Did you try the partition mbr option?
On May 23, 2017 04:11, "Bruce Edge" wrote:
> Confirming this is still a problem with 20170516 images.
>
> --
> This mailing list is governed under the HashiCorp Community Guidelines -
> https://www.hashicorp.com/community-guidelines.html. Behavior in
> viola
04-lts --image-project ubuntu-os-cloud \
> --maintenance-policy TERMINATE --restart-on-failure
>
>
>
>
>
>
> On Monday, May 22, 2017 at 1:07:18 PM UTC+3, Rickard von Essen wrote:
>>
>> That will definitely not work.
>>
>> Could you provide the a gclo
That will definitely not work.
Could you provide the a gcloud example of how to create a instance with
GPU? I didn't find how to this in the go Google cloud lib by quickly
browsing it.
/ Rickard
On May 22, 2017 10:54, wrote:
> Hi,
> I'm trying to create image with googl's new GPU,
> I tried to
See my answer to your stackoverflow question
http://stackoverflow.com/a/44094893/226174
/ Rickard
On May 21, 2017 10:22, "Adron Hall" wrote:
> I'm trying to run a Packer template to build a basic AWS EBS based
> instance. I keep getting the following error however.
>
> ==> amazon-ebs: Error que
And you can use process substitution:
packer build -var-fil=<(var_fn) template.json
See http://www.tldp.org/LDP/abs/html/process-sub.html
/ Rickard
On May 17, 2017 10:53 PM, "Matthew Hooker" wrote:
> Hi Andrew,
>
> Glad to hear our tools are working out for you.
>
> We don't allow you to pass
You should add: "disable_sudo": true, to salt-masterless and it will be
alright, see
https://www.packer.io/docs/provisioners/salt-masterless.html#disable_sudo
// Rickard
On 11 May 2017 at 21:45, Adam wrote:
> Just trying to test out Packer builds. My environment:
>
> Packer box: CentOS 6.8 Vir
Just unset tools_upload_flavor which will disable upload. Since you install
open-vm-tools you don't need it.
See https://www.packer.io/docs/builders/vmware-iso.html#tools_upload_flavor
On 10 May 2017 at 21:25, Steven Langlois wrote:
> I'm building a VMWare CentOS7.3 image on an ESX server runni
Yes it is, see https://www.packer.io/docs/builders/qemu.html#iso_url
On May 8, 2017 9:32 PM, "esio" wrote:
> Hello.
>
> Is it possible to use QEMU builder but without install OS from scrath but
> use prepared qcow2 or raw image instead? Something like in VirtualBox
> builder?
>
> For me OS insta
Could you add some more context, full template, scripts and full output
etc.
On May 4, 2017 2:29 AM, "Farye Nwede" wrote:
> I'm not sure if this is the best place for this. I'm encountering this
> error. My host system is Ubuntu 14.04. The system that I'm trying to build
> is Ubuntu 16.04. Any i
Check
On May 1, 2017 2:58 PM, "Luke Farnell" wrote:
> My first comment contains a link to the Github Repo with all the scripts
> and templates
>
> On Monday, May 1, 2017 at 2:04:04 AM UTC-4, Rickard von Essen wrote:
>>
>> Please provide templates and scripts a
Without digging deeper into this it seems like a bug. Can you create a
minimal example to reproduce and open an issue on github.
/ Rickard
On May 1, 2017 12:57 PM, "Kristopher Landon" wrote:
> I should add this is windows 2012 R2 with latest patches and Packer V 1.0.0
>
> On Monday, May 1, 2017
command line ? I have googled but couldn't find a way to
> create vm.
>
>
> On Apr 27, 2017, at 3:16 PM, Rickard von Essen <
> rickard.von.es...@gmail.com> wrote:
>
> Most likely you can't run vmware workstation in a docker container. Try
> and run your containe
Please provide templates and scripts as text (attachments).
On Apr 30, 2017 9:49 PM, "Luke Farnell" wrote:
> Photo's are attached.
>
> Luke
>
> On Sunday, April 30, 2017 at 2:59:47 PM UTC-4, Luke Farnell wrote:
>>
>> Currently I am trying to write a PR for including the IP address of the
>> host
Ram, Dan is building on a remote vmware ESXi only running packer inside a
container, not vmware.
On Apr 27, 2017 23:38, "Ram Maram" wrote:
> Hi Dan,
>
> We are trying to build vmware packer images inside container, we have
> created a docker image with vmware workstation, vmware Vix and packer.
38).
>
> -- RGL
>
> On Thursday, April 27, 2017 at 4:26:26 PM UTC+1, Rickard von Essen wrote:
>>
>> This is a concurrency error in your code.
>>
>> Somewhere in your code you write to the cancel channel after you have
>> closed it. Try adding a printo
VirtualBoxs network stack is a bit odd. The default NAT network doesn't
allow it possible for the host to access the guest on it's IP. Instead it
requires that port forwarding is setup to the ssh port on the guest. This
is done by default by Packer and that's why it ssh to 127.0.0.1 on some
random
Most likely you can't run vmware workstation in a docker container. Try and
run your container and enter it and create a VM manually. Does that work?
On Apr 27, 2017 20:54, "Ram Maram" wrote:
> Hi,
>
>
> I'm trying to build vmware packer images inside docker container (redhat
> 7), but I'm getti
It's just a bundled default docker registry (
https://github.com/docker/distribution/blob/master/README.md) so it should
be fine.
On Apr 27, 2017 18:14, "'Mitchell Cooper' via Packer" <
packer-tool@googlegroups.com> wrote:
> Does anyone know if packer supports using the docker-push post processor
This is a concurrency error in your code.
Somewhere in your code you write to the cancel channel after you have
closed it. Try adding a printout before you close and write to it.
// Rickard
On 27 April 2017 at 13:34, Rui Lopes wrote:
> That's what I did. I just modified it slightly to remove s
You need to run amazon-chroot from an AWS EC2 instance, the only way to run
that from travis is to issue a ssh command from travis that runs the build
on an instance you have running. (I'm not sure you can ssh out from Travis.)
// Rickard
On 24 April 2017 at 19:26, Stephen wrote:
> Hi,
>
> I've
.redhat.com/pulp/repos//content/dist/rhel/rhui/server/
> 7/%24releasever/x86_64/extras/os/repodata/repomd.xml: [Errno 14] HTTPS
> Error 404 - Not Found
> amazon-ebs: Trying other mirror.
> (...snip...)
> ==> amazon-ebs: Terminating the source AWS instance...
> ==> amazon-e
Haven't done anything similar, but a few comments:
This looks wrong:
"sudo mkfs.ext4 /dev/xvdf",
"sudo mv /home/ec2-user /home/ec2-user.save",
"sudo mkdir -p /home",
"sudo shopt -s dotglob; cp -R /home/ec2-user.save/
/home/ec2-user/",
hanks,
> rg
>
> On Friday, April 21, 2017 at 3:53:53 PM UTC-4, Rickard von Essen wrote:
>>
>> Run a shell provisioner that saves the IP in a file. Use a file
>> provisioner to download it and read it in the shell-local script.
>>
>> Or use some aws tag specific
Run a shell provisioner that saves the IP in a file. Use a file provisioner
to download it and read it in the shell-local script.
Or use some aws tag specific to this build and run the aws cli to find the
IP of that instance.
On Apr 21, 2017 21:20, wrote:
> Hi everyone,
>
> tl;dr: I'm looking f
Hi,
The GCE API for creating an Image[1] only allows one disk when creating an
Image. The API for launching an Instance[2] allows a list of disks but
since only one can be used in the resulting Image I we haven't implemented
support for that. And no one have requested support for it.
// Rickard
There is no direct support for this if you need it to span multiple
provisioners.
I would sugest that you write some metadata to a file and use the file
provisioner to download it and wrap your packer run in a script that uses
that to unregister the subscription after packer has completed regardle
I think you shouldn't use launch_block_device_mappings that will replace
the (source) ami snapshot with a 60 GB empty ebs volume.
On Apr 11, 2017 22:16, "Shalin Patel" wrote:
> Hi
> I am trying to create new AMI based off official coreos AMI. I am trying
> to expand root volume to bigger size.
mages for them to work on one or the other
> platform. So why exactly can't I start from a single source image and get
> different outputs still puzzles me.
>
> Best regards,
>
> Vincenzo
>
> On Tuesday, April 11, 2017 at 1:14:08 PM UTC+2, Rickard von Essen wrote:
>>
You can often use a local builder for example qemu to build an image and
then post-process it to upload it to a cloud provider, but that is a _very_
advanced use case and if you don't get all the specific details for _that_
cloud right you image will be dead in the water. This is something I don't
It runs with whatever you have in "ssh_username".
On 6 April 2017 at 22:04, Isaiah Frantz wrote:
> ok, lookes like it runs post boot operations as the vagrant user, good to
> know. Please update the docs to indicate this.
>
> slipped this in the top of my script and sat through a 5th kickstart c
It's not realy the purpose of packer to create ISO images. Check mkisofs
instead. If you want to distribute a VM template packer is the right tool.
On 6 April 2017 at 11:06, JAI RAJ wrote:
> Rickard,
> if have an iso file in which required applications are installed in it in
> my virtual box.No
No
On Apr 6, 2017 7:57 AM, "JAI RAJ" wrote:
using packer we can process a json file and generate appropriate templates
like iso,ami files.
Is it possible to generate json file from given iso or ami files.?
--
This mailing list is governed under the HashiCorp Community Guidelines -
https://www
Awesome!
Four years after Mitchells initial commit. Thanks to lots of hard work from
HashiCorp and the community.
// Rickard von Essen
On 4 April 2017 at 23:22, Vasiliy Tolstov wrote:
> 2017-04-04 23:55 GMT+03:00 Matthew Hooker :
> > Hi all,
> >
> > Today we released Pack
What's the content of install-hdfs.sh ? The exit code is simply from that
script.
On 4 April 2017 at 16:55, Laeeq Ahmed wrote:
> Complete log here.
>
> https://pastebin.com/v6JBh9R6
>
>
> On Tuesday, 4 April 2017 16:50:27 UTC+2, Laeeq Ahmed wrote:
>>
>> Hi,
>>
>> What can be the reason for exit
x27;t know about per-instance. But still is there a way to
>> make the script not run at all when launched by packer?
>>
>> Cheers
>>
>> On 29 March 2017 at 14:33, Rickard von Essen > > wrote:
>>
>>> Just put it in per-instance, then cloud-init wi
You are running with -debug, don't. If you need to troubleshoot run with
-on-error=ask.
What are you trying to build? Your shutdown_command is strange? The command
is issued on the VM that is built, I would expect something like /sbin/halt
-h -p etc.
If you are just starting with packer I really
Check your path, at least Fedora includes a tool (crackerlib) which
includes a binary also called packer.
/ Rickard
On Mar 31, 2017 05:36, "Saed Alavinia" wrote:
> Hello,
>
> I am trying to run packer_0.12.3_linux_amd64 on my RHEL for Developers
> Image. But when I unzip packer_0.12.3_linux_amd
Preemptible is just this option in GCE
https://cloud.google.com/compute/docs/instances/preemptible. It allow you
VM to be stoped if someone else needs the resources. This gives a lover
price per minute but risk that your VM is interrupted.
If you want to launch a VM of the newly built image you ha
1.0 has already branched off https://github.com/mitchellh/packer/tree/v1.0
So masters next build is 1.1
// Rickard
On 30 March 2017 at 07:38, Matt Black wrote:
> Hi Matt,
>
> How comes the new version is reporting 1.1.0? I thought I'd messed
> something up building a testing version on my syst
Just install sudo in the docker image in a shell provisioner with "only": [
"docker" ].
On Mar 29, 2017 6:22 PM, "Kevin Pauli" wrote:
> Newbie here, trying to figure out how to build both a docker image and an
> amazon-ebs ami from the same script. Here I have stripped it down to the
> essentia
n 29 March 2017 at 14:33, Rickard von Essen
> wrote:
>
>> Just put it in per-instance, then cloud-init will see to that it runes
>> once per instance.
>>
>> https://cloudinit.readthedocs.io/en/latest/topics/modules.ht
>> ml#scripts-per-instance
>>
>>
Just put it in per-instance, then cloud-init will see to that it runes once
per instance.
https://cloudinit.readthedocs.io/en/latest/topics/modules.html#scripts-per-instance
On 29 March 2017 at 13:55, Michael Seiwald
wrote:
> Hello,
>
> I'm building an AMI that already includes a script that is
ther Packer should not change this hardcoded
> configuration.
>
> Dne úterý 28. března 2017 22:59:32 UTC+3 Rickard von Essen napsal(a):
>>
>> The bottom note on https://www.packer.io/docs/builders/openstack.html
>> suggest that OS_TENANT_NAME should be mcafee-siem and not soc.
&g
The bottom note on https://www.packer.io/docs/builders/openstack.html
suggest that OS_TENANT_NAME should be mcafee-siem and not soc.
On 28 March 2017 at 21:28, Rickard von Essen
wrote:
> That is interesting.
>
> I assume that this is a private OpenStack installation.
>
> Did y
>
> export OS_PROJECT_NAME=mcafee-siem
> export OS_USER_DOMAIN_NAME=soc
> export OS_PROJECT_DOMAIN_NAME=soc
> export OS_TENANT_NAME=soc
>
> But in Packer I cannot change project name, in the docummentation is
> written that PROJECT_NAME is same as TENANT_NAME or similar.
>
&g
penStack builder and CentOS images ?
>
> El lunes, 27 de marzo de 2017, 15:31:38 (UTC+2), Lorena Lobato escribió:
>>
>> I will give a try and let you know, thanks
>>
>> El viernes, 24 de marzo de 2017, 21:45:16 (UTC+1), Rickard von Essen
>> escribió:
>&g
"communicator": "winrm",
"winrm_username": "packer",
"winrm_insecure": true,
"winrm_timeout": "300m",
"winrm_use_ssl": true,
"metadata": { "windows-startup-script-cmd"
Actually after some googling I think it kind of works but then we should
have an example in the docs.
On 25 March 2017 at 17:30, Sachin wrote:
>
> Thank you Rickard,
>
> I will report bug.
>
> On Saturday, March 25, 2017 at 12:10:56 PM UTC-4, Rickard von Essen wrote:
>&
Did you get this to work for you?
// Rickard
On 22 February 2017 at 16:32, Rickard von Essen wrote:
> I think you might be missing to allow the firewall traffic on port:
> tcp/5986 from anywhere. By default it allows SSH and RDP but not WinRM.
> Except for that everything looks fin
07:55:19 [ERROR] connection
>>> error: http response error: 401 - invalid content type
>>> 2017/03/25 07:55:19 packer.exe: 2017/03/25 07:55:19 [ERROR] WinRM
>>> connection err: http response error: 401 - invalid content type
>>>
>>> On Saturday, Marc
By default the GCE firewall doen't allow WinRM connections (from the
outside). Have you opened up port 5986 in the firewall? If that doesn't
help can you run PACKER_LOG=1 packer build template.json and attach/gist
the output?
// Rickard
On 25 March 2017 at 12:02, Sachin wrote:
> *Hello,*
>
> *I
bato wrote:
>
> Same case but with Linux images. When I reboot the machine in the
> provisioner stops without any error and skipping the post commands.
>
> Is there anyway to avoid this? Need to reboot the machine during
> provisioning time to make effective some puppet chang
me case but with Linux images. When I reboot the machine in the
>> provisioner stops without any error and skipping the post commands.
>>
>> Is there anyway to avoid this? Need to reboot the machine during
>> provisioning time to make effective some puppet changes.
>&
Check how other do it:
https://github.com/chef/bento
https://github.com/boxcutter
And the documentation:
https://wiki.debian.org/DebianInstaller/Preseed
https://www.centos.org/docs/5/html/Installation_Guide-en-US/ch-kickstart2.html
On 24 March 2017 at 11:43, Shiraaz Moollatjie wrote:
> Hi,
>
>
It's hard to give any further debugging advice unless you can capture the
traffic or access the logs on keystone and see why your auth is denied.
On 23 March 2017 at 11:07, wrote:
> http://pastebin.com/A2Wx9whX Here is the full log.
>
> Dne čtvrtek 23. března 2017 11:15:14 UTC+2 peet.p...@gmail.
Shutdown is initiated by an api towards openstack.
On Mar 23, 2017 23:46, "Blake Garner" wrote:
> I'm starting to work with the OpenStack builder and noticed that no
> shutdown_command or shutdown_timeout exist. Am I missing something here?
> Using the shutdown commands is how I trigger windows
: "vagrant"
> }
> }
>
>
> If i run this without elevated user/pass, the loop runs fine for over 2
> hours. Googling what this option does show it is creating a windows
> scheduled task and running as an elevated user. Is packer setting a limit
> for how long t
You couldn't connect to the instance, but without more details it's hard to
say why. Could you provide your template?
On Mar 23, 2017 20:11, wrote:
> I am getting below error. Can you please suggest me what to do to fox this
> issue.
>
> amazon-ebs output will be in this color.
>
> ==> amazon-eb
Looks good to me. Can you run with PACKER_LOG=1 packer build template.json
and pastbin the output.
On 23 March 2017 at 10:15, wrote:
> Hello, I am trying to automate some deployment to an Openstack-based
> cloud. It's based on Openstack, but there might be some minor differences
> between those.
Script exited with non-zero exit status: 259.
This means that your script returned a non 0 exit code (259). Without
knowing much about Windows this seams to be a restriction set there.
Googling gives alot of similar issues for different softwares written in
different languages. E.g.
https://githu
Upgrade packer to 0.12.3 and report back again.
On Mar 20, 2017 6:38 PM, "Steven Langlois" wrote:
> Sorry, should have mentioned that I am running packer 0.10.2.
>
>
> --
> This mailing list is governed under the HashiCorp Community Guidelines -
> https://www.hashicorp.com/community-guidelines.h
or this platform...
> find: `./pkg/linux_amd64': No such file or directory
>
> ==> Results:
> total 0
>
> Which I think means it failed to copy the binary in the end? There's
> nothing in the $GOPATH/bin folder and nothing in the packer/bin folder
> either
&
gt;
> Which is progress! So now I need to figure out why govendor isn't a
> command.
>
> On Monday, 20 March 2017 22:03:03 UTC+13, Rickard von Essen wrote:
>>
>> I tested my docker command and it's not completely accurate. This works
>> on my Mac:
>&g
, it doesn't seem to
> produce a package either.
>
>
> On Monday, 20 March 2017 21:36:21 UTC+13, Rickard von Essen wrote:
>>
>> It should work to build on Windows, unfortunately I can't really help
>> with that. We build each commit in appveyor and that is green so
oping to avoid having to create a Linux machine to build
> Packer on, but if that's the only way then I guess I'll have to make one.
>
> On Sun, Mar 19, 2017 at 9:48 PM, Rickard von Essen <
> rickard.von.es...@gmail.com> wrote:
>
>> If you have Docker for Wind
If you have Docker for Windows the simplest way might be cross compile in a
container.
Go to the directory where you have packer checked out and run something
like:
docker run -v %PWD%:/usr/src/packer -w /usr/src/packer golang:1.8 bash -c
XC_ARCH=amd64 XC_OS=windows make dev
This should give you
Could you provide the values for your user vars and the output when you run
ansible with -.
On Mar 7, 2017 10:53 PM, "Alvaro Miranda Aguilera"
wrote:
> I see
>
> thanks for the explanation. missed that.
>
>
>
> On Tue, Mar 7, 2017 at 10:49 PM, Chris Stevens
> wrote:
>
>> I'm temporarily usi
succeeds)?
>
> I added: "virtualbox_version_file":"", to my .json, did not help, still
> uploading. Any idea?
>
>
> Dne pátek 3. března 2017 12:50:15 UTC+2 Rickard von Essen napsal(a):
>>
>> Try setting ssh_skip_nat_mapping to true and run packer wit
Try setting ssh_skip_nat_mapping to true and run packer with:
PACKER_LOG=1 packer build -on-error=ask template.json
You will get a log of debug info and if it fails you can use the console to
do additional debugging.
https://www.packer.io/docs/builders/virtualbox-iso.html#ssh_skip_nat_mapping
On
I see that you have a kickstart (.ks) file in the http directory, but if
this is based on Debian it probably uses preseed. See
https://github.com/chef/bento/blob/master/http/debian-8/preseed.cfg for an
example. Also check the boot_command for the corresponding Debian vagrant
box in the root of the
controller is not a known dns name in the dns server you are using. Ask
your OpenStack admin if you need to use a specific dns server or if there
is a misconfiguration on their end.
On Feb 28, 2017 16:07, "balubhai555" wrote:
> Hello Team,
>
> Currently, I am new to Openstack. I want to trying t
>
> Thanks!
>
> On Monday, February 27, 2017 at 4:05:08 PM UTC-5, Rickard von Essen wrote:
>>
>> Could you provide some more context here? Give us a example packer
>> template (.json) and what do you mean with a container ? a docker
>> container, if so can you sp
Could you provide some more context here? Give us a example packer template
(.json) and what do you mean with a container ? a docker container, if so
can you specify the command you started it with.
It might just be that env var USER is not set. If I run: docker run -it
ubuntu:trusty bash USER is
Try dropping these two lines:
"user": "ec2-user",
"sftp_command": "/usr/libexec/openssh/sftp-server -e"
On 27 February 2017 at 14:45, Tony Mills
wrote:
> Hi All,
>
> I am trying to get packer running ansible and am wondering if there is a
> fix for the following problem:-
>
> packer itself is
Use:
"ansible_env_vars": [
"ANSIBLE_ROLES_PATH=/some_custom_path/:$ANSIBLE_ROLES_PATH" ]
See https://www.packer.io/docs/provisioners/ansible.html#ansible_env_vars
On 27 February 2017 at 12:29, Yasin Amadmia <
yasin.amad...@travisperkins.co.uk> wrote:
> Hi,
>
> Is it possible to provide a `roles
Hi Kevin,
I think this is definitely a missing feature that we would like to have in
Packer (builder/vmware-vmx with support for remote ESXi). Just 3 days ago
the was a PR opened which adds this support,
https://github.com/mitchellh/packer/pull/4591 . It might not make it to the
1.0 release so I r
Also if you set DIGITALOCEAN_API_TOKEN packer will use that. I.e. export
DIGITALOCEAN_API_TOKEN=$DO_API_TOKEN
If that doesn't work it would be good I'd you attach a log from
PACKER_LOG=1 packer build template.json, just redact all sensitive
information.
On Feb 25, 2017 5:27 PM, "Alvaro Miranda Ag
IMHO it's best to wrapp packer for these kind of things. Use the manifests
post-processor to write out information about the resulting artifact, if
needed.
On Feb 24, 2017 4:50 PM, "Basil Peace" wrote:
> On Wednesday, August 10, 2016 at 8:15:53 PM UTC+3, Yuriy Medvedev wrote:
>>
>> Anybody try a
Looks like you don't have a proper end double quote (zoom in).
"PACKER_BOX_NAME": "ubuntu-16.04.2-server-i386”,
On 24 February 2017 at 12:24, Himanshu Pant wrote:
> I am trying to get an application server up and running. I changed the
> Ubuntu distribution in the template file to point to late
_file set to the location of the key and unfortunately
> it still failed (see log file
> <https://gist.github.com/newtonne/014629e976326c7cd6365b54b82dfb48>).
> Note that the build works for aws-key even if I don't
> set ssh_private_key_file.
>
> On Wednesday, February
When you set ssh_keypair_name to server_key you also have to update
ssh_private_key_file. But from the logs it looks like you haven't done that.
On 22 February 2017 at 20:11, notmyrealname
wrote:
> Packer v0.12.2
> Mac OS X 10.11.6
>
> As suggested, I've moved this from the Github issues page to
I think you might be missing to allow the firewall traffic on port:
tcp/5986 from anywhere. By default it allows SSH and RDP but not WinRM.
Except for that everything looks fine with me. The user should be created
and the password is retrieved over the API in the same way as the gcloud
tool do.
ht
There is no api, and I don't expect there to be one since this is more of a
build tool than Nomad, Consul, Vault, etc. Also see
https://github.com/mitchellh/packer/issues/1840
But it should be fairly easy to run packer by OS exec. I think the
difficult part will be how to handle config. I think yo
Could you describe your setup you are running packer in and maybe provide
your template/scripts?
On 14 February 2017 at 14:54, wrote:
> Hi,
> many time, my packer build goes in error because a file upload
> provisionners step failed
> Here is the log :
>
> 2017/02/14 14:44:05 ui: ==> vmware-is
Packer is looking for dhcp. Please see error
> attached
>
> Thanks
> ShyanOn Tuesday, 14 February 2017 16:07:00 UTC+5:30, Rickard von Essen
> wrote:
> > "With mentioning ssh_host, ssh_port packer is still looking for Dhcp"
> Sorry I don't follow this.
> >
"With mentioning ssh_host, ssh_port packer is still looking for Dhcp" Sorry
I don't follow this.
On 14 February 2017 at 10:27, Shyam J wrote:
> With mentioning ssh_host, ssh_port packer is still looking for Dhcp. Am I
> missing any other parameter?
>
> Thanks
> Shyam
>
> --
> This mailing list i
didn't complete successfully and had errors:
> --> virtualbox-iso: unexpected EOF
>
> ==> Builds finished but no artifacts were created.
> 2017/02/13 21:25:13 /Users/dan/bin/packer: plugin process exited
> 2017/02/13 21:25:13 /Users/dan/bin/packer: plugin process exited
It should work even with static IP's. Worst case you need to set ssh_host.
On Feb 13, 2017 11:13, "Shyam J" wrote:
Thanks a lot. It worked
After that I see it is looking for Dhcp, but I don't have a Dhcp in that
network and configured in kick-start static IP details.
Is packet co figured to use
Packer doesn't work with dvs (distributed virtual switches), I recommend
unsetting those in the vmx_data section and add "ethernet0.networkName":
"VM Network" or whatever your ESXi network is called.
/ Rickard
On Feb 13, 2017 7:50 AM, "Shyam J" wrote:
> ,
>
> I am trying to create template in a
Rerun with PACKER_LOG=1 and post the out here.
On Feb 13, 2017 00:40, "Dan Linder" wrote:
> I am trying to build a RHEL 6.8 server VM from ISO, but it fails on the
> "StepUploadVersion":
>
> *==> virtualbox-iso: Waiting for SSH to become available...*
> *==> virtualbox-iso: Connected to SSH!*
>
It's not possible to interpolate values in user variables.
On Feb 9, 2017 5:52 PM, "Kevin Moser" wrote:
> Is there a way to have a variable reference another variable, but still
> allow it to be overriden in a var file or var on the cmd line?
>
> Example:
>
> {
> "variables":{
> "foo": "",
What's the output of: which -a packer
?
On Feb 7, 2017 7:07 PM, "Srinivas Reddy" wrote:
> I installed packer and set the path in ./bashrc_profile
> PATH=$PATH:/usr/local
>
> under /usr/local/
>
> libexec *packer* sbin share src
>
> I have the packer installed . When i run packer it keeps ha
Disclaimer: I've never used azure-arm, but if you don't get any answer here
I think it's valid to open a Github issue and request a "feature" or a
clarification in the docs about this.
// Rickard von Essen
On 6 February 2017 at 13:57, Johan Elmerfjord wrote:
> Hi,
The docs contain all information, most template variables and functions are
local per configuration option and are described under each
builder/provisioner/etc.
If you are just looking for some great examples of how to build vagrant
boxes see https://github.com/chef/bento
If you are looking for s
Use the -on-error=ask argument to packer build to pause the build when a
provisioner fails.
On Feb 2, 2017 23:37, "John C" wrote:
> I'm looking for a way to timeout a specific provisioner step, and if the
> step fails, either move on and continue the build, or at least stop so it
> can be debugg
To get started with building debian from ISO on VirtualBox check
https://github.com/chef/bento/, just remove the vagrant post-processor.
For KVM and Xen you should use the qemu builder, there is code in bento
also for that.
For DO you have to start with the official Debian droplet and apply your
No, sshd is installed and enabled by default.
The best way to troubleshoot this is by setting an incredible high
ssh_timeout, say 12h and use headless false. After that start packer and
use the VirtualBox ui to access the console where you can login as root.
Now you can troubleshoot the network fr
pastebin.com/c3Eu9Aat
>
> or like this:
> http://pastebin.com/7UYizYDq
>
> Which means that by the error code the server closed the connection.
> No clue why it happens,
>
> Any help would be appreciated
> Ido.
>
> On Thursday, January 19, 2017 at 9:28:25 PM UTC
Your kickstart file is called kickstarter.cfg but your boot_command
references kickstart.cfg. Also you should not use remote_host since your
kickstart uses dhcp. Let packer handle networking as much as possible.
On Jan 19, 2017 9:39 PM, "Scott Bailey" wrote:
> Hi,
>
> I am new to all this kinda
It's not very useful when everything is a user variable. Can supply the
values of those, say which distribution the source ami is and run with
`PACKER_LOG=1 packer build template.json` and attach the full log.
On Jan 19, 2017 6:29 PM, wrote:
> Hi all,
> Trying to use amazon-ebs builder with ansi
801 - 900 of 1053 matches
Mail list logo
