IT continuity plan: what to do when everything stops

A server goes down, a ransomware attack locks every file, or your internet provider has a multi-hour outage — all common scenarios, yet few companies have a written plan for them. An IT continuity plan does not prevent a disaster; it shortens the time it takes to get back to normal: what you restart first, who makes decisions, and where you find emergency passwords when the main system is down.

RTO and RPO, explained simply

RTO (Recovery Time Objective) is the maximum acceptable time to restart a system after an outage — for example, "the invoicing server restarts within 4 hours at most." RPO (Recovery Point Objective) is how much recent data you can afford to lose — if you back up once a day, your RPO is 24 hours, meaning you could lose a full day of work. These two numbers, set in advance for every important system, are the backbone of any continuity plan. For example, a one-day RTO may be acceptable for an internal file server, while an online invoicing system needs an RTO measured in hours.

Realistic scenarios, not just theory

A good plan starts from concrete scenarios, not a generic checklist:

• Server failure — a hardware fault cuts off access to files and applications, sometimes for a few hours, sometimes for days, depending on spare part availability.
• Ransomware — data gets encrypted and a ransom is demanded for decryption, and paying it does not guarantee recovery.
• Fire or flooding — physical equipment becomes unusable, no matter how well it was maintained.
• Internet provider outage — the office loses connectivity for hours, and anything that depends on the cloud stops working.
• A key employee is unavailable — the only person who knows the admin password is on sick leave or has left the company.

These are not hypothetical situations — they happen at small companies just as often as at large ones, only the relative impact is bigger: a company with 12 employees feels a day of downtime far more than a corporation with redundant departments. For each one, the plan needs to answer a single question: what do we do in the first 30 minutes?

The order in which systems come back online

Not every system restarts at the same time — nor should it. A retail company, for instance, restarts invoicing and the internet connection first, then the file server, and only later the secondary applications. The order is set in advance, based on what actually stops the business and what can wait a few hours without harm. Document this order once, in a simple document that non-technical staff can also read, so anyone involved in a crisis knows what comes next, not only the system administrator.

Where to keep the contact list and emergency passwords

In a real outage, passwords stored in a password manager installed on the failed server help no one. The continuity plan needs an offline copy, kept in a safe physical location (or a separate account, reachable independently of the main infrastructure), with: the IT provider's contact, the internet provider's contact, the server administrator account, and who makes decisions when the manager is unreachable. Without this document, the first hours of a crisis go into figuring out who remembers a password. Ideally, this list is updated every time a provider or a critical password changes, not once every three years, when no one can tell if the information is still valid.

Why an untested plan is not a plan

A plan written and left in a drawer is a list of intentions, not a working plan. Many companies only discover during a real crisis that a backup does not restore correctly, or that a written-down password is outdated. An annual restore exercise — rebuilding a system from a backup copy, timed, in a test environment — shows exactly where the plan has gaps, before it actually matters. The exercise does not need to be complicated: pick one important system, restore a backup copy in a separate environment, time how long it takes, and compare the result against the RTO you set.

How to start

Start with a security audit that shows the real weak points of your infrastructure. Continue with backup and disaster recovery, the technical part of the plan, and with continuous monitoring, so you learn about a problem before your customers do. Contact us to build the plan together.