Tag Archives: data storage

Instagram Moved 20 Billion Photos To Facebook – Did Any Fall Through The Cracks?

Unbeknownst to the hundreds of millions of Instagram users, the infrastructure that holds all of their photographic creations has quietly been dismantled, relocated and rebuilt. Impressive stuff, when you consider the quantities of data involved – the service now stores in excess of 20 billion digital photos.

Since 2010 Instagram had been using Amazon’s cloud computing service, but the number of virtual machines required to run Instagram on Amazon was getting in “the thousands”. Historically, when Facebook had acquired other, smaller properties, the process involved shutting down the service in order to incorporate it into the world of Facebook. However, according to Facebook engineer George Cabrera, in the case of Instagram “the service couldn’t take any disruption”.

So in what Facebook called ‘Instagration’, the team in essence had to carry out multiple organ transplants while both patients were still conscious. Or as Mike Kreiger, founder of Instagram, explained:

“The users are still in the same car they were in at the beginning of the journey, but we’ve swapped out every single part without them noticing.”

The project was complicated, but the team behind the data transfer have successfully transitioned Instagram to now run from its own dedicated machines inside one of Facebook’s facilities. If you’re particularly interested in what the team had to do to make sure the whole thing went off without a hitch, then check out this article by WIRED’s Cade Metz:

How Facebook Moved 20 Billion Instagram Photos Without You Noticing

While everything seemed to go smoothly, whenever your content or data is held by a network or third-party that may encounter scaling issues or could indeed be absorbed into another company’s infrastructure, there will also be a concern that data could be lost along the way. While the ‘Instagration’ is a great example of data migration done well, it’s also a reminder that there’s always the potential that the actions of others could cost you what’s yours.

We believe that the individual should be the one who owns and controls their personal data, and that is why we built the SocialSafe application to allow you to download your Instagrams, Facebook Messages, Tweets, Followers and so on, all to your very own personal data store. There’s no reason to think that the networks themselves won’t look after the originals properly, but accidents can and do happen, so why take the risk?

Start building you own personal data library today, by downloading SocialSafe for free, and backing up your memories from the social networks you use.

What Does A Facebook ‘Like’ Really Say About You? Why Fragmented Data Is Wrong

The significance associated with ‘Liking’ something on Facebook can vary greatly depending on who sees what you Like, and what they already know about you.

First of all, you may ‘Like’ a page purely for the purposes of staying up to speed on a brand or public individual’s activity, even if they were a competitor or your ideological polar opposite. Think about a Greenpeace activist ‘Liking’ Shell, Chevron or BP, just so that they know what they are up against. They don’t necessarily ‘like’ (with a small ‘l’) what the company is doing, but Liking (big ‘L’) on Facebook gives pseudo-endorsement thanks to the arbitrary name given to consciously opting in to see updates from that company, politician, brand etc.

Then there’s the issue of whether or not you actually ever clicked Like in the first place…

Just over a year ago an article on ReadWrite highlighted many instances in which people were seeing promoted posts, suggested pages and news stories indicating that their friends ‘Like’ certain products, brands of pubic figures. The author followed this up with some of his friends who – according to Facebook – had given their tacit endorsement to some very unlikely pages, and they were left scratching their heads. Most of them had no recollection or staunchly denied having ever clicked Like in the first place.

When asked about this, Facebook fell back on human error, with a spokesman saying that people can “Like” things by accident, perhaps by inadvertently pressing a button on the mobile app. The author of the article, remained skeptical however, saying that the frequency with which so many of his friends could make the same ‘mistake’ was just too suspicious, although it could be the work of third-parties selling Likes to pages.

Whatever the reasoning behind the Likes, this story demonstrates that if someone is supposedly showing their support or endorsing particular brands on Facebook, it might not necessarily be an accurate picture of them. Singular aspects of an individual’s different online profiles – when viewed in isolation – may paint them in a certain light, but it’s only when all of the component parts of their online self are brought together that the picture become clear and complete.

These days people use a variety of services to engage in lots of activities, creating data about themselves that is fragmented and stored by a multitude of different networks and companies. At SocialSafe we believe that that the individual should be the biggest single owner of their own data, which is why – starting with social networks – we are enabling our users to created their own aggregated personal data store.

Some of the most successful sports teams, bands and other groups have been described as ‘greater than the sum of their parts’.  The power of several otherwise unremarkable or average members combining into something that performs at a level far higher than simply adding up the achievements or skills of all the individuals when operating separately. Surely the same could be true if you brought together all of your fragmented personal data?