Google goes dark for 2 minutes, kills 40% of world's net traffic

[Panthro]

You can all relax now. The near-unprecedented outage that seemingly affected all of Google's services for a brief time on Friday is over.

The event began at approximately 4:37pm Pacific Time and lasted between one and five minutes, according to the Google Apps Dashboard. All of the Google Apps services reported being back online by 4:48pm.

[theregister.co.uk...]

I noticed the down time yesterday. Interested to see if we get any information about the cause. Swamp gas maybe?

[graeme_p]

About half of that 40% dip would have been Youtube, I wonder what the rest were?

Search is busy but not bandwidth hungry. Google Drive and apps are much less widely used, so its not obvious to me - no doubt WW will, as usual, produce someone who knows the answers :)

[brotherhood of LAN]

The Reg article references GoSquared who seem to be a web analytics company, for me it'd seem the 40% is an estimate based on referrals they know about. So it's not necessarily 'all web traffic' but visitors based on referrers.

[lucy24]

Whew. The alternative would be to postulate that 40% of all www users worldwide are incapable of going to any site, anywhere, ever, by any means other than opening google dot yourcountry and searching for the desired site name. (No bookmarks? No browser history? No links from the page they happened to be on when google dot anything fell off the earth?)

[Mr Bo Jangles]

Whew. The alternative would be to postulate that 40% of all www users worldwide are incapable of going to any site, anywhere, ever, by any means other than opening google dot yourcountry and searching for the desired site name. (No bookmarks? No browser history? No links from the page they happened to be on when google dot anything fell off the earth?)

Love it!
(and a degree of accuracy in it I would think)

[graeme_p]

No, because its all Google services, not just search.

It was also only for two minutes, so some people would have just tried again - notice the spike after the dip?

I was wrong about YouTube because, look at the original article:

[engineering.gosquared.com ]

The drop was in page views per minute, not data volume.

I wonder how meaningful a measure it is given how many busy sites have infinite scroll and other AJAX elements.

[creeking]

installing those PRISM logging servers took a few minutes.

[jmccormac]

nstalling those PRISM logging servers took a few minutes.

NSA is screwed if it is relying on Google to detect dodgy links. :)

Regards...jmcc

[engine]

hehehe, creeking, nice one.

It's quite possible that some of that 40% was from Android OS mobile users that probably wouldn't realise it'd gone offline, and just thought it was their connection.

[seoskunk]

Great News 60% of the net don't use Google........

[lucy24]

Great News

In a random two-minute period, only 40% of the net was using a google service.