A network software problem was behind Monday’s disastrous system crash at the London Stock Exchange.

A network software problem was behind Monday’s disastrous system crash at the London Stock Exchange.

But the LSE said it would not point the finger of blame at Accenture, which built its crucial TradElect platform. It said TradElect was not at the source of the problem.

Trading was stopped on Monday until the final half hour of activity, after a connectivity problem meant the LSE was forced to take all traders offline until the system was rectified, with only 30 minutes’ trading left in the day. The LSE has begun an investigation into the cause of the problem, but has not disclosed any more information.

Angry traders had walked out of the building in protest, unable to conduct business on what would have been one of the best days for the market as stocks rallied after the bail out of US mortgage lenders Fannie Mae and Freddie Mac.

An LSE spokesperson told Computerworld UK the problem was related to network software, and not the LSE’s key TradElect platform. He also said that an undisclosed “fix” had been applied to the software and that the LSE was “confident” the systems would now function properly.

“Two particular software activities” had taken place, causing the problem, the LSE admitted. But it categorically refused to disclose any more information.

Last year, it moved IT management in-house and away from supplier Accenture.

But Accenture still provides software development services to the exchange, and built the LSE’s TradElect platform, which the exchange said was not the cause of the problem. The platform runs on HP ProLiant Servers and Microsoft .Net and SQL Server systems.

Accenture would not comment, and referred all queries to the LSE.

The problems are “thought to have occurred” on the trading gateway between the LSE’s Extranex private network (linking the exchange and clients) and the TradElect electronic trading platform, the Financial Times said.

The LSE moved to quash speculation that the meltdown was due to heavy trading, saying it was “not about volumes”.

But observers were not convinced. Bob McDowall, analyst at Tower Group, said the problems were “clearly attributable to volumes” and that the LSE’s explanation was “not very satisfactory”. He said the exchange needed to explain the situation “in a non-technical way to all stakeholders”.

The exchange has been extensively upgrading its systems in order to stave off new competitors, such as Chi-X and Turquoise. But it denied rumours that system upgrades had caused the problem. The issues were “not upgrade-related”, the spokesperson said.

McDowall at Tower Group said the technical issues would not help the LSE’s appearance against its competitors. “The LSE’s competitors are only competing on the top 100 to 150 stocks,” he said. “But over time there’s a real threat.” The LSE was “not doing any good by not explaining the problem”, he said.

Industry participants also said better testing was needed. Performance monitoring firm Macro 4 said proper load testing was vital to make sure such critical systems work under stress, but added 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 this”.

Macro 4 said it was important to identify problems in the code which joins all the parts of the system together, including databases, software and servers, because “this is often where many of the glitches actually reside”. All the managers of the different system components would need to collaborate to identify and fix problems, Macro 4 said.

Now read

London Stock Exchange in denial over outage