Security Policy & Data Protection - OpenISP
Last Updated: September 13, 2026
Version: 1.0
1. INTRODUCTION
The OpenISP Security Policy outlines how OpenISP protects your data, systems, and infrastructure from unauthorized access, misuse, and threats. This policy applies to:
- Tenant ISP operators and their staff
- Subscribers whose data is processed
- OpenISP staff and authorized partners
- Third-party service providers with data access
2. SECURITY ARCHITECTURE
2.1 Defense-in-Depth Model
OpenISP implements multiple layers of security:
┌─────────────────────────────────────────────────────┐
│ Layer 1: Perimeter Security (Firewall, WAF, DDoS) │
├─────────────────────────────────────────────────────┤
│ Layer 2: Network Security (TLS, VPN, Segmentation) │
├─────────────────────────────────────────────────────┤
│ Layer 3: Application Security (Auth, Input Validation) │
├─────────────────────────────────────────────────────┤
│ Layer 4: Data Security (Encryption, Backup, Redaction) │
├─────────────────────────────────────────────────────┤
│ Layer 5: Physical Security (Data Center, Access) │
├─────────────────────────────────────────────────────┤
│ Layer 6: Operational Security (Monitoring, Incident Response) │
└─────────────────────────────────────────────────────┘
3. ENCRYPTION & CRYPTOGRAPHY
3.1 Encryption Standards
| Data State | Algorithm | Standard | Key Size |
|---|---|---|---|
| Transit (TLS) | AES-256-GCM | TLS 1.3+ | 256-bit |
| Transit (VPN) | ChaCha20-Poly1305 | WireGuard | 256-bit |
| Stored (Passwords) | Argon2id | OWASP | Variable salt |
| Stored (Sensitive Data) | AES-256-CBC | FIPS 140-2 | 256-bit |
| Database Backup | AES-256 | Encryption at rest | 256-bit |
| File Upload | AES-256-GCM | OpenSSL | 256-bit |
3.2 Encrypted Data Fields
The following fields are encrypted at rest:
Tenant Data:
- API tokens
- RADIUS server secret
- MikroTik API password
- WireGuard private keys
- Database connection strings
- Payment gateway credentials
- Third-party API keys
Subscriber Data:
- PPPoE password
- Device credentials
- TOTP secrets (2FA)
- Payment method references
3.3 Encryption Key Management
| Key Type | Generation | Storage | Rotation | Access |
|---|---|---|---|---|
| Master Key | Random 256-bit | Hardware Security Module (HSM) | Annual | Root only |
| Database Key | Derived from Master | HSM-backed key vault | Annual | DB service only |
| TLS Private Key | RSA 2048-bit | Encrypted on disk | Annual | TLS service only |
| Session Keys | Generated per session | Memory only | Per session | Current request only |
Key Backup: Master keys backed up to secure, geographically distributed vaults. Encryption keys never transmitted in plaintext.
3.4 Cipher Suites
TLS 1.3 Preferred Cipher Suite (in order):
TLS_AES_256_GCM_SHA384TLS_AES_128_GCM_SHA256TLS_CHACHA20_POLY1305_SHA256
Legacy Support (TLS 1.2):
ECDHE-ECDSA-AES256-GCM-SHA384ECDHE-ECDSA-AES128-GCM-SHA256
Explicitly Disabled: NULL, EXPORT, anon-DH, DES, RC4, MD5
4. AUTHENTICATION & ACCESS CONTROL
4.1 Authentication Methods
| User Type | Primary Method | Secondary (2FA) | Recovery |
|---|---|---|---|
| Tenant Admin | Username/Password (Argon2id) | TOTP or SMS | Backup codes |
| Tenant Staff | Username/Password | TOTP or SMS | Backup codes |
| Subscriber (PPPoE) | Username/Password (encrypted) | Optional TOTP | Account email/phone |
| Subscriber (Hotspot) | Mobile number (SMS/USSD) | Optional | Registration email |
| System Services | API token (HMAC-SHA256) | Mutual TLS (optional) | Token rotation |
4.2 Password Requirements
Tenant Admins:
- Minimum 12 characters
- Uppercase, lowercase, numbers, symbols
- Not in breach database (checked against HIBP)
- No more than 2 consecutive repeated characters
- No personal information (name, email, etc.)
- Expires every 90 days (optional force-reset on next login)
Subscribers:
- Minimum 8 characters (configurable per ISP)
- No character restrictions (usability > enforcement)
- No automatic expiration (reset on breach only)
- ISP may enforce stronger policies
4.3 Multi-Factor Authentication (MFA)
MFA Methods:
- TOTP (Time-based One-Time Password): Authenticator app (Authy, Google Authenticator, Microsoft Authenticator)
- SMS: One-time code sent via SMS (less secure; support for accessibility)
- Email: One-time code sent to registered email (async, less secure)
- Backup Codes: 10 single-use codes generated during MFA setup
MFA Enforcement:
- Required for all Tenant administrators (no exceptions)
- Optional for Subscriber hotspot access
- Mandatory within 7 days of account creation (for admins)
Recovery:
- Backup codes stored encrypted
- Account recovery requires email + phone verification
- Support can reset MFA with ID verification (security questions)
4.4 Session Management
| Aspect | Requirement |
|---|---|
| Session Timeout | 24 hours inactivity (re-login required) |
| Session ID | 256-bit random, cryptographically secure |
| Cookie Flags | HttpOnly, Secure, SameSite=Strict |
| Concurrent Sessions | Max 5 per admin account (older sessions terminated) |
| Session Fixation | Prevented (session ID regenerated after login) |
| CSRF Protection | Token per-request, verified on state-changing operations |
4.5 Role-Based Access Control (RBAC)
Platform Roles:
| Role | Permissions | Scope |
|---|---|---|
| SuperAdmin | All functions, cross-tenant view, policy changes | Platform-wide |
| Tenant Owner | Full account access, user management, billing, API keys | Single tenant |
| Tenant Admin | Dashboard, customer management, router configs, support | Single tenant |
| Tenant Staff | Limited dashboard, view-only analytics, basic support | Single tenant |
| Customer | Self-service portal, billing history, usage stats | Single subscriber |
Least Privilege: Each role has minimum permissions needed for function. Permissions cannot be combined/elevated except through role change.
4.6 API Authentication
API Token Format: openisp_[32 random characters]_[32 hash]
| Aspect | Requirement |
|---|---|
| Token Generation | One per API key request, stored hashed (SHA-256) |
| Token Lifetime | No expiration (manual revocation only) or custom expiration |
| Token Rotation | Manual rotation recommended every 90 days |
| Token Scope | Limited to specific operations (read, write, delete) |
| Rate Limiting | 100-1000 requests/minute per token (depending on endpoint) |
| IP Whitelisting | Optional per API key (recommended for production) |
| Mutual TLS | Optional for high-security integrations |
All application HTTP requests are also limited to 100 requests per minute per client IP. This broad ceiling covers public and authenticated traffic, including requests to unknown paths; stricter route-specific and user-specific limits still apply. Rate-limited requests receive HTTP 429 with a Retry-After value. This application-level protection complements, but does not replace, perimeter DDoS and WAF controls.
5. NETWORK SECURITY
5.1 HTTPS/TLS Enforcement
Requirements:
- All HTTP traffic redirected to HTTPS
- HSTS header:
Strict-Transport-Security: max-age=31536000; includeSubDomains; preload - Certificate: RSA 2048-bit or ECDSA 256-bit, issued by trusted CA
- Certificate validation: All browsers must pass validation checks
- Certificate pinning: Optional for API endpoints
SSL Labs Rating Target: A+ (100/100 score)
5.2 Network Segmentation
┌────────────────────────────────────────────┐
│ DMZ (Public-facing) │
│ ├── Web app (reverse proxy) │
│ ├── API Gateway (rate limiting) │
│ └── DNS (resolvers) │
├────────────────────────────────────────────┤
│ Application Tier │
│ ├── Web servers (no internet access) │
│ ├── API servers (no database direct access)│
│ └── Message queue (internal only) │
├────────────────────────────────────────────┤
│ Database Tier │
│ ├── PostgreSQL (encrypted connections) │
│ ├── Redis cache (memory only, no disk) │
│ └── Key vault (hardware security module) │
├────────────────────────────────────────────┤
│ Operational Tier │
│ ├── Logging service (centralized) │
│ ├── Monitoring (observability) │
│ └── Backup systems (air-gapped) │
└────────────────────────────────────────────┘
Access Rules:
- DMZ ↔ App Tier: Application-level only (no direct DB access)
- App Tier ↔ DB Tier: Encrypted connections, IP-whitelisted
- DB Tier: No outbound internet access
- Operational Tier: Management networks, VPN only
5.3 Firewall Rules
Inbound (Public):
- Port 443 (HTTPS): Worldwide
- Port 80 (HTTP redirect): Worldwide
- Port 53 (DNS): Configured servers only
Inbound (Admin VPN only):
- Port 22 (SSH): VPN clients + bastion host only
- Port 5432 (PostgreSQL): Application and FreeRADIUS hosts only
- Port 6379 (Redis): Disabled; management interface via SSH tunnel
Outbound (App → External):
- Port 443 (HTTPS): Whitelist only (no open outbound)
- Port 25 (SMTP): Email servers only
- Port 53 (DNS): ISP resolvers + Cloudflare (1.1.1.1)
Outbound (DB → External): Blocked (except backup to S3-like storage)
5.4 Reverse Proxy & WAF
Cloudflare/WAF Features:
- OWASP ModSecurity ruleset
- DDoS protection (Anycast + rate limiting)
- Bot detection (JavaScript challenge for suspicious traffic)
- IP reputation blocking (known botnet IPs)
- Geofencing (optional country restrictions)
- Request size limits: 100 KB headers, 10 MB body
Custom Rules:
- Block requests with SQL injection patterns
- Block requests with XSS payloads
- Block requests with directory traversal sequences
- Block requests with excessive headers
5.5 DDoS Protection
Strategy:
- Cloudflare Anycast + local scrubbing center
- Traffic rate-limiting: 1000 req/sec per IP (adaptive)
- Connection rate-limiting: 100 new connections/sec per IP
- Bandwidth limiting: 10 Gbps per customer + 100 Gbps total capacity
- Automatic failover to backup data center
Monitoring:
- Real-time traffic analysis
- Anomaly detection (ML-based)
- Automatic mitigation triggers
- Manual escalation for large attacks (> 50 Gbps)
6. APPLICATION SECURITY
6.1 Secure Development Practices
Code Review:
- All code changes reviewed by ≥2 developers
- Security review for features handling sensitive data
- Automated SAST (Static Application Security Testing) checks
- No code merged without passing reviews
Testing:
- Unit tests with ≥80% code coverage
- Integration tests for auth, payment, API flows
- Manual penetration testing quarterly
- Automated DAST (Dynamic Application Security Testing) in staging
Dependencies:
- Automated vulnerability scanning (Snyk, Dependabot)
- Weekly security patch reviews
- No use of packages with unresolved critical CVEs
- Pinned versions in production (no auto-updates)
6.2 Input Validation & Sanitization
Rule: All user input is untrusted until validated.
Validation:
- Whitelist approach (allow known-good patterns)
- Type checking (string vs. integer, email format, etc.)
- Length validation (min/max limits)
- Format validation (regex for email, phone, IP, etc.)
- Range validation (dates, quantities, amounts)
Sanitization:
- HTML entity encoding for output (prevents XSS)
- SQL parameterized queries (prevents SQL injection)
- URL encoding for URL parameters
- JSON serialization for API responses
- File upload: Scan for malware, reject executable types
Example Invalid Inputs Rejected:
name: "<script>alert('xss')</script>" // XSS attempt
email: "[email protected]' OR '1'='1" // SQL injection attempt
phone: "+1 (555) 123-4567 DROP TABLE users;" // Injection
amount: "-999999" // Negative amount
date: "2099-13-45" // Invalid date
6.3 Output Encoding
HTML Context:
<!-- Input: <script>alert('xss')</script> -->
<!-- Output: <script>alert('xss')</script> -->
<div>{{ user_input }}</div>
JavaScript Context:
// Input: '); alert('xss'); //
// Output: \'); alert(\'xss\'); //
var user = JSON.stringify(userInput);
URL Context:
// Input: /user/123?redirect=https://evil.com
// Output: /user/123?redirect=https%3A%2F%2Fevil.com
6.4 Cross-Site Request Forgery (CSRF) Protection
Implementation:
- CSRF tokens generated per session
- Token validated on all state-changing requests (POST, PUT, DELETE)
- Token sent in request body or
X-CSRF-Tokenheader - SameSite=Strict cookie flag prevents cross-site cookie sending
Example Flow:
1. User logs in → Session created, CSRF token generated
2. User visits form → CSRF token embedded in <form hidden>
3. User submits form → Token included in POST request
4. Server validates token before processing
6.5 Authentication Bypass Prevention
Protections:
- No hardcoded credentials in code
- Credentials not logged or exposed in errors
- Failed login attempts rate-limited (5 per minute per IP)
- Account lockout after 10 failed attempts (1 hour lockout)
- No "remember me" option persisting passwords
- No password recovery flows that expose information
6.6 Session Fixation Prevention
Protection:
- New session ID generated after login
- Old session terminated
- Session ID not in URL (uses secure cookies only)
- Session cookie marked HttpOnly (inaccessible to JavaScript)
7. DATA PROTECTION
7.1 Data Classification
| Classification | Examples | Protection Level |
|---|---|---|
| Public | Service documentation, public API docs | Standard HTTPS |
| Internal | System logs, internal metrics, IP addresses | Encryption + access control |
| Confidential | Passwords, API keys, RADIUS secrets | Encryption + MFA + audit logging |
| Restricted | Subscriber PII, payment data, health data | Encryption + MFA + encryption at rest + minimal access |
7.2 Data Minimization
Principle: Collect and retain only necessary data.
Application:
- Subscriber usernames only (no real names unless needed)
- Phone numbers only for SMS (not for calls)
- No email collection unless for communication
- No family info, employment, location (unless explicit consent)
- RADIUS logs: Limited to auth success/failure + bytes (no content)
Deletion:
- Inactive accounts deleted after 12 months (soft-delete)
- Support tickets deleted after 24 months
- Error logs deleted after 90 days
- Payment records retained 7 years (tax law)
7.3 Data Retention Policy
| Data Type | Retention Period | Reason |
|---|---|---|
| Active subscriber data | Until deletion | Current service |
| Deleted subscriber data | 90 days | Soft-delete recovery |
| Payment records | 7 years | Tax/audit compliance |
| RADIUS accounting | 12 months | Billing reconciliation |
| Access logs | 6 months | Security audit |
| Error logs | 90 days | Troubleshooting |
| Backups | 30 days | Disaster recovery |
| Support tickets | 24 months | Dispute resolution |
| Audit logs (admin actions) | 12 months | Change tracking |
Purge Process:
- Scheduled daily for logs, weekly for soft-deleted data
- Data overwritten 3 times (secure deletion)
- Verification that data is unrecoverable
- Audit log of purged records
7.4 Backup & Recovery
Backup Schedule:
- Database: Daily incremental, weekly full backup
- Configuration: Real-time replication
- Backups: Encrypted at rest, stored geographically distributed
Recovery SLA:
- Recovery Time Objective (RTO): 4 hours
- Recovery Point Objective (RPO): 1 hour
- Test recovery: Monthly (restore to test environment)
- Document recovery procedures: Maintained and updated
Backup Storage:
- Off-site storage (geographically separate)
- Immutable backups (cannot be modified after creation)
- Encryption: AES-256, separate keys from live database
- Retention: 30 days (older backups deleted)
7.5 Data Portability
Tenants may request data export in standard formats:
Formats:
- CSV (subscriber list, billing records, usage stats)
- JSON (configurations, API responses)
- PDF (invoices, reports)
- SQL dump (raw database schema + data)
Process:
- Request via support portal [email protected]
- Verification of requestor identity (admin account required)
- Export generated and encrypted
- Download link provided (valid 7 days)
- Completion within 5 business days
Cost: Free for exports ≤ 1 GB; $100 per 10 GB for larger exports
8. INCIDENT RESPONSE
8.1 Incident Response Plan
Incident Response Team:
- Security Officer (incident commander)
- Database Administrator (forensics, recovery)
- Application Engineer (code review, patch)
- Communications Manager (external communications)
- Legal Counsel (breach notification, compliance)
On-Call Schedule: 24/7/365 rotation, response within 15 minutes
8.2 Incident Categories & Response Times
| Severity | Definition | Response SLA | Example |
|---|---|---|---|
| Critical | Active data breach, service down, ransomware | 15 min | Unauthorized database access in progress |
| High | Data exposure (past breach), significant availability | 1 hour | Customer data exposed in logs |
| Medium | Partial service issue, potential vulnerability | 4 hours | Weak TLS cipher suite enabled |
| Low | No customer impact, informational | 1 business day | Configuration drift, minor log anomaly |
8.3 Incident Response Steps
1. Detection & Triage (0-15 min)
- Automated alerts: Intrusion detection, log analysis, uptime monitoring
- Manual reports: Abuse reports, customer complaints
- Initial assessment: Severity level, scope, systems affected
- Incident ID assigned, timeline started
2. Containment (15-60 min)
- Isolate affected systems (disconnect from network if necessary)
- Preserve evidence (memory dump, disk snapshot, logs)
- Stop ongoing damage (revoke credentials, block IPs, rate limit)
- Initial notification to affected parties (if safe to do so)
3. Investigation (1-24 hours)
- Forensic analysis: Examine logs, memory, disk
- Determine root cause: How did breach happen?
- Scope assessment: What data/systems affected?
- Timeline reconstruction: When did it start/end?
4. Remediation (1-7 days)
- Patch vulnerabilities
- Revoke compromised credentials
- Enhance monitoring
- Deploy additional controls
5. Notification (24-72 hours)
- Notify affected Tenants/Subscribers per legal requirements
- Provide: What, when, scope, impact, mitigation, credit/offer
- Offer: Free credit monitoring, identity theft protection (if PII exposed)
- Contact method: Email + phone + in-app notification
6. Recovery (1-30 days)
- Restore from backup (if compromised)
- Rebuild affected systems
- Restore normal monitoring
- Customer verification of fix
7. Post-Incident (30+ days)
- Root cause analysis: Why did this happen?
- Preventive measures: How to prevent recurrence?
- Process improvements: Update incident response plan
- Communication: Share learnings with team
- External notification: Regulatory filings, public disclosure (if applicable)
8.4 Breach Notification Timeline
Within 24 hours of discovery:
- Confirm breach scope and impact
- Initiate notification process
- Notify affected Tenant administrators (email + SMS)
Within 72 hours:
- Detailed notification to Tenants with:
- What data was affected
- Who was affected (number of subscribers)
- When the breach occurred
- How we discovered it
- What we're doing to fix it
- Credit/compensation offer (if applicable)
- Support contact information
Within 30 days:
- Formal incident report with:
- Root cause analysis
- Timeline of events
- Evidence of remediation
- Preventive measures
- Third-party forensics report (if commissioned)
Regulatory Notification:
- GDPR: Notify Supervisory Authority (if >1000 people affected)
- CCPA: Notify California Attorney General
- Industry regs: Comply with local telecommunications authority
8.5 Evidence Preservation
Chain of Custody:
- Document who accessed evidence, when, why
- Maintain integrity of forensic samples
- Encrypt evidence in storage
- Restricted access (security team only)
- Retention: 3 years minimum (per legal holds)
Forensic Tools:
- Memory analysis: Volatility Framework
- Disk forensics: Autopsy, EnCase
- Log analysis: Splunk, ELK Stack
- Network forensics: Wireshark, tcpdump
9. MONITORING & LOGGING
9.1 Centralized Logging
Logs Collected:
- Web server access logs (all HTTP requests)
- Application logs (errors, warnings, business events)
- Database logs (authentication, queries, connections)
- System logs (kernel, services, security)
- Network logs (firewall, IDS/IPS, VPN)
- API logs (all API requests with user, time, action)
- Authentication logs (logins, MFA, token generation)
Log Aggregation:
- Centralized SIEM (Security Information & Event Management)
- Real-time streaming (no buffering)
- Log retention: 6 months hot, 12 months archive
- Encryption: Logs encrypted in transit and at rest
Log Sanitization:
- Passwords never logged
- API keys redacted (only last 4 chars visible)
- Credit card numbers never logged (only last 4 digits)
- PII minimized in logs (first/last initial for names)
9.2 Security Monitoring
Alerts Triggered:
- Failed login attempts (>5 in 5 minutes)
- Privilege escalation attempts
- Unusual API activity (high volume, unusual patterns)
- Database access from unexpected IPs
- Certificate expiration warnings (30 days before)
- Suspicious file uploads
- New user account creation (admin review)
- Configuration changes (approval notifications)
Response to Alerts:
- Auto-escalate to security team (1-5 minute review)
- False positive rate monitored (target <5%)
- Tuning: Reduce false positives weekly
9.3 Real-Time Monitoring
Metrics Tracked:
- API response times (P50, P95, P99 latency)
- Database connection pool usage
- Error rates by endpoint
- RADIUS authentication success rate
- TLS handshake failures
- Certificate validation failures
- DDoS attack patterns
- Resource utilization (CPU, memory, disk, network)
Dashboards:
- Executive dashboard (uptime, incident count, trends)
- Operations dashboard (real-time metrics, alerts)
- Security dashboard (logins, API activity, anomalies)
- Customer support dashboard (ticket volume, response times)
10. VULNERABILITY MANAGEMENT
10.1 Vulnerability Identification
Sources:
- Automated scanning: Snyk (dependencies), Qualys (infrastructure)
- Manual testing: Penetration testing (quarterly)
- Bug bounty program: External security researchers
- Security advisories: CVE feeds, vendor notifications
- Code review: Static analysis during development
10.2 Vulnerability Scoring
CVSS Scoring:
- Critical (9.0-10.0): Fix immediately
- High (7.0-8.9): Fix within 1 week
- Medium (4.0-6.9): Fix within 1 month
- Low (0.1-3.9): Fix within 3 months
Internal Scoring Factors:
- Exploitability: Proof-of-concept available?
- Impact: Data exposure? Service unavailable?
- Affected customers: How many users impacted?
- Workarounds: Can customers mitigate?
10.3 Patching Policy
Timing:
- Critical: Emergency patch, immediate deployment
- High: Scheduled patch (within 1 week, tested first)
- Medium: Regular patch cycle (monthly)
- Low: Bundled in quarterly releases
Process:
- Develop patch in isolated branch
- Test in staging environment (replicate production)
- Code review by security + development team
- Deployment to production (rolling update, no downtime)
- Monitor for issues (rollback plan on standby)
- Document patch + validate fix
Communication:
- Critical: Customer notification before patch
- High: Notification with patch (1-2 business days)
- Medium/Low: Included in release notes
11. PHYSICAL SECURITY
11.1 Data Center Security
Facility Controls:
- Biometric access (fingerprint + badge)
- Video surveillance (24/7, 30+ day retention)
- Visitor logs (all visitors signed in/out)
- Background checks for data center staff
- Security guards on-site (24/7)
Environmental Controls:
- Temperature monitoring (18-24°C, alarms if off-range)
- Humidity monitoring (30-50%, automatic adjustment)
- Fire suppression (FM-200, no water damage to equipment)
- Power backup (UPS + diesel generators, 72-hour fuel)
- Redundant cooling systems
Physical Safeguards:
- Server racks in locked cages (key access only)
- Laptop/portable device encryption (BitLocker, FileVault)
- Secure disposal: Equipment shredding + data wipe (DOD 5220.22M)
- USB port restrictions (admin approval for external drives)
12. COMPLIANCE & STANDARDS
12.1 Compliance Frameworks
OpenISP aligns with:
| Framework | Scope | Status |
|---|---|---|
| ISO 27001 | Information Security Management | Aiming for certification |
| SOC 2 Type II | Service Organization Controls | Annual audit |
| GDPR | EU data protection | Compliant |
| CCPA | California privacy | Compliant |
| PCI-DSS | Payment card security | Compliant (scope: payment data handling) |
| HIPAA (if applicable) | Health information protection | Compliant (if processing health data) |
| NIST Cybersecurity Framework | Risk management | Aligned |
12.2 Audit & Certification
Annual Audits:
- SOC 2 Type II audit (independent auditor)
- Security assessment (penetration testing)
- Code security review (SAST tools)
- Infrastructure review (vulnerability scanning)
Audit Reports:
- Available to Tenants under NDA
- Requested via [email protected]
- Completed within 5 business days
13. SECURITY RESPONSIBILITIES
13.1 OpenISP Responsibilities
OpenISP provides:
- Secure infrastructure (encryption, backups, monitoring)
- Regular security updates (patching, library updates)
- Access controls (authentication, authorization, RBAC)
- Incident response (breach investigation, notification)
- Compliance reporting (audit reports, certifications)
OpenISP does NOT:
- Prevent Subscribers from using your service illegally
- Monitor Subscriber content (encrypted traffic, HTTPS)
- Control your routers or network configuration
- Manage your ISP's backups or disaster recovery
- Implement GDPR compliance on your behalf
13.2 Tenant ISP Responsibilities
Tenants must:
- Keep credentials confidential (password, API keys)
- Enable MFA on all admin accounts
- Regularly rotate RADIUS secrets and API tokens
- Monitor their Subscriber access and activity
- Secure their network equipment (physical security)
- Maintain separate backups of critical data
- Comply with local data protection laws
- Notify Subscribers of OpenISP's role as processor
- Report suspected breaches immediately
- Implement secure firewalls protecting routers
13.3 Subscriber Responsibilities
Subscribers should:
- Keep passwords confidential
- Don't share account credentials
- Keep devices updated with security patches
- Use strong, unique passwords
- Report suspected account compromise
- Comply with ISP's Acceptable Use Policy
14. SECURITY INCIDENT RESPONSE CONTACTS
14.1 Report a Security Issue
Security Email: [email protected]
PGP Key: Available at openisp.io/security/pgp (optional)
Response Time: Within 4 hours (critical issues)
14.2 Report a Vulnerability
Vulnerability Disclosure Policy: openisp.io/security/disclosure
Responsible Disclosure:
- Report privately via email (not public issues)
- Allow 90-day response period before public disclosure
- Do not exploit vulnerability beyond proof-of-concept
- Do not access other Tenants' data
- Coordinate public disclosure with OpenISP
Recognition:
- Vulnerability credit in release notes (if consenting)
- Bug bounty payment (based on severity)
- Swag/merchandise (as thank you)
14.3 Emergency Contact
For critical security incidents:
- Email: [email protected] + call ops team
- Incident response team mobilized within 15 minutes
15. SECURITY TRAINING & AWARENESS
15.1 Employee Security Training
All OpenISP staff receives:
- Onboarding security training (within first week)
- Annual security refresher
- Role-specific training (developers, DBA, ops, support)
- Phishing simulations (monthly, with feedback)
- Data protection & GDPR training (annual)
Topics:
- Password management & MFA
- Phishing & social engineering
- Malware & ransomware
- Data classification & DLP
- Secure coding practices
- Incident response procedures
- Customer privacy & confidentiality
15.2 Tenant Security Guidance
OpenISP provides documentation:
- Security best practices guide
- MFA setup instructions
- API key rotation guide
- Backup & disaster recovery guide
- Subscriber communication templates
Resource: support.openisp.io/security
16. SECURITY POLICY UPDATES
16.1 Policy Review & Updates
This policy is reviewed:
- Annually (formal review)
- After security incidents (immediate update)
- When regulations change (within 30 days)
- When new threats emerge (urgent update)
Notification: Material changes communicated to Tenants via email.
END OF SECURITY POLICY & DATA PROTECTION
Effective Date: [INSERT DATE]
Last Updated: September 13, 2026
Version: 1.0
APPENDIX: Security Checklist for Tenants
Pre-Production Checklist
- Enable MFA on all admin accounts
- Change default RADIUS secret (generate 24+ char random)
- Rotate all API tokens
- Verify WireGuard private keys are secured
- Configure firewall to restrict router admin access
- Review and customize Subscriber privacy policy
- Set up HTTPS certificate (valid, not self-signed)
- Enable database backups (outside OpenISP)
- Test backup restoration procedure
- Document incident response plan
- Train staff on security procedures
- Set up centralized logging for your equipment
- Configure rate limiting on payment endpoints
- Implement DPI/traffic shaping policy (if desired)
- Get cyber insurance (recommended)
Ongoing Maintenance
- Rotate API tokens monthly
- Rotate RADIUS secrets quarterly
- Review access logs weekly
- Check for failed authentication attempts
- Update router firmware as released
- Monitor for security advisories
- Test backup restoration (quarterly)
- Review Subscriber account changes (inactive accounts)
- Audit admin actions (who made configuration changes?)
- Refresh staff security training (annually)