add proposal doc for backing up ghe to s3 (#344)

aws/proposals/github-concerns/README.md

@@ -0,0 +1,139 @@
+# GitHub Enterprise Backup to S3
+
+This proposal outlines a plan to safely export all GitHub Enterprise content (code repositories, issues, wikis, releases, projects, etc.) to AWS S3 storage in the event of a lapse in funding for our GitHub Enterprise licenses.
+
+## Why
+
+During a government funding lapse, as documented in [aws/documentation/gov-funding-lapse/README.md](../documentation/gov-funding-lapse/README.md), there's a risk that the Census Bureau may be unable to maintain payments for GitHub Enterprise licenses. If access to GitHub Enterprise is suspended due to non-payment, all code repositories, issues, wikis, releases, projects, and other critical development assets could become temporarily inaccessible, severely impacting development activities and knowledge management once funding is restored.
+
+Creating a backup on AWS S3 provides a safety net that ensures:
+1. Continuity of access to critical code and documentation during funding lapses
+2. Protection against data loss or temporary service suspensions
+3. Ability to continue some development activities in an alternative environment if necessary
+4. Quick restoration of normal operations after funding is restored
+
+## What
+
+We propose implementing a backup process that will:
+
+1. Export the following GitHub Enterprise assets to S3:
+   * Git repositories (code, commits, branches, tags)
+   * Issues and pull requests (including comments and metadata)
+   * Wikis
+   * Releases and assets
+   * Project boards
+   * GitHub Pages content
+   * Organization and team metadata
+
+2. Use AWS services to:
+   * Store exports in a secure S3 bucket with versioning enabled
+   * Implement lifecycle policies to manage storage costs
+   * Encrypt data at rest using AWS KMS
+   * Maintain access logs for all backup operations
+
+3. Create a documentation and recovery process for:
+   * Accessing backed-up content during GitHub outages
+   * Restoring repositories if needed
+   * Continuing development using the backups during an extended outage
+
+## Where
+
+The backup infrastructure will be deployed in AWS with the following components:
+
+1. Primary storage bucket:
+   * Region: us-gov-east-1 (GovCloud East)
+   * Bucket naming format: `v-s3-ghe-backup-{org}-{env}`
+
+2. Redundant backup bucket (optional):
+   * Region: us-gov-west-1 (GovCloud West) 
+   * Cross-region replication from primary bucket
+
+3. AWS accounts:
+   * Primary: ent-gov-operations-prod
+   * Potential backup hosting in lab-gov-operations-nonprod for dev/testing
+
+## When
+
+This capability is needed within the next 30 days to prepare for any potential funding lapses that might occur in the next fiscal year. The implementation timeline includes:
+
+1. Week 1: Design and approval of backup architecture
+2. Week 1-2: Development of backup scripts and infrastructure
+3. Week 2: Testing in lab environment
+4. Week 2: Documentation and production deployment
+5. Week 3-4: Validation and team training
+
+## Who
+
+This initiative will be led by the Secure Cloud Team (SCT) with support from:
+
+1. Primary stakeholders:
+   * GitHub Enterprise administrators (ADSD)
+   * Cloud Infrastructure team (SCT/CSVD)
+   * Security operations team (CSVD Sec/OIS)
+   * Development team leads who depend on GitHub (SCT/CSVD/ADSD)
+
+2. Users:
+   * All development teams currently using GitHub Enterprise
+   * Operations personnel who may need to restore content during outages
+   * Security teams who need to maintain audit trails
+
+## How
+
+The backup process will be implemented using the following approach:
+
+1. Automation scripts:
+   ```python
+   # Example GitHub backup script pseudocode
+   from datetime import datetime, timezone
+   import github
+   import boto3
+
+   timestamp = datetime.now(timezone.utc).isoformat()
+   s3_client = boto3.client('s3')
+
+   # Connect to GitHub using token or app authentication
+   gh = github.Github(auth_token)
+
+   # For each org in our GitHub Enterprise
+   for org in gh.get_organizations():
+       # Export repositories, issues, wikis, etc.
+       # Upload to S3 with appropriate prefixes
+   ```
+
+2. Infrastructure as Code:
+   * AWS S3 bucket creation via Terraform
+   * IAM roles and policies for secure access
+
+3. Monitoring and auditing:
+   * CloudWatch alarms for backup failures
+   * S3 access logging
+
+4. Recovery process:
+   * Documentation for accessing backups
+   * Scripts for bulk restoration if needed
+
+## Tagging
+
+In accordance with the [government funding lapse guidelines](../documentation/gov-funding-lapse/README.md), we'll use the following tags:
+
+* `gfl_shutdown_excluded`: `true` - These backups should continue even during funding lapses
+* `backup_source`: `github_enterprise`
+* `data_classification`: `internal`
+
+## Links
+
+Product Documentation Links:
+* [GitHub Enterprise API](https://docs.github.com/en/enterprise-server/rest)
+* [AWS S3](https://docs.aws.amazon.com/AmazonS3/latest/userguide/what-is-s3.html)
+* [AWS S3 Cross-Region Replication](https://docs.aws.amazon.com/AmazonS3/latest/userguide/replication.html)
+* [GitHub Enterprise Backup Utilities](https://github.com/github/backup-utils)
+
+Internal Links:
+* [Government Funding Lapse Plans](../documentation/gov-funding-lapse/README.md)
+* [Cloud Architecture Outline](../documentation/architecture-outline.md)
+* [AWS Account Naming](../documentation/account-setup/README.md)
+
+# CHANGELOG
+
+- 1.0.0 -- 2024-06-15
+  - initial proposal