An iPrimus datacentre in King St, Melbourne, suffered a complete power outage on Sunday, including failure of the emergency diesel backup system.
The outage resulted in numerous ISPs issuing statements to businesses and residential customers explaining why their Internet connections and VoIP services were down for several hours.
For ISP, Internode, the power failure that occurred about 2pm left all Victorian corporate Ethernet customers without connectivity and DSL customers with higher than normal congestion and latency. All services in Tasmania were unavailable.
Internode managing director, Simon Hackett, said power failures in capital cities are relatively rare and that Internode, along with other ISPs, were just tenants at the Primus datacentre. The primary failure modes, therefore, were beyond their capacity of control in this case.
"Our core data centres in South Australia are directly built, managed, and operated by Internode, and the main one actually uses two generator sets to back up mains power, with automatic 'failover' should one of them not start up during a power outage. It appears Primus have a single generator set in their Melbourne facility which (obviously) didn't start up properly on this occasion," he told ARN in an email, adding that the ISP was awaiting a post incident report from Primus before deciding what additional steps to take in response to the outage.
"The presence of multiple carrier and path diverse inter-capital links, and the use of multiple data centres means that the loss of access to that datacentre did not lead to connectivity issues outside of Melbourne (which it may otherwise have done), and it also substantially reduced the disruption to customers in Melbourne and Victoria for us, compared to the experience of customers using an ISP with only one point of presence in that state.
"The Primus datacentre concerned is a major communications site for much of the industry, so you can rest assured that we'll be ensuring, with Primus, that this doesn't become a habit. We don't expect that it will," Hackett said.
PIPE Networks CEO, Bevan Slattery, said on the Whirlpool forum that all PIPE Networks services within the facility, including peering, telehousing and Ethernet, had been affected.
“Details are still being determined, however, there appears to have been a loss of mains power, combined with a failure of backup diesel generation, and therefore the depletion of the UPS batteries,” he wrote, before the issue had been resolved.
Slattery told Whirlpool readers he had sent PIPE Network’s national head of fibre operations, technicians and engineers to Melbourne to help investigate the cause of the outage. ARN understands he is still on-site and is unavailable for comment.
Several other threads posted Sunday on Whirlpool indicated that customers with Westnet, iiNet, Netspace, Internode, and TPG nationwide were also among those affected, however Victoria and Tasmania suffered the worst outages.
According to Internode, the datacentre’s power provider, CitiPower, had returned power by about 6pm.
"Had the power outage been for only a few minutes, the Primus battery systems would have kept the centre running. Had that facility been destroyed rather than just being out of power, we would have moved our entire operations to our own second site in Melbourne. But an outage for a few hours means the most rational approach is to re-start our systems 'in place' as soon as practicable - which is what we did," Internode's Hackett said.
iPrimus and iiNet are yet to respond to requests for comment.