Re:Cassandra vs MongoDB For Time Series Data
by relistan on Thursday August 22, 2013 @05:29AM
(#6256102)
Disk usage patterns are the problem with most relational stores (and were for Mongo). When you do a query on time series data, you want to look up a start, read some data, and stop when you hit the end point. Getting it laid out largely sequentially on disk is a big win there.
[ Parent
| on HN ]