BT Internet Outage Was Our Fault, Says Equinix (theregister.co.uk)
Kat Hall, reporting for The Register: Telecity's owner, Equinix, has 'fessed up to a "brief outage" which subsequently knocked 10 per cent of BT internet users offline this morning as well as a number of other providers. A spokesman from the group, which slurped up Telecity for 2.3bn euro last year, confirmed that the outage occurred at its LD8 site in the Docklands. The company has nine London sites which service more than 600 businesses.The outage occurred due to power failure, which lasted for around 75 minutes. ( Update: Some readers note that the outage lasted for as long as three hours. ) BT wasn't the only ISP that suffered an outage earlier this morning. All services have been restored, according to Ars Technica. Update: 07/20 14:57 GMT: It was apparently a faulty UPS that caused the outage.
Well, for one thing it shows how easy it is to pull the plug on a lot of people at once. A single service like this shouldn't have this kind of power. It illustrates the necessity of having alternate hookups that can *route around the damage*
“He’s not deformed, he’s just drunk!”
This is why I have redundancy and backups. This morning I had connectivity issues on my 100Mbps BT line, so I switched to my 200Mbps Virgin Media line and all was well.
It does surprise me how many people that depend largely on the Internet for work don't get a second Internet connection (and do so on separate infrastructure) and then complain when they have down time and how some how they couldn't get 100% uptime all year.
Change is certain; progress is not obligatory.
Reminds me of a company I know that built a brand new data center, put in an over-sized UPS system, over-sized electric generator, state of the art power monitoring/transfer system, and tested the generator bi-annually. Only there were three problems when the area finally suffered a blackout:
1.) They never tested fail-over to the UPS, they had only tested starting up the generator and then manually switched off mainline power once the generator was fully operational to see if it worked.
2.) The UPS installer never bothered to connect the batteries to the power control unit so when power did fail everything immediately lost power (these were racks of large batteries wired together, not the plug-and-play consumer stuff.) When the generator kicked in, everything tried to turn back on at the same time and tripped the breaker. LOL.
3.) The air conditioner was not connected to the fancy power transfer system and after a little over an hour, servers started throttling and eventually shut down from the heat. This all happened in early August on one of the hottest days of that year.
-==- Buy a Mac and leave me alone!