Google Storage Lessons: Factoids with a Slice of Baloney
July 15, 2016
I read “Lessons To Learn From How Google Stores Its Data.” I noted a couple of interesting factoids (which I assume are spot on). The source is an “independent consultant and entrepreneur based out of Bangalore, India.”
The factoids:
- Google could be holding as much as 15 exabytes on their servers. That’s 15 million terrabytes [sic] of data which would be the equivalent of 30 million personal computers.
- “A typical database contains tables that perform specific tasks.”
- According to a paper published on the Google File System (GFS), the company duplicates each data indexed as many as three times. What this means is that if there are 20 petabytes of data indexed each day, Google will need to store as much as 60 petabytes of data.
As you digest these factoids, keep in mind the spelling issues, the obvious, and the reference to a decade old Google article.
Now the baloney. Google keeps it code in one big thing. Google scatters other data hither and yon. Google struggles to retrieve specific items from its helter skelter set up when asked to provide something to a person with a legitimate request.
In short, Google is like other large companies wrestling with new, old, and changed data. The difference is that Google has the money and almost enough staff to deal with the bumps in the information superhighway.
The Google sells online ads; it does not lead the world in each and every technology, including data management. Bummer, right?
Stephen E Arnold, July 15, 2016
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
