Skip to main content
Institute of Advanced Technological and Commercial Studies

Resources

Service continuity and the recovery plan

A continuity plan answers two quantified questions: how long the service may stay unavailable, and how much data may be replayed. Those two values drive the whole architecture that follows.

Two values decide everything else

Acceptable downtime states how long the business holds without the service; acceptable data loss states which backup point you are willing to fall back to. An organisation that sets those two figures with the business immediately holds the specification for its architecture: backup frequency, equipment redundancy, presence of a fallback site, power autonomy. Without them, any sizing is debated endlessly.

  • An acceptable downtime, agreed with each business unit
  • An acceptable data loss, which sets backup frequency
  • An off-site backup whose restore is dated and signed
  • A written switchover procedure, rehearsed at least once a year
  • Sized power autonomy, uninterruptible supplies and generator included

The switchover drill, the only check that counts

  1. 01

    Announce the window

    A drill is scheduled with the business. Announced, it becomes training rather than an event endured.

  2. 02

    Actually switch over

    The service restarts from the fallback site. A paper-only switchover leaves every surprise intact.

  3. 03

    Time it

    Actual time is compared with acceptable downtime. The gap becomes the year's action plan.

  4. 04

    Return and record

    Returning to the primary site is part of the drill, and the report stands as evidence to an auditor.

The programmes that build this skill

Infrastructure, security and power converge on continuity. All three programmes approach it from their own craft, with taught volumes as published in the catalogue.

ProgrammeLevelDurationCredits and hoursFees
Networks and CloudBachelor's degree2 years120 credits · 1200 h5 600 000 GNF per year
Cybersecurity programme — Bac+5 levelMaster's degree2 years120 credits · 1200 h7 500 000 GNF per year
Industrial Maintenance and EnergyBachelor's degree2 years120 credits · 1200 h5 400 000 GNF per year

Continuity questions

How often should a restore be tested?

Quarterly for critical data, annually for everything else. A dated, signed test is the only proof that a backup genuinely exists: a file sitting on a disk is not one.

Do I need a second site?

The answer follows from acceptable downtime: a few hours makes it necessary, a few days allows rebuilding from backups. That is why those two values are set before any architecture decision.

How is acceptable downtime set?

With the business, by quantifying what one hour without the service costs: suspended orders, halted production, obligations falling due. That costing turns a matter of opinion into a documented decision.

Service continuity and the recovery plan — Resources | IHETC — IHETC