bigtable

Maybe they should be looking at bigtable or a bespoke system.

Twitter At Scale: Will It Work? Nik Cubrilovic 2005

BigTable Replication (New Since OSDI'06) • Configured on a per-table basis • Typically used to replicate data to multiple bigtable clusters in different data centers • Eventual consistency model: writes to table in one cluster eventually appear in all configured replicas

Recently Uploaded Slideshows longhao 2010

BigTable Replication (New Since OSDI'06) • Configured on a per-table basis • Typically used to replicate data to multiple bigtable clusters in different data centers • Eventual consistency model: writes to table in one cluster eventually appear in all configured replicas

Recently Uploaded Slideshows longhao 2010

The most interesting question is "how to do bigtable from java", and the answer is interesting too.

Planet Apache 2009

Follows the bigtable model, so it's more complicated than it needs to be. (300+kloc vs 50 for Cassandra; many more components).

Spyced 2009

Specifically, we will cover bigtable, dynamo, and cassandra, and how to port a typical relational schema to cassandra's ColumnFamily model.

Spyced 2009

While Google, in the recent stage, binds its bigtable databases with its own web application.

螺旋进化 2009

As it just came to my mind the performance of TokuDB at some extent resembles one of google bigtable (or SStables design) - they also have very fast inserts avoiding sequential writes, compression and rather fast scans.

random($foo) 2009

Putting together the different components google has built up (bigtable, mapreduce, gfs, pagerank+all the other variables added to the search algorithm, ad relevance) would at least be a start – then augment it with MS’s own innovations, things that google is way behind on like maybe a rich client interface or maybe giving voice commands to the engine that it can understand.

Hey Microsoft, How ’bout We Do That First Deal You Offered? Michael Arrington 2005