Sam Halicke's questions

I have an Amazon RDS instance set up in the 'classic' EC2 (no VPC.)

I also have a VPC set up that holds our newer, migrated applications and such.

However, it would seem that there is no way to specify a combination of either CIDR or EC2 security groups in the DB Security Groups that will allow my VPC instances to connect. The allowed security groups are only the 'classic' security groups.

I cannot find any documentation that suggests this is not doable, but I cannot find any documentation in the affirmative, either. Googling lands on a lot of results the 'other way', e.g. RDS instances in VPC and instances in EC2. I have the opposite problem.

Am I looking at a point-in-time restore to a new instance(s) in VPC to get connectivity to work? I understand I can maintain a public availability requirement that I have, but I would like to simply allow my VPC instances to connect for the time being.

I'm experiencing the following.

If I download and run a base box (ubuntu/trusty64), make some modifications like installing packages, then shutdown and repackage the box, on subsequent imports it fails to connect on vagrant up.

This is not the transient wait-for-connection error, this hangs forever:

default: SSH address: 127.0.0.1:2200
default: SSH username: vagrant
default: SSH auth method: private key
default: Warning: Connection timeout. Retrying...
default: Warning: Remote connection disconnect. Retrying...

Workflow is:

vagrant box add ubuntu/trusty64
vagrant init
# change config.vm.box in Vagrantfile to ubuntu/trusty64
vagrant up
vagrant ssh -c "sudo apt-get -y install git"
vagrant halt
vagrant package --output some-box.box
vagrant destroy

vagrant box add --name some-box ./some-box.box
# change config.vm.box in Vagrantfile to some-box
vagrant up

... this does not work.

I suspect this has something to do with VirtualBox networking and some sort of mapping or value that I need to capture or reset prior to packaging but I don't know what.

Using ssh-agent and private keys per the usual. Everything's working as normal.

My question regards best practices on flushing keys from ssh-add on activity like sleep, suspend, hibernate, etc. I thought about writing a simple wrapper around those commands, but then wondered if are they even called? Or does the kernel initiate this activity directly? Are the PM utilities strictly userland?

I would like this additional layer of security beyond locking my screen, etc. and was wondering if anyone else had solved this elegantly or has best practices to recommend. Thanks.