Software bug caused LSE outage

gallery

Since electronic systems for trading on the London Stock Exchange (LSE) ground to a halt for most of yesterday, speculation has been rife as to the cause of the outage.

An LSE spokesperson told IT PRO that it had identified the cause of the issue, taken remedial action and was confident the issue would not reoccur.

He refused to be drawn much further on the specific cause other than to say: “It was actually a software issue, and not one to do with hardware as some in the media have suggested. This is why we couldn’t switch over to our backup hardware.”

But an IT performance management expert said a tighter focus on key application development, monitoring and management processes could help to prevent IT failures like the LSE’s from occurring in the first place.

David Chalmers, Macro 4 director of product strategy, observed that traders reportedly said the exchange was suffering an “overloading” caused partly because the FTSE 100 ended down sharply last Friday, forcing traders who were betting on further falls in the index to close their positions.

But he said that it can be extremely difficult and time consuming to generate the required artificial ‘load’ in order test IT systems for high volume activity like that generated by last Friday sharp index decline.

“In a similar example from earlier this year Heathrow’s terminal five baggage handling system was reportedly only tested at full load 20 times and ‘the testing regime did not adequately replicate the first days of operations,’” he added.

Chalmers also said the current generation of application performance monitoring systems were good at monitoring the performance of individual components, like databases, software and servers. “But they are poor when it comes to identifying problems in the code which joins all the parts together. And this is often where many of the glitches actually reside,” he said.

And he stressed that proper test procedures and adequate monitoring were false economies if communication at systems management level is not effective.

“In many big organisations different aspects of an IT application – networks, databases, storage – are looked after by different managers and monitoring systems. It can take time to get all these specialists together and it is often difficult get anyone to admit it’s their part of the system that could be at fault,” Chalmers added.

Email to a friend

Print this page