-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/44871/#review123815
-----------------------------------------------------------


Ship it!




2:44 for a destroy/up after the initial image pull from Montana.  Thats 
qualitatively much better than before!


The e2e ran green for me, albeit on my 2nd attempt after a trip home.  I don't 
have details for you on the initial failure except for the inability of the 
client to find the scheduler.
The connection difficulty somehow went away on the second run.

I'm away until Sunday so this LGTM as it stands, but you may want to replace me 
with another reviewer if there are substantial changes.


.gitignore (line 1)
<https://reviews.apache.org/r/44871/#comment186092>

    Not related, but noticing this is a lie iiuc.  Should be 'Use /dist/ to 
ignore dist in project root.' (reference 'PATTERN FORMAT' section here: 
https://www.kernel.org/pub/software/scm/git/docs/gitignore.html).



examples/vagrant/provision-dev-cluster.sh (line 87)
<https://reviews.apache.org/r/44871/#comment186102>

    Is this `/home/vagrant/aurora`?  As things stand almost all paths - save 
for `pushd aurora` in the `prepare_extras` function - are absolute; so this 
relative path reads risky / uncertain.


- John Sirois


On March 15, 2016, 5:59 p.m., Bill Farner wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/44871/
> -----------------------------------------------------------
> 
> (Updated March 15, 2016, 5:59 p.m.)
> 
> 
> Review request for Aurora, John Sirois and Zameer Manji.
> 
> 
> Repository: aurora
> 
> 
> Description
> -------
> 
> Relevant docs on packer can be found here: https://www.packer.io/docs/
> 
> To support this, i have created the [`apache-aurora` 
> org](https://atlas.hashicorp.com/apache-aurora).  If we move forward with 
> this patch, i will be looking to add other committers to the org and improve 
> the bus factor.
> 
> 
> Diffs
> -----
> 
>   .gitignore 1af09a251b3f76c13813033d32aa7efba9aef304 
>   Vagrantfile 2d6c2ae598e80035840f7e517e161be266b581dd 
>   build-support/packer/README.md PRE-CREATION 
>   build-support/packer/aurora.json PRE-CREATION 
>   build-support/packer/build.sh PRE-CREATION 
>   examples/vagrant/provision-dev-cluster.sh 
> df33a30c1dcc0ea832899c849be05fe9deb47433 
> 
> Diff: https://reviews.apache.org/r/44871/diff/
> 
> 
> Testing
> -------
> 
> I can run `vagrant up`.  Currently looking into a possible configuration 
> issue in end-to-end tests, to be corrected before submitting.
> 
> Reviewers - you can do this as well by applying the patch.  You will observe 
> a long fetch on the first run (much like you would with the current repo if 
> you wiped your local vagrant boxes).  However, subsequent `vagrant destroy`, 
> `vagrant up` cycles should be relatively fast, about 1m30s for me.
> 
> 
> Thanks,
> 
> Bill Farner
> 
>

Reply via email to