539 karma · joined September 17, 2013
Github/twitter: @micahhausler
[ my public key: https://keybase.io/micahhausler; my proof: https://keybase.io/micahhausler/sigs/rg60fRi4UGRr1oFpkl2hOHdQFc4-uUekdcDHbhayDm8 ]
Secondly, It seems that the author did not follow the stated docker procedure for security vulnerabilities (https://www.docker.com/resources/security/). In my view this taints the author's position of revealing a vulnerability to the community and reduces it to mere attention-getting.
Chattanooga has its own public utility that provides power and Gigabit fiber. See http://www.salon.com/2014/07/18/comcasts_worst_nightmare_how...
EDIT: Its slightly off, but still about one a month
Secondary, a native Darwin Docker server would be killer.
The mobile implications could be huge if this made it out of Windows Server.
There are several reasons that we use this:
* 'Template' out our infrastructure. If us-east-1 is ever down, we can easily re-create and spin up in us-west or wherever we'd like.
* If us-east-1 ever goes offline (a-la hurricane sandy), we have a record of what size/how many resources we had in production. Especially helpful since our infrastructure is dynamic. We add autoscale groups as we gain customers.
* We can easily foriegn key models in other apps in our project to an Autoscale group/ELB's etc. Ex: a customer is assigned to a particular autoscale group or load balancer.
* The loss in accuracy is acceptable, as we don't use every feature of AWS, so we only (at this point) have built out the necessary features.
Features:
* Models have several extra/modified methods:
.sync(force=False) # push changes to AWS
.should_sync() # if last .last_aws_sync() > self.last_update_time
.destroy() # for deleting in AWS
.delete(destroy=False) # override of .delete to call .destroy() on delete
* Uses django-encrypted-fields for storing passwords, SSH keys
* Uses django-netfields for validating subnets
* Uses jsonfield for validation and storage of IAM policies in postgres.
Each model has an additional interface for making calls to AWS. This keeps .models from being too huge.
There is still much to model and build out, but we're actively using it in production.