Princeton IT ServicesResources / Blogs
Hadoop on Android
Could it be possible to run Hadoop on a cluster of mobile phones? This can be considered as the future scope of Hadoop. Recently a team at Google has pushed Hadoop to the limits.
Android and Hadoop are two technologies to talk about. These individual technologies solve many real time problems.
“How about combining them ?” is a question for people interested in future development of technology.
Could it be possible to run Hadoop on a cluster of mobile phones? This can be considered as the future scope of Hadoop. Recently a team at Google has pushed Hadoop to the limits by creating a cluster whose size is on the order of 1,00,000 nodes, running on the recently released Nexus One mobile phone hardware, powered by Android. By pushing computation out to these devices, the Nexus One team was able to solve the difficult rendering and scaling problems in situ.
The features that seem to make this idea work are :
1. Hadoop requires local disk memory, network speed and RAM. These can be built in mobile phones.
2. Hadoop is completely written in java. Why should not we think about Hadoop in java micro edition?
The challenges :
1. The battery life for phones can be a big concern.
Four role-based workspaces in one app