Skip to content
TheChatArchiveThe Chat Archive
DeepSeekdeepseek-r1-0528· DeepSeekHugging Face LMArena Human Preference 140K

can you create me an isp disaster recovery plan

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with Coherecommand-a-03-2025’s answer.
  1. User

    can you create me an isp disaster recovery plan

  2. DeepSeek Assistant

    Comprehensive ISP Disaster Recovery Plan (DRP)


    1. Introduction & Objectives

    Purpose: Ensure rapid restoration of critical ISP services (internet, voice, cloud) during disasters (natural, cyber, infrastructure failures).
    Scope: Network infrastructure, data centers, customer support, core systems (BGP, DNS, DHCP).
    Goals:

    • Max 4-hour recovery for critical services (RTO)
    • Max 15-min data loss (RPO) for core systems
    • 99.99% uptime post-recovery

    2. Disaster Recovery Team (DRT)

    RoleResponsibilitiesContact Info
    DR CoordinatorOverall plan execution, coordination24/7 phone
    Network EngineersRestore routers/switches, BGP, fiber linksOn-call roster
    Systems AdminsServer/cloud recovery, DNS, DHCPSlack/phone
    CS LeadCustomer comms, outage updatesEmail/SMS
    Vendor ManagerCoordinate with hardware/bandwidth providersDedicated hotline

    Activation Triggers:

    • Data center power failure >15 mins
    • Fiber cut affecting >50% users
    • Ransomware attack

    3. Risk Assessment & Business Impact

    Top Risks:

    • Fiber cuts (60% impact): Use diverse paths + satellite backup.
    • DDoS attacks: Deploy scrubbing centers (e.g., Cloudflare).
    • Data center fire: Geo-redundant DCs (min 500 miles apart).
    • Power grid failure: Tier-4 UPS + generators (72-hr runtime).

    Critical Systems:

    SystemRTORPOImpact Severity
    Core routing1 hr0 minsCritical ($$$)
    Customer DB4 hrs15 minsHigh ($$)
    Billing24 hrs1 hrMedium ($)

    4. Recovery Strategies

    Network Redundancy:

    • BGP: Multi-homed with 3+ upstream providers.
    • Last-mile: Wireless backup (LTE/5G) for fiber cuts.
    • PoPs: Automatic failover to secondary PoP if latency >100ms.

    Data & Systems:

    • Real-time replication: VM snapshots to DR site (every 15 mins).
    • DNS: Anycast DNS with global nodes (AWS Route 53).
    • Cloud: Hybrid cloud failover (On-prem + AWS/Azure).

    Infrastructure:

    • Power: N+1 generators, ATS switches, dual-grid feeds.
    • Cooling: Redundant chillers with temp sensors.

    5. Communication Plan

    Internal:

    • Alerting: PagerDuty for DRT alerts + War room (Microsoft Teams).
    • Escalation: Automated SMS/email after 30-min non-response.

    External:

    • Customers: Status page (e.g., Statuspage.io), SMS, social media.
    • Partners: Direct NOC hotline for carriers/vendors.
    • Media: Pre-approved press releases (CEO sign-off).

    Template:

    "Incident [DD/MM HH:MM]: [Issue] affecting [Area]. ETA for resolution: [Time]. Updates: [URL]"


    6. Data Backup & Recovery

    Backup Schedule:

    • Core configs: Hourly versioned backups (Git).
    • Customer data: 15-min incremental (SAN replication).
    • Full backups: Daily encrypted offsite (AWS S3 IA).

    Recovery Process:

    1. Validate backup integrity (hash checks).
    2. Restore VM snapshots from DR site.
    3. Verify data consistency (automated scripts).

    7. System Recovery Procedures

    Network Restoration:

    Step 1: Failover traffic to secondary PoP via BGP prepending.  
    Step 2: Deploy pre-configured routers/switches (Cisco/Juniper).  
    Step 3: Test latency/jitter (<50ms, <5ms variation).  
    

    Cyberattack Response:

    • Isolate infected segments.
    • Restore clean backups.
    • Forensic audit pre-reconnection.

    8. Vendor Management

    Vendor TypeSLA RequirementsContact Protocol
    Bandwidth2-hr rerouting guaranteeDedicated NOC line
    Hardware4-hr emergency deliveryWeb portal + hotline
    Cloud99.995% uptimeAPI-driven alerts

    9. Testing & Maintenance

    Testing Schedule:

    • Quarterly: Simulated fiber cut/DDoS (4-hr window).
    • Biannual: Full DR site failover (12-hr downtime).
    • Annual: "Blackout" test (no power/cloud, 24 hrs).

    Metrics:

    • Recovery time deviation (<10% of RTO).
    • Backup success rate (100% target).

    10. Training & Awareness

    • DR Drills: Role-based scenarios for DRT (every 6 months).
    • New Hire Training: Mandatory DR module (e.g., KnowBe4).
    • Customer Awareness: Outage FAQ + prep tips (e.g., UPS for modems).

    11. Plan Review & Updates

    • Review: Quarterly by DR Coordinator + CTO.
    • Update Triggers:
      • New infrastructure deployment.
      • Failed DR test.
      • Post-disaster analysis (within 72 hrs).
    • Version Control: SharePoint + physical copies at DR sites.

    Approvals:

    • CTO: ___________________ Date: _________
    • DRT Lead: _______________ Date: _________

    [Customize with ISP name, contact details, and technical diagrams]

    Key: Prioritize automation (scripts for failover/backup) and partner SLAs. Test relentlessly!

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.