Over four million motorists across NSW were unable to access the RTA's myRTA.com self-service portal for nearly two days due to a core systems upgrade, Computerworld has learnt.
Sources familiar with the outage say the RTA posted a message on its Web site notifying the public that myRTA.com services are "temporarily unavailable".
"MyRTA.com services will be unavailable from 02:00 pm Saturday 28 May until 06:00 am Monday 30 May," the notice read. "This is to allow for system maintenance. Normal services are expected to resume after the outage. We apologize for any inconvenience caused."
Contrary to the "system maintenance" message, CIO Greg Carvouni told Computerworld the RTA is in the process of changing data centres and core systems making downtime "unavoidable".
"There are many complex database issues [associated with the upgrade]," Carvouni said. "This was a one in three- or four-year event."
Conceding the job was not some routine 'maintenance', Carvouni said the message on the Web site was not as clear as it could have been but the public don't want to know about core IT operations.
"This is the last stage of the implementation and is expected to be complete in three months," he said, adding that the commissioning of two Sun E6900 systems is part of the upgrade.
When asked if the upgrades affected the business-critical DRIVES transaction processing system, Carvouni said there are ways of making core systems available.
The downtime comes within three months of the RTA announcing a three-year extension to its outsourcing contract with Fujitsu for its applications and data centre operations.
Under the new contract, Fujitsu will continue to host about 800 corporate applications for the RTA - including the DRIVES information management system for vehicle licensing and registration - from two "state-of-the-art" data centres in Sydney, according to a media statement by Fujitsu in March.
Ironically, the same media release has Carvouni quoted as saying the new contract will "ensure that our IT systems continue to operate at the very high levels of reliability necessary to help manage the safe and efficient use of the road transport network".
To avoid downtime during disruptive changes or unexpected failures, an appropriate level of disaster recovery is usually planned and executed. The RTA chose not to keep the lights on during the upgrade.
However, independent IT analyst Kevin McIsaac said the amount an organization spends on disaster recovery is dictated by the importance of the service.
"Cleary-defined SLAs (service-level agreements) are typically not done well and if no implicit SLA was given by the RTA [for myRTA.com] it can't be expected to run non stop," McIsaac said. "If it was Internet banking or an emergency service you expect those to be always on."
McIsaac said when disaster recovery strategies are put in place costs go up significantly so end users need to set realistic uptime expectations. "What do you save by not having a redundant system?" he said, adding that it may have been a known quiet period for myRTA.com.
Allowing for a few days per year of downtime to avoid costly 'always on' disaster recovery may be a rational business choice, McIsaac said.