Primus completes $2m upgrade to fight outages

By Munir Kotadia on Aug 17, 2009 5:24 PM
Filed under Hardware

Bullet-proof from now on?

Following three service outages so far this year, Primus Telecom over the weekend completed an upgrade to the power systems of its Melbourne data centre.

Primus Telecom CEO Ravi Bhatia said the company decided to upgrade its systems following an outage in February.

Since then, customers have reported outages on at least two further occasions, most recently two weeks ago.

"The first time we had an outage, which was at the beginning of February, it was extremely concerning to us and the company opened its wallet and decided to fix it," Bhatia told iTnews.

He said the most recent outage was "truly unfortunate" because it happened just a week before a scheduled upgrade of the facility's power systems.

An earlier upgrade in May was said to have made the Primus data centre "beyond bulletproof".

However, today Bhatia warned that a multitude of factors were combining to threaten Australia's ability to feed its unquenchable thirst for electricity.

"Most of our power plants are close to the end of their design life; a lot of these utilities are being privatised and there is significant uncertainty in the future because of the carbon trading scheme. High voltage and high power is a subject [this IT community] does not understand," Bhatia said.

Primus completes $2m upgrade to fight outages

7 comments in this discussion

""High voltage and high power is a subject [this IT community] does not understand," Bhatia said." Perhaps we understand the need for it quite a lot better than some of the relevant corporate ..."

By anonymous

Tags

primus, outage, power, hosting, fuse, ravi, bhatia, datacentre

AAPT loses power in battery swapout

Enterprise Data upgrades Sydney1 data centre

NetRegistry services hit by DDoS attack

Firmware update blamed for Office365 outage

Breaking Stories

Roses Only goes with its heart for hosting

Bankwest credits IT for flat operating expenses

Conroy warns of cloud uncertainties

Hiring strategy vital to government IT transformation

Melbourne scientists ready 2013 bionic eye trial

Comments: 7

DJ Aug 18, 2009 8:47 AM	So, what Bhatia is saying is that he is running an enterprise datacentre but has no clue about power. How on Earth can you continue to blame everyone else, when you (Primus) are responsible for your datacentre ? The cost to your customer's businesses from these outages - which could have been avoided - is more than just downtime, it's reputation and and financial, not to mention the wasted time dealing with customers who want to know why their systems are down again for the third time. Not good enough Primus. It's now clear you did not perform monthly maintenance, generator inspections or UPS testing. Busted.
Bob Aug 18, 2009 9:41 AM	Completely agree. There seems to be a consistent attitude of blame someone else for the inability of providers to perform as stated. Primus blames electricity companies for multiple outages, Optus blames Apple for it's bottom line and everyone blames Telstra for not spoon feeding them. If you can't provide a quality service profitably then get out of the way and stop complaining if someone else can.
megamouse Aug 18, 2009 4:54 PM	What a load of ...... Power is the lifeblood of ANY IT facility. I was ops manager for a national datacentre that operated flawlessly through the Newcastle earthquake. How did we do it? simple, install a proper UPS with batteries and our own diesel generator big enough to hold the entire datacentre plus 50%. It was tested monthly by simply flipping the main building breaker and allowing the system to kick in. If your system can't cope with this, then it ain't bulletproof! We ran the genset for four days nonstop and did not lose a single transaction. Too easy when it's done properly. P.S. I recommend that a datacentre should NEVER be located in any high rise or other situation where you have to make any compromise. Out in the burbs in a nondescript warehouse that is away from floods and fire is the way to go. And do keep the maintenance up to the system + fuel. Too simple = no excuses
rickripe Aug 19, 2009 10:00 AM	Yes, I agree a complete boatload of bollocks. Ravi's basically blaming everyone else and not taking any responsibility for the outages. What's particularly galling is the reference to power plants when the outages this year have been due to sub station issues, losses of a single phase inside the building and then yesterday the loss of a "static transfer switch". The last two are wholly inside and under Primus's control. NO EXCUSES. We've been running servers in other data centres for 8 years in AUS and 6 years in the US - we've had one minor outage in AU 5 years ago and none in the US. It comes down to pretty poor risk management on their part - no, strike that - it comes down to pretty poor management. When will Primus learn that they have to invest not only in the infrastructure but also in the staff and the management. We're pulling our servers out of there as this whole debacle has cost us money (in rebates to our clients) and it's far cheaper to go to another data centre - even with higher floor space/data rates. I simply can't believe that these guys think they are running even close to a Tier 3 data centre (http://www.adc.com/us/en/Library/Literature/102264AE.pdf).
sunpazed Aug 19, 2009 1:35 PM	Doing power upgrades is one thing. Managing your service and systems responsibly is another. Primus has had another (although small) outage that affected the Co-Lo room last night. We lost power to some of our servers completely for a few minutes. Mr Bhatia, this further highlights your failure as a CEO, and your organisations incompetence at delivering a so called 'bullet-proof' solution. From: Melbourne NOC [mailto:melb_noc@primustel.com.au] Sent: Tuesday, 18 August 2009 7:00 PM To: Melbourne NOC Subject: Dear Customers, We have just had an issue with 1 of our static transfer switches in the Co-Lo room which means that half the power in the room would have been lost for the period of the outage. The static transfer switch is back online. Contractors are currently on the way to investigate. Further updates to come. Regards, Primus DC Melbourne NOC
DJ Aug 19, 2009 4:37 PM	Ravi, it's time for you to go. You can't keep crapping on about power when other datacentres in the same area have experienced zero outages whilst Primus continues to invoke downtime on customers who are paying top dollar for an enterprise service. Clearly you are not in control of the DC environment and your management team don't "have things in hand". Get rid of Ravi and the middle management who are supposed to be managing power. Replace them with people who actually know what they are doing and value client business. The whole bulletproof thing was a dumb statement and something you are a long way from achieving.
anonymous Aug 20, 2009 3:21 PM	"High voltage and high power is a subject [this IT community] does not understand," Bhatia said." Perhaps we understand the need for it quite a lot better than some of the relevant corporate management executives.