Skip to content
BENIBIX Learn

Data Downtime

Also known as Data Incident, Data Outage

Time when data is missing, late or wrong. It counts from the break, not the alert.

Think of it like

It is like a clock that stopped in the night. It still shows a time, so people trust it.

Example

A load breaks at 02:00. A manager spots a blank chart at 09:00, and it is fixed by 11:00. That is 9 hours of data downtime, and 7 of them went by with nobody knowing.

Why it matters

Every hour of downtime is an hour of choices made on bad numbers. Measuring it shows where to improve.

Definition

Data downtime is the time when data is missing, late or wrong. The tables still exist and the dashboards still load, so it is easy to miss. It starts when the data breaks, not when someone notices. It ends when the data is fixed. So the total is the time to find it plus the time to fix it. Data observability tools exist to shrink the first part.
Nine hours of downtime, and seven of them passed before anyone knew. Faster alerts shrink that part.

Common mistake

Counting only the time spent fixing. The hours before anyone noticed are usually the biggest part.

Key takeaways

  1. 1
    Missing, late or wrong. All three count as down.
  2. 2
    From the break. The clock starts when it breaks.
  3. 3
    Find it faster. Alerts cut the longest part.