IT Glossary
What is failover?
The mechanism by which a standby system automatically takes over the work when the primary one fails.
Aircraft carry two engines not because one is unreliable but because the consequence of a single failure is unacceptable. Failover applies the same logic to business systems: a second server, database or data path sits ready, health is monitored continuously, and when the primary stops responding the standby takes over the work without waiting for a person to notice. Done well, the switch happens in seconds and customers see nothing but a slow page. Done badly it happens at three in the morning, to a standby nobody had tested since installation, and the outage lasts longer than it would have without the mechanism. The distinction that matters commercially is between failover and backup, which are frequently confused in vendor proposals. Backup protects the data and is measured in how much you can afford to lose and how long restoring takes. Failover protects availability and is measured in how many seconds the takeover costs. You need both, they cost different amounts, and buying one while believing you bought the other is the most expensive misunderstanding in the whole area of business continuity.
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247
Why it matters for your business
Continuity without human intervention
The switch happens automatically at the moment of failure, rather than after someone notices, finds the runbook and reaches a keyboard.
Major maintenance without stopping
Upgrades and patching run on one side while the other serves traffic, so security work stops competing with trading hours.
A verifiable commercial argument
Availability commitments to enterprise customers become demonstrable with test evidence rather than asserted in a proposal.
Frequently asked questions
What does active-passive mean compared with active-active?
In active-passive, one system serves everything while the standby idles, ready to take over. It is simpler and cheaper, and the switch costs seconds to minutes. In active-active, both handle live traffic and losing one merely reduces capacity, giving near-instant continuity at higher cost and considerably more design complexity, particularly around keeping data consistent between the two.
Is data lost during a database failover?
It depends on how replication is configured. Synchronous replication confirms each write on both nodes before acknowledging it, so nothing is lost but every transaction pays a latency cost. Asynchronous replication is faster and can lose the last few seconds of writes. Decide this against the value of a transaction rather than by default, and write the chosen figure into the continuity plan.
How often should failover be tested?
At least twice a year, and always after significant infrastructure changes, with the results written down. Untested failover is a belief rather than a capability, and the failure modes are consistently mundane: an expired certificate on the standby, a firewall rule nobody replicated, a configuration file that drifted. Scheduled tests find those while a calm engineer is watching.
Let’s talk about your project
Message us on WhatsApp or send an email — you talk directly to a developer.
office@northdan.com · +40 752 070 247