroups.com> wrote:
> Hi,
>
> I have a packer template written about 9 months ago on a previous Packer
> version (v0.7, that worked fine) that isn't working with Pac
I was trying to repo with your template and realises that you are missing
user_data_file *)
https://www.packer.io/docs/builders/amazon-ebs.html#user_data_file
On 13 September 2017 at 22:46, TJ Biddle wrote:
> Which looks like setting a password from the get-go fails as well: "The
> password doe
quot;source_template": "{{user `source_template`}}",
>> "template_os": "{{user `template_os`}}",
>> "zone": "{{user `zone`}}",
>> "instance_name": "{{ user `instance_name`}}-{{timestamp}}",
>
Are you running the latest version (released today). I have been working on
some cloudstack improvements that was released now some of them involved
improved ssh support.
/ Rickard
On Sep 13, 2017 16:38, "David Curran" wrote:
I'm trying to automate the build process for templates and have run i
This works
https://github.com/Diabol/packer-demo/blob/master/windows-cloud.json
using Windows Server 2016
On 13 September 2017 at 15:11, Khalid Hosein wrote:
> amazon-ebs
>
> On Tuesday, September 12, 2017 at 3:16:08 PM UTC-4, Rickard von Essen
> wrote:
>>
>> Wh
601 - 700 of 1053 matches
Mail list logo
