Column-Family Database Features

Column-Family Database Features

1. Cassandra

   • Basic unit
     • Column
       • Name-value as key
       • Timestamp value
         • Expires data
         • Resolves write conflicts
       • Super column
         • Multiple super columns
           • Super column family
           • Not recommended for
             • Rare column access
         • Content
           • Map of columns
   • Keyspace
     • Receives and stores column families
     • Configuration
       • Number of data replicas required

2. Consistency

   • Successful write operation
     • Commit log
     • Memtable
       • In-memory structure
   • SSTable
     • Data updates
   • Compatation process
     • Manages unused structures
   • Low consistency
     • Suitable for
       • Use of non-recent data
       • High read or write performance
       • Probable data loss
   • QUORUM consistency
     • Provides the newest column
     • Successful
       • Most nodes write
   • All consistency
   • Hinted handoff
     • Online nodes in charge of
       • Own responsabilities
       • Offline nodes tasks

3. Transactions

   • Single row
     • Atomic transaction
   • External libraries
     • ZooKeeper
     • Cages

4. Availability

   • More availability
     • Lower consistency level
   • No master node
   • Quorums

5. Query features

   • Rows sorted by column names
   • Basic queries
     • Steps
       • Specify the keyspace
       • Define the column family
       • Query for data
     • Queries
       • SET
       • DEL
       • GET
         • Whole column family
         • One specific column
           • Efficiency

6. Scaling

   • Large number of nodes