Update production environment for general availability. [20h]

[dpinney]

1. Find bigger VM size.
2. Fix email issue.
3. Redeploy newest code.
4. Stop and archive old VMs.
5. Update the SSL cert.
6. Make sure wsgi app isn't in werkzeug debug mode.
7. Update users and feeders.
8. Final git pull update.
9. Fix email key issue.

[dpinney]

We're going with m3.large.

All options:

[dpinney]

Working with AWS tech support to get our account unblocked. Looks like an AWS bug.

[dpinney]

Still waiting on AWS support.

[dpinney]

Deleted a corrupt key and pushed AWS support to unblock. They will be hearing from me every hour on the hour...

[dpinney]

Unblocked! Turned out the problem was that no access keys of any kind can be shared publicly, and Amazon scans the internet to try to find them. Yikes. Well, we complied with their requirements and all privileges are restored.

[dpinney]

Production updated. Waiting for last code changes to go public.

[dpinney]

Few status updates: [X]Get code on new VM. [X]What's our code pull options? We just need to do a git pull in the omf directory. [X]Publish update to doc on wiki? No, too risky. [X]Get cloudDeploy.sh updated. Moved /omf/data to safe place, then run pulls. [X]We're getting OS permission denied errors on spawning process. It's a problem with the embedded copy of Gridlab. Switch to system copy and we're cool. [X]Add omf to path via omf.pth. [X]Fix voltage drop. We needed to set the matplotlib backed to Agg in /etc/matplotlibrc.

[dpinney]

Almost there!

[X]Migrate past version prod users. Tested locally first. Just dropped em in and tried to log in. Excess user info deleted. [X]Migrated feeders from old prod database. [X]Archive the data from past production.

[dpinney]

[X]Fix email. Issue was expired keys. Added new key security infrastructure.