- Distributed Database
- Distributed File System
- we use the structure master slave which we have a main master and other slaves
- in the master the users create users and there is a backup database to insert all the users that should be inserted in the slaves
- we have implemented backupdatabase program to check periodically all the slaves and insert the user in it
- we have to ensure that all databases are consistent as the user try to enter can log in from any slave
- if a user tries to log in to database then he selects from slaves and if all slaves are dead then we select from master
- a lot of other things
- we make a master tracker and data node keeper
- master tracker handle the connections between the client and data nodes to upload and download
- master tracker has database so can store where client files in which data nodes
- we make replications to three data nodes for each file and we make sure they are alive
- all and more than in the document requirement we have handle it