Does Asia have to rethink its Internet architecture after the Boxing Day earthquake‾
'I was starting to get calls from 9 pm. When my guy called up and said the cable's broken, I thought, 'my holiday is gone'.'
That's how the news of Asia's worst cable disaster broke for Andrew Kwok, vice president international business for Hong Kong telco HGC.
He wasn't the only one to spoil his holiday. Around the region, hundreds of engineers worked day and night to reconnect lost bandwidth after a once-in-a-century earthquake shredded Asia's trans-Pacific connectivity.
The impact on broadband-happy east Asia was catastrophic. Hong Kong and Singapore lost 80%-90% of connectivity, Taiwan nearly 100%.
For 48 hours Internet users had virtually no offshore connectivity. By the weekend the Web was back up, but extremely slow. After a week or so, with new paths running to the US via China, Europe or Australia, it began to approach normal.
Questions since have rumbled through the industry. Why were so many cables placed in the same location‾ How can we diversify our connectivity‾ Will it happen again‾ Can subsea capacity avoid quake-prone zones‾
The story unfolded like this.
Cables break slowly. At first some water gets in, then some fibers are torn, then some more and then soon enough the cable is completely out.
After the earthquake off Taiwan's southern tip at 8.30 pm on Boxing Day, the shifting seabed began to stretch some of the dozens of cable segment that run through it.
By lunch the following day, eight international cables had been severed in 16 places. As every Asian Internet user now knows, the prime role of those cables was to connect across the Pacific to the US, the largest single source of Internet content.
The result was Asia's biggest ever loss of Internet capacity. Users were unable to access major US sites such as Google, Yahoo, YouTube and CNN.
Echoing a universally-held view, VSNL vice president of global transmission services Byron Clutterbuck said it was fortunate that the event occurred in one of the quietest periods of the year.
But some did better than others out of the break.
Like Asia Netcom. The regional carrier lost two cable segments on its EAC system, which runs between Singapore and Japan, the first going down about 2 am on the 27th and the other at 11 am.
But the Hong Kong-based operator was the least affected. ANC president and CTO Wilfred Kwan attributed it to the multiple redundancy paths created by EAC's rings as well as recent completion of a 50% upgrade.
It was also a matter of luck, too, that the system was hit in two segments but could still operate.
A spokesman admitted ANC had been 'lucky' with the upgrade, too, but noted it underlined ANC's advantage in managing its own infrastructure.