Technical Glitches Disrupt Online Exams: Why Students Want Vendor Changes—and How to Stay Exam‑Ready

- What went wrong: Widespread complaints of system failures, abrupt cancellations, and poor centre management triggered nationwide protests under #SSCMisManagement. Read more here: SSC Exam 2025 Glitches

- Re‑exam announced: SSC confirmed a retest for ~55,000 candidates on August 29, 2025, after detecting inconsistencies in exam log data.

- Ripple effect: SSC deferred CGL 2025 to September, citing platform review and operational checks

- Day‑2 disruptions (May 15, 2025): Technical failures in Delhi and J&K, including cancellations in Srinagar; NTA moved to reschedule affected candidates.

- California Bar Exam (Feb 2025): Severe platform and proctoring issues led the State Bar of California to sue its vendor (Meazure Learning); the vendor later moved to dismiss fraud claims. The saga has spurred legal and legislative scrutiny of vendor claims and exam tech readiness.

1. Capacity & Load Miscalculations : Under‑provisioned servers or weak autoscaling can cause timeouts, login failures, and frozen screens during peak loads.

2. Insufficient Redundancy: Lack of geo‑redundant data centres, power backups, or secondary proctoring flows turns local faults into mass failures.

3. Fragile Centre Operations: Device misconfiguration, patching lapses, or bandwidth bottlenecks at centres cascade into last‑minute cancellations.

4. Opaque Incident Management : Delayed alerts, unclear reschedule rules, and slow refunds deepen candidate anxiety and erode trust.

5. Vendor Overselling : Marketing claims about uptime/capacity that aren’t backed by audited stress tests—an issue that’s now under legal spotlight elsewhere too.

1. Preparation windows shrink when exams move or re‑tests cluster too close to other papers.

2. Travel, lodging, and day‑off costs pile up after cancellations.

3. Confidence dips after a botched first attempt—even when re‑exams are promised.

4. Equity concerns rise when disruptions are centre‑ or region‑specific (e.g., J&K), disproportionately affecting certain cohorts.

1. Independent Tech Audits before each major cycle (infrastructure, DDoS posture, peak concurrency, failover drills).

2. Transparent Stress‑Test Reports published pre‑exam (target concurrency, error budgets, performance baselines).

3. Dual‑Vendor or Dual‑Track Options for critical phases (e.g., staggered providers or hybrid CBT + OMR fallback).

4. Centre Accreditation 2.0 (mandatory bandwidth tests, device baselines, invigilator tech training with certification).

5. Live Status Dashboards (uptime, incident ticker, regional health) are visible to the public during the exam window.

6. Clear Retest & Refund SLAs auto‑triggered by incident severity thresholds.

7. Candidate Protection Protocols (extra attempts if tech fault >X minutes, provisional scoring where feasible).

8. Real‑Time Grievance Channels (WhatsApp bot + hotline + email with ticket IDs; 24–48 hr resolution windows).

9. Post‑Mortem Reports within 7–10 days, with corrective actions and vendor penalties where due.

10. Legally Binding Performance Clauses tied to payouts; recurring SLO breaches trigger blacklisting or re‑tendering.

1. Arrive early (or log in early for remote) to buffer for checks.

2. Carry redundant IDs and print backups of admit cards.

3. Cross‑check centre bandwidth rules; if remote, ensure stable power + mobile hotspot ready.

4. Know escalation paths: helpline, on‑site supervisor, e‑mail template for grievances.

5. Document issues (timestamps, screenshots, invigilator name); keep proof for re‑exam eligibility.

6. Pack essentials: water, transparent stationery kit, prescribed meds, spectacles.

7. Sleep and nutrition routine 48 hours prior—cognitive stability matters.

8. Rehearse CBT navigation via mocks to reduce time lost to UI unfamiliarity.

9. If an incident occurs: within 7–10 days, with corrective actions and vendor penalties where due.

10. After disruption: file a complaint within published timelines; track ticket ID.

11. Switch to recovery mode: resume revision schedule; treat re‑exam as a fresh attempt.

12. Lean on your coaching network for updates, logistics help, and mock/test recalibration.

1. Verified load testing results at ≥2× expected peak concurrency.

2. 99.9%+ uptime SLOs during live windows; penalty ladder for breaches.

3. Geo‑redundant data centres with automatic failover; RTO/RPO targets defined.

4. Strict proctoring SLAs (response time, handoff rules).

5. Offline fallback (local caching + buffered sync) to salvage attempts if short outages occur.

6. Centre readiness audits (bandwidth, hardware, power backups)—publish pass/fail lists pre‑exam.

7. Incident classification matrix with scripted playbooks and public dashboards.

8. Auto‑refunds / auto‑retests triggered by Class‑A incidents.

9. Security posture: DDoS mitigation, zero‑trust endpoints, patch timelines.

10. Accessibility compliance (screen readers, adjustable fonts, colour-contrast, extra time).

11. Candidate supports SLAs (multi‑lingual, 24×7 during windows).

12. Data protection (encryption, retention, deletion on schedule).

13. Independent audit rights (before, during, after exam).

14. Whistle‑blower channel for centre staff and candidates.

15. Exit plan & knowledge transfer if the vendor is replaced.

- Map likely centres (Mukherjee Nagar, Pitampura, Dwarka, Noida border, Rohini belt) a week in advance.

- Trial commute at exam time to gauge traffic buffers.

- Keep a “tech kit” power bank, extra mask, pen drive if permitted, and prescription glasses.

- Have a coaching fallback: if your centre cancels, check your batch’s WhatsApp group for immediate retest guidance—your institutes above usually post verified updates first.

- If you’re in West/North‑West Delhi, batch at Rajouri Garden / Uttam Nagar / Pitampura to shorten the day of travel risk.