Servers Australia has instigated an emergency -- and permanent -- data centre move for customers affected by an outage at a Vocus Communications data centre in Sydney after power issues hit the site early on 13 February.
As reported by media outlet, iTnews, Vocus Networks’ Alexandria data centre ran into trouble just before 7AM, with the telco advising customers that a power outage occurred in a core point of presence (POP) in Sydney due to unsuccessful uninterruptible power supply (UPS) maintenance.
The company said that power had been restored to the site, but that ongoing network connectivity issues were still being experienced due to the event, and that Vocus engineers were en route to the site to continue investigating.
By 8AM, Vocus told customers that services were gradually being restored, and that multiple engineers were on-site troubleshooting and restoring services.
By 11:50AM, Vocus said that engineers were working to restore ADSL services connected to its Ipera Network. The company said it believed services should be restored by the end of the business day.
At around 10AM, however, Servers Australia managing director, Jared Hirst, took to telco industry online forum, Whirlpool, to express his disappointment over the outage and that his company would be enacting an emergency move for any customers in the affected site.
Moreover, Hirst said that Servers Australia will be vacating the Vocus site entirely within the next few weeks.
“Unfortunately, there has been a complete loss of power at the Vocus Data Centre whilst they had scheduled maintenance this morning. All A and B power feeds have been lost for a brief 1 minute,” Hirst wrote.
“All Servers at this site and switches have A and B power to avoid outages, especially during maintenance. I am extremely disappointed that this has happened. This is a very small site for Servers Australia, as most servers are located at Equinix or our own DC's.
“I will not tolerate a loss of both A and B power under any circumstances, as this means that there is NO redundancy,” he said.
Hirst told ARN that the emergency move, which will see the relocation of at least 16 racks, will not be temporary, and comes after a series of issues with the Vocus site.
"It’s riddled with power and cable management [issues] and has been for some time," Hirst said. "We have sustained many outages here in the past 12 months due to poor cable management and dirty power. Our plan is to move everyone in the next three months."
On its own website, Servers Australia network engineer, Jaden Roberts said that the company was working with Vocus to obtain information on the root cause of the outage.
“In the meantime, Servers Australia has immediately made emergency space available to those customers wishing to move from this facility immediately,” he said. “We understand that this facility has not been stable, and we understand that outages are not acceptable.”
Servers Australia said that it expects to provide a full incident report within seven days.
“The majority of services have been restored though we are still working through the list of remaining affected services,” Roberts said in a status update posted just after midday.
Updated at 15:45 to include additional comment by Jared Hirst.