This document discusses using MongoDB for social analytics at BuddyMedia. It describes how BuddyMedia started with Cassandra but found it too complex, so they chose MongoDB which was simpler to develop with. Raw events are collected and processed to update metrics in MongoDB aggregated at different resolutions (minute, hour, day). Examples of queries on the metrics data are provided to retrieve metrics for specific clients or time periods.
This document discusses using MongoDB for social analytics at BuddyMedia. It describes how BuddyMedia started with Cassandra but found it too complex, so they chose MongoDB which was simpler to develop with. Raw events are collected and processed to update metrics in MongoDB aggregated at different resolutions (minute, hour, day). Examples of queries on the metrics data are provided to retrieve metrics for specific clients or time periods.
Double page spread before creation powerpoint uploadjadeparrettharris
THIS IS TO SHOW THE PLANNING WHICH I CARRIED OUT BEFORE MAKING MY DOUBLE PAGE SPREAD, ANALYSING PREVIOUS DOUBLE PAGE SPREADS AND THINKING ABOUT WHAT I MAY DO.
This summary provides an overview of Buddy Media's experience using MongoDB for three stages of a project:
Stage 1 was a non-critical logging application where they learned MongoDB is not like MySQL and to use subdocuments instead of rows/tables.
Stage 2 involved critical user data with spikes where they learned to use modifier operators carefully and implement indexing and replica sets.
Stage 3 was for real-time analytics of user events, requiring flexibility and high write volumes. They store aggregated metrics instead of individual events and use upserts and $inc to update in bulk, providing faster performance than SQL for their needs.
Double page spread before creation powerpoint uploadjadeparrettharris
THIS IS TO SHOW THE PLANNING WHICH I CARRIED OUT BEFORE MAKING MY DOUBLE PAGE SPREAD, ANALYSING PREVIOUS DOUBLE PAGE SPREADS AND THINKING ABOUT WHAT I MAY DO.
This summary provides an overview of Buddy Media's experience using MongoDB for three stages of a project:
Stage 1 was a non-critical logging application where they learned MongoDB is not like MySQL and to use subdocuments instead of rows/tables.
Stage 2 involved critical user data with spikes where they learned to use modifier operators carefully and implement indexing and replica sets.
Stage 3 was for real-time analytics of user events, requiring flexibility and high write volumes. They store aggregated metrics instead of individual events and use upserts and $inc to update in bulk, providing faster performance than SQL for their needs.