Bug#877451: packer: New Release of Packer (v1.1.0)

[JD Friedrikson]

Package: packer
Version: 0.10.2+dfsg-6
Severity: wishlist
Tags: upstream

Hello,

Packer recently released v1.1.0:
https://github.com/hashicorp/packer/blob/v1.1.0/CHANGELOG.md

One of the really nice things about this version is that it now supports
builds for `lxc` and `lxd`. Is work already being done to package the
new version? If not, is there anything I can do to help the process
along?

Regards,
JD


-- System Information:
Debian Release: 9.1
  APT prefers stable
  APT policy: (600, 'stable'), (500, 'stable'), (300, 'testing'), (1, 
'unstable')
Architecture: amd64 (x86_64)

Kernel: Linux 4.9.0-3-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_US.utf8, LC_CTYPE=en_US.utf8 (charmap=UTF-8), 
LANGUAGE=en_US.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages packer depends on:
ii  libc6  2.24-11+deb9u1

Versions of packages packer recommends:
pn  docker.io  
ii  qemu   1:2.8+dfsg-6+deb9u2

Versions of packages packer suggests:
ii  ansible  2.3.2.0+dfsg-1
pn  chef 

-- no debconf information

Bug#874365: ruby-gdk3: Typelib file for namespace 'Gdk' (any version) not found (GLib::Error)

[JD Friedrikson]

Package: ruby-gdk3
Version: 3.1.0-1
Severity: grave
Justification: renders package unusable

Hello there,

I was attempting to make a small thing with gtk3 (via ruby) this morning
and ran into an error with one of its dependencies:

```
/usr/lib/ruby/vendor_ruby/gobject-introspection/loader.rb:37:in `require': 
Typelib file for namespace 'Gdk' (any version) not found (GLib::Error)
from /usr/lib/ruby/vendor_ruby/gobject-introspection/loader.rb:37:in 
`load'
from /usr/lib/ruby/vendor_ruby/gdk3.rb:56:in `init'
from /usr/lib/ruby/vendor_ruby/gtk3.rb:64:in `init'
from /usr/lib/ruby/vendor_ruby/gtk3.rb:40:in `const_missing'
from ./pomodoro:5:in `'
```

It seems to fail with some pretty basic usage:

```
#!/usr/bin/env ruby

require 'gtk3'

window = Gtk::Window.new('test')
window.show
Gtk.init
Gtk.main
```

As the error refers missing files, I figure this would be a bug related
to the package. Please let me know if I'm mistaken.

Cheers,
JD

-- System Information:
Debian Release: 9.1
  APT prefers stable
  APT policy: (600, 'stable'), (500, 'stable'), (300, 'testing'), (1, 
'unstable')
Architecture: amd64 (x86_64)

Kernel: Linux 4.9.0-3-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_US.utf8, LC_CTYPE=en_US.utf8 (charmap=UTF-8), 
LANGUAGE=en_US.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages ruby-gdk3 depends on:
ii  ruby1:2.3.3
ii  ruby-atk3.1.0-1
ii  ruby-cairo-gobject  3.1.0-1
ii  ruby-gdk-pixbuf23.1.0-1
ii  ruby-gobject-introspection  3.1.0-1
ii  ruby-pango  3.1.0-1

ruby-gdk3 recommends no packages.

ruby-gdk3 suggests no packages.

-- no debconf information

Bug#874194: ansible: --force doesn't replace unversioned roles

[JD Friedrikson]

Package: ansible
Version: 2.3.1.0+dfsg-2
Severity: normal

Hello!

Currently ansible doesn't allow for updating roles via forced installs.
This affects all versions packaged by Debian. This is considered a bug
that was fixed upstream:
https://github.com/ansible/galaxy-issues/issues/249
https://github.com/ansible/ansible/pull/23391

Here's the patch:

```
commit 78836ec0b9719a62fa0f8619707a9f411ed4a4f0
Author: Michael Scherer 
Date:   Fri Apr 21 13:40:47 2017 +0200

Fix --force for unversionned requirements (#23391)

In current stable (2.2), ansible galaxy install --force do erase
a role, even if the version is not set. This commit should restore
that specific behavior, in accordance to people reports:
  https://github.com/ansible/ansible/issues/11266#issuecomment-273801480

It was also the behavior planned in the initial discussion:
"if you're not fixing versions in your roles file, then it's fine
to expect that the role will be reinstalled each time you run
ansible-galaxy install.", cf https://github.com/ansible/ansible/pull/12904

diff --git a/lib/ansible/cli/galaxy.py b/lib/ansible/cli/galaxy.py
index 8554b68c4..6baf7affb 100644
--- a/lib/ansible/cli/galaxy.py
+++ b/lib/ansible/cli/galaxy.py
@@ -389,8 +389,9 @@ class GalaxyCLI(CLI):
 (role.name, 
role.install_info['version'], role.version or "unspecified"))
 continue
 else:
-display.display('- %s is already installed, skipping.' % 
str(role))
-continue
+if not force:
+display.display('- %s is already installed, skipping.' 
% str(role))
+continue
 
 try:
 installed = role.install()
```

Please let me know if there's anything I can do to help this along.

Cheers,
JD

-- System Information:
Debian Release: 9.1
  APT prefers stable
  APT policy: (600, 'stable'), (500, 'stable'), (300, 'testing'), (1, 
'unstable')
Architecture: amd64 (x86_64)

Kernel: Linux 4.9.0-3-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_US.utf8, LC_CTYPE=en_US.utf8 (charmap=UTF-8), 
LANGUAGE=en_US.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages ansible depends on:
ii  python2.7.13-2
ii  python-crypto 2.6.1-7
ii  python-httplib2   0.9.2+dfsg-1
ii  python-jinja2 2.8-1
ii  python-netaddr0.7.18-2
ii  python-paramiko   2.0.0-1
ii  python-pkg-resources  33.1.1-1
ii  python-yaml   3.12-1

Versions of packages ansible recommends:
ii  python-cryptography  1.7.1-3
ii  python-jmespath  0.9.3-1
ii  python-kerberos  1.1.5-2+b2
ii  python-libcloud  1.5.0-1
ii  python-selinux   2.6-3+b1
pn  python-winrm 
ii  python-xmltodict 0.10.2-1

Versions of packages ansible suggests:
pn  cowsay   
pn  sshpass  

-- no debconf information

Bug#873890: ansible: enabling systemd service fails in chroot

[JD Friedrikson]

Package: ansible
Version: 2.2.1.0-2
Severity: normal

Hello,

To summarize, my issue is the same as the one here:
https://github.com/ansible/ansible/issues/21026

Fixes have already been merged:
https://github.com/ansible/ansible/pull/21072
https://github.com/ansible/ansible/pull/23904 # mildly related, but maybe 
should be brought up

and should be already included in 2.3.1.0+dfsg-2 (testing and unstable).

Here are the git commits that resolve this:

```
commit 2f0070639f3af9afca3b0280f89346c12e7b7e12
Author: Brian Coca 
Date:   Mon Feb 6 13:05:29 2017 -0500

attempt to fix systemd in chroot env

fixes #21026

diff --git a/lib/ansible/modules/system/systemd.py 
b/lib/ansible/modules/system/systemd.py
index c82bed60d..c970bd3ff 100644
--- a/lib/ansible/modules/system/systemd.py
+++ b/lib/ansible/modules/system/systemd.py
@@ -297,6 +297,7 @@ def main():
 
 # check service data, cannot error out on rc as it changes across 
versions, assume not found
 (rc, out, err) = module.run_command("%s show '%s'" % (systemctl, unit))
+
 if rc == 0:
 # load return of systemctl show into dictionary for easy access and 
return
 multival = []
@@ -327,6 +328,18 @@ def main():
 if is_systemd and 'LoadError' in result['status']:
 module.fail_json(msg="Error loading unit file '%s': %s" % 
(unit, result['status']['LoadError']))
 
+elif out.find('ignoring request') != -1:
+# fallback list-unit-files as show does not work on some systems 
(chroot)
+# not used as primary as it skips some services (like those using 
init.d) and requires .service notation
+if unit.endswith('.service'):
+service = unit
+else:
+service = '%s.service' % unit
+(rc, out, err) = module.run_command("%s list-unit-files '%s'" % 
(systemctl, service))
+if rc == 0:
+is_systemd = True
+
+
 # Does service exist?
 found = is_systemd or is_initd
 if is_initd and not is_systemd:
```
```
commit fa93bf70968f0b65422bd020a4adf9484dde8751
Author: Brian Coca 
Date:   Wed Feb 8 15:05:11 2017 -0500

allow more than .service, onus on user

diff --git a/lib/ansible/modules/system/systemd.py 
b/lib/ansible/modules/system/systemd.py
index c970bd3ff..831637064 100644
--- a/lib/ansible/modules/system/systemd.py
+++ b/lib/ansible/modules/system/systemd.py
@@ -33,7 +33,7 @@ options:
 name:
 required: true
 description:
-- Name of the service.
+- Name of the service. When using in a chroot environment you 
always need to specify the full name i.e. (crond.service).
 aliases: ['unit', 'service']
 state:
 required: false
@@ -330,12 +330,8 @@ def main():
 
 elif out.find('ignoring request') != -1:
 # fallback list-unit-files as show does not work on some systems 
(chroot)
-# not used as primary as it skips some services (like those using 
init.d) and requires .service notation
-if unit.endswith('.service'):
-service = unit
-else:
-service = '%s.service' % unit
-(rc, out, err) = module.run_command("%s list-unit-files '%s'" % 
(systemctl, service))
+# not used as primary as it skips some services (like those using 
init.d) and requires .service/etc notation
+(rc, out, err) = module.run_command("%s list-unit-files '%s'" % 
(systemctl, unit))
 if rc == 0:
 is_systemd = True
```

As this does affect usability for the chroot connector I feel that these fixes 
should be backported.

Please let me know if I'm not doing this correctly or if you need anything more 
from me.

Cheers,
JD


-- System Information:
Debian Release: 9.1
  APT prefers stable
  APT policy: (500, 'stable'), (1, 'unstable')
Architecture: amd64 (x86_64)

Kernel: Linux 4.9.0-3-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_US.utf8, LC_CTYPE=en_US.utf8 (charmap=UTF-8), 
LANGUAGE=en_US.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages ansible depends on:
ii  python2.7.13-2
ii  python-crypto 2.6.1-7
ii  python-httplib2   0.9.2+dfsg-1
ii  python-jinja2 2.8-1
ii  python-netaddr0.7.18-2
ii  python-paramiko   2.0.0-1
ii  python-pkg-resources  33.1.1-1
ii  python-yaml   3.12-1

Versions of packages ansible recommends:
ii  python-kerberos   1.1.5-2+b2
ii  python-selinux2.6-3+b1
pn  python-winrm  
ii  python-xmltodict  0.10.2-1

Versions of packages ansible suggests:
pn  cowsay   
pn  sshpass  

-- no debconf information

Bug#863850: [pkg-lxc-devel] Bug#863850: systemd service file does not stop systemd-based containers

[JD Friedrikson]

Hi again,

> You are absolutelly right! Thanks for catching this.
> And thanks for fixing it upstream!

You're welcome! I enjoy fixing things.

> I fear it's too late for Stretch 9.0, but I'll see to include that into
> 9.1.

It appears that my endeavors have been foiled. Very well, then. Is there any 
place where we can document the issue or do we just leave it here until 9.1?

Cheers,
JD

Bug#863850: systemd service file does not stop systemd-based containers

[JD Friedrikson]

Package: lxc
Version: 1:2.0.7-2

Hello,

Debian's packaged version of LXC currently is not able to stop systemd-based 
containers as they have not responded to SIGPWR as of 
https://github.com/lxc/lxc/commit/8eb62c245e9b67b451ba0766f3ecd7c6f2081d73 .

The appropriate way to stop systemd via a signal is to use SIGRTMIN+3 (or, I 
think, SIGRTMIN+4). The lxc-stop binary automatically determines whether the 
container will respond to this signal and handles it appropriately. Therefore, 
we should use that binary with ExecStop instead of using a signal (in the 
service file).

This has already been fixed upstream:

https://github.com/lxc/lxc/commit/c08d29b6d134fbb94d2cff0454ce27eb66930c4d

It would be cool if we could package this fix before the release. Here's a 
patch:

"""

diff --git a/config/init/systemd/l...@.service.in 
b/config/init/systemd/l...@.service.in
index 44d11e8e..a2aa2211 100644
--- a/config/init/systemd/l...@.service.in
+++ b/config/init/systemd/l...@.service.in
@@ -8,9 +8,9 @@ Documentation=man:lxc-start man:lxc
[Service]
Type=simple
KillMode=mixed
-KillSignal=SIGPWR
TimeoutStopSec=120s
ExecStart=@BINDIR@/lxc-start -F -n %i
+ExecStop=@BINDIR@/lxc-stop -n %i
# Environment=BOOTUP=serial
# Environment=CONSOLETYPE=serial
Delegate=yes
--
2.13.0

"""

Please let me know if you need anything more from me.

Cheers,
JD

Bug#861282: packer: FTBFS

[JD Friedrikson]

> 
> Thanks for your verifacation under qemu.
> So I'm going to release, as the autoremoval dealline is drawing very 
> near..
> 
> If you meet other issues, please report again.
> 
> Cheers,
> --
> Roger Shimizu, GMT +9 Tokyo
> PGP/GPG: 4096R/6C6ACD6417B3ACB1
> 

Hi Roger,

Thanks so much for your help! I can confirm that the patches work with 
Virtualbox too.

Cheers,
JD

Bug#861282: packer: FTBFS

[JD Friedrikson]

> 
> Thanks for your verifacation under qemu.
> So I'm going to release, as the autoremoval dealline is drawing very 
> near..
> 
> If you meet other issues, please report again.
> 
> Cheers,
> --
> Roger Shimizu, GMT +9 Tokyo
> PGP/GPG: 4096R/6C6ACD6417B3ACB1
> 

Hi Roger,

Thanks so much for your help! I can confirm that the patches work with 
Virtualbox too.

Cheers,
JD

Bug#861282: packer: FTBFS

[JD Friedrikson]

> 
> Thanks for your info!
> I pushed a commit to new branch fix_861282, that applied:
> ee5d13611fb8aca1f1014f9bcd65c18fffdd1b2b
> a0052fdb687f80ac07e67d7a0f39dcf3a66e32dd
> 
> (https://anonscm.debian.org/cgit/pkg-go/packages/packer.git/log/?h=fix_861282)
> 
> Please help to test whether it fixes your problem.
> 
> Cheers,
> --
> Roger Shimizu, GMT +9 Tokyo
> PGP/GPG: 4096R/6C6ACD6417B3ACB1
> 

The new changes package nicely and pass the tests just fine. I can confirm that 
the fix works for qemu. I'm having some issues with Virtualbox, but I've also 
always had trouble with Virtualbox and packer with the machine; the binary I've 
compiled straight from Packer's git project also fails in the same way so it's 
probably not related.

Can someone step in and confirm that the new patches work with Virtualbox?

Thanks,
JD

Bug#861282: packer: FTBFS

[JD Friedrikson]

Thanks for the patch! I'm having a bit of trouble with the build:


"""
?   github.com/mitchellh/packer/version[no test files]
dh_auto_test: go test -v -p 1 -timeout 30m -run 
^Test([^M]|M[^u]|Mu[^x]|Mux[^B]|MuxB[^r]|MuxBr[^o]|MuxBro[^k]|MuxBrok[^e]|MuxBroke[^r])
 github.com/mitchellh/packer github.com/mitchellh/packer/builder/amazon/chroot 
github.com/mitchellh/packer/builder/amazon/common 
github.com/mitchellh/packer/builder/amazon/ebs 
github.com/mitchellh/packer/builder/amazon/instance 
github.com/mitchellh/packer/builder/azure/arm 
github.com/mitchellh/packer/builder/azure/common 
github.com/mitchellh/packer/builder/azure/common/constants 
github.com/mitchellh/packer/builder/azure/common/lin 
github.com/mitchellh/packer/builder/azure/common/logutil 
github.com/mitchellh/packer/builder/azure/pkcs12 
github.com/mitchellh/packer/builder/azure/pkcs12/rc2 
github.com/mitchellh/packer/builder/digitalocean 
github.com/mitchellh/packer/builder/docker 
github.com/mitchellh/packer/builder/file 
github.com/mitchellh/packer/builder/googlecompute 
github.com/mitchellh/packer/builder/null 
github.com/mitchellh/packer/builder/openstack
  github.com/mitchellh/packer/builder/parallels/common 
github.com/mitchellh/packer/builder/parallels/iso 
github.com/mitchellh/packer/builder/parallels/pvm 
github.com/mitchellh/packer/builder/qemu 
github.com/mitchellh/packer/builder/virtualbox/common 
github.com/mitchellh/packer/builder/virtualbox/iso 
github.com/mitchellh/packer/builder/virtualbox/ovf 
github.com/mitchellh/packer/builder/vmware/common 
github.com/mitchellh/packer/builder/vmware/iso 
github.com/mitchellh/packer/builder/vmware/vmx 
github.com/mitchellh/packer/command github.com/mitchellh/packer/common 
github.com/mitchellh/packer/common/json github.com/mitchellh/packer/common/ssh 
github.com/mitchellh/packer/common/uuid 
github.com/mitchellh/packer/communicator/none 
github.com/mitchellh/packer/communicator/ssh 
github.com/mitchellh/packer/communicator/winrm github.com/mitchellh/packer/fix 
github.com/mitchellh/packer/helper/builder/testing 
github.com/mitchellh/packer/helper/communicator 
github.com/mitchellh/packer/helper/config g
 ithub.com/mitchellh/packer/helper/flag-kv 
github.com/mitchellh/packer/helper/flag-slice 
github.com/mitchellh/packer/packer github.com/mitchellh/packer/packer/plugin 
github.com/mitchellh/packer/packer/rpc 
github.com/mitchellh/packer/plugin/example 
github.com/mitchellh/packer/post-processor/amazon-import 
github.com/mitchellh/packer/post-processor/artifice 
github.com/mitchellh/packer/post-processor/atlas 
github.com/mitchellh/packer/post-processor/compress 
github.com/mitchellh/packer/post-processor/docker-import 
github.com/mitchellh/packer/post-processor/docker-push 
github.com/mitchellh/packer/post-processor/docker-save 
github.com/mitchellh/packer/post-processor/docker-tag 
github.com/mitchellh/packer/post-processor/shell-local 
github.com/mitchellh/packer/post-processor/vagrant 
github.com/mitchellh/packer/post-processor/vagrant-cloud 
github.com/mitchellh/packer/post-processor/vsphere 
github.com/mitchellh/packer/provisioner 
github.com/mitchellh/packer/provisioner/ansible github.com/mitche
 llh/packer/provisioner/ansible-local 
github.com/mitchellh/packer/provisioner/chef-client 
github.com/mitchellh/packer/provisioner/chef-solo 
github.com/mitchellh/packer/provisioner/file 
github.com/mitchellh/packer/provisioner/powershell 
github.com/mitchellh/packer/provisioner/puppet-masterless 
github.com/mitchellh/packer/provisioner/puppet-server 
github.com/mitchellh/packer/provisioner/salt-masterless 
github.com/mitchellh/packer/provisioner/shell 
github.com/mitchellh/packer/provisioner/shell-local 
github.com/mitchellh/packer/provisioner/windows-restart 
github.com/mitchellh/packer/provisioner/windows-shell 
github.com/mitchellh/packer/scripts github.com/mitchellh/packer/template 
github.com/mitchellh/packer/template/interpolate 
github.com/mitchellh/packer/version returned exit code 1
debian/rules:17: recipe for target 'override_dh_auto_test' failed
make[1]: *** [override_dh_auto_test] Error 1
make[1]: Leaving directory '/tmp/packer'
debian/rules:13: recipe for target 'build' failed
make: *** [build] Error 2
dpkg-buildpackage: error: debian/rules build gave error exit status 2
debuild: fatal error at line 1116:
dpkg-buildpackage -rfakeroot -us -uc -i -b failed
"""


Am I doing something wrong here?

JD

> 
> Thanks for your info!
> I pushed a commit to new branch fix_861282, that applied:
> ee5d13611fb8aca1f1014f9bcd65c18fffdd1b2b
> a0052fdb687f80ac07e67d7a0f39dcf3a66e32dd
> 
> (https://anonscm.debian.org/cgit/pkg-go/packages/packer.git/log/?h=fix_861282)
> 
> Please help to test whether it fixes your problem.
> 
> Cheers,
> --
> Roger Shimizu, GMT +9 Tokyo
> PGP/GPG: 4096R/6C6ACD6417B3ACB1
> 

Bug#861282: [No Subject]

[JD Friedrikson]

Here are the four patches that are related to this issue:

https://github.com/hashicorp/packer/commit/28ee60d216e49d565d654443b57295ce37197db1
https://github.com/hashicorp/packer/commit/249cb690e04477582aefadf4350a087bf4e33a87
https://github.com/hashicorp/packer/commit/ee5d13611fb8aca1f1014f9bcd65c18fffdd1b2b
https://github.com/hashicorp/packer/commit/a0052fdb687f80ac07e67d7a0f39dcf3a66e32dd

28ee60d216e49d565d654443b57295ce37197db1 is the patch that is currently 
packaged with debian

249cb690e04477582aefadf4350a087bf4e33a87 concerns vendor files which Debian 
doesn't appear to carry as part of the source package

ee5d13611fb8aca1f1014f9bcd65c18fffdd1b2b and 
a0052fdb687f80ac07e67d7a0f39dcf3a66e32dd both concern drivers to proprietary 
software (which Debian also does not seem to carry) so both need to be trimmed 
before applying

After removing the references to missing files, 
a0052fdb687f80ac07e67d7a0f39dcf3a66e32dd works just fine.

ee5d13611fb8aca1f1014f9bcd65c18fffdd1b2b is the one I'm having trouble with 
currently as it's failing for openstack's and amazon's ssh.go files. Not sure 
what changes we're carrying for those as I haven't dug too deep, but that's 
where I'm at now.

JD

Bug#861282: Not Yet Resolved

[JD Friedrikson]

Hello,

Packer is still broken due to the ABI changes. AFAIK the patch that was 
supposed to fix this does nothing to address use of the ABI in the builders. 
Since then, there have also been some patches upstream (as Matt had mentioned) 
to resolve the fallout due to the change:

https://github.com/hashicorp/packer/pull/4900/files
https://github.com/hashicorp/packer/pull/4910/files

Here's some output to look at for what's failing:

"""
Package: packer
Version: 0.10.2+dfsg-5
Installed-Size: 31971
Maintainer: Daniel Stender 
Architecture: amd64
"""

"""
==> virtualbox-iso: Waiting for SSH to become available...
2017/05/21 12:37:21 ui: ==> virtualbox-iso: Waiting for SSH to become 
available...
2017/05/21 12:37:21 packer: 2017/05/21 12:37:21 [INFO] Attempting SSH 
connection...
2017/05/21 12:37:21 packer: 2017/05/21 12:37:21 reconnecting to TCP connection 
for SSH
2017/05/21 12:37:21 packer: 2017/05/21 12:37:21 handshaking with SSH
2017/05/21 12:37:21 packer: 2017/05/21 12:37:21 handshake error: ssh: must 
specify HostKeyCallback
2017/05/21 12:37:21 packer: 2017/05/21 12:37:21 [DEBUG] SSH handshake err: ssh: 
must specify HostKeyCallback
"""

"""
==> qemu: Waiting for SSH to become available...
2017/05/21 12:40:33 ui: ==> qemu: Waiting for SSH to become available...
2017/05/21 12:40:33 packer: 2017/05/21 12:40:33 [INFO] Attempting SSH 
connection...
2017/05/21 12:40:33 packer: 2017/05/21 12:40:33 reconnecting to TCP connection 
for SSH
2017/05/21 12:40:33 packer: 2017/05/21 12:40:33 handshaking with SSH
2017/05/21 12:40:33 packer: 2017/05/21 12:40:33 handshake error: ssh: must 
specify HostKeyCallback
2017/05/21 12:40:33 packer: 2017/05/21 12:40:33 [DEBUG] SSH handshake err: ssh: 
must specify HostKeyCallback
"""

As of right now, packer (per Stretch) won't successfully build anything that 
depends on provisioning via SSH. Can we please look into this and do what's 
needed so that it's ready for the release?

Thanks,
JD