Typically, when we think of airlines experiencing data loss due to downtime we tend to focus on the chaos that this scenario creates at the front desk or at the gate. Not being able to access, and even losing, customer booking data certainly sounds like an airline’s worst nightmare.
For example, in 2017 Delta Airlines had two system-wide outages within in a six month period – costing the company a reported $100 million dollars. Delta isn’t the only airline that has experienced IT failures though. In that same year Southwest Airlines cancelled more than 2,000 flights after a router went down, costing them estimated $54 million.
So Why is Preserving In-Flight Data So Vital?
Data loss has become increasingly more than just a customer experience crisis for airlines. During a flight there is a significant amount of data being generated. So much so that NASA has partnered with Southwest Airlines to develop a data mining application that will allow the airline to manage it all.
A NASA press release states that, “When an airplane flies, hundreds of data streams fly from it every second – pilot reports, incident reports, control positions, instrument positions, warning modes.” This is data that safeguards the lives of passengers, and when looked at predictively can, “find issues before they become incidents.”
Stratus’s Downtime Prevention Buyer’s Guide explains that the reason downtime negatively impacts all this rich, in-flight data is because, “When a system outage occurs all data and transactions not yet written to disk are at risk of being lost or corrupted.” This means data that could be used to predict and prevent future catastrophes may be lost forever.
The guide recommends that when installing or upgrading technology, airlines should ask the question, “How does your solution preserve in-flight data?” Stratus stresses that while failure may be “tolerable” for some applications, “the loss of in-flight data can have serious consequences ranging from a scrapped batch or lost revenue to compliance issues or even loss of life.”
“Many availability solutions are not designed to ensure transaction and data integrity in the event of a system failure. Depending on how the hardware is configured, standalone servers and high availability clusters can typically preserve the integrity of database transactions, but any in-memory data not yet written to disk will be lost upon failure.”
The guide goes on to say that, “Fault-tolerant solutions are built from the ground up to provide higher levels of data integrity. Fully replicated hardware components and mirrored memory ensure that all in-flight transactions are preserved — even when a hardware component fails.”