i) HiCoMo (High Commit Mobile Transaction Model)
Overview:
HiCoMo stands for High Commit Mobile Transaction Model. It is designed to ensure higher transaction commit rates in mobile environments, where frequent disconnections and limited resources pose challenges to traditional database transaction models.
Key Features:
-
Agent-based approach: The transaction is split into two parts — Mobile Part Transaction (MPT) and Fixed Part Transaction (FPT).
-
Execution Location:
-
MPT is executed on the mobile host (MH).
-
FPT is executed on the fixed host (server or base station).
-
-
Checkpointing and Recovery: Uses checkpointing to handle mobile disconnections.
-
Commit Strategy: Only commits when both MPT and FPT succeed, ensuring ACID properties.
Advantages:
-
Increased transaction commit rates.
-
Reduces the effects of disconnections by allowing local processing.
-
Efficient recovery using checkpoints.
Disadvantages:
-
Complex management of agent-based execution.
-
Requires synchronization between mobile and fixed parts.
Use Case:
-
Banking apps or stock trading apps where partial commit is not acceptable and consistency is crucial.
ii) Kangaroo Mobile Transaction Model
Overview:
The Kangaroo model is inspired by the mobility of kangaroos — "hopping" from one place to another. It addresses the issue of mobile users changing their point of attachment (e.g., moving between base stations) during transaction execution.
Key Features:
-
Hopping Transaction: As the mobile host moves, the transaction also migrates along with the user, hopping from one mobile support station to another.
-
Mobile Agents: Each transaction is encapsulated within a mobile agent that can move across different servers.
-
Location Transparency: The model provides location-independent transaction processing.
-
Recovery: Maintains state across hops, ensuring that even if the mobile device disconnects, the transaction can continue.
Advantages:
-
Supports high mobility without disrupting transactions.
-
Provides seamless migration of transactions.
-
Ensures better fault tolerance in mobile environments.
Disadvantages:
-
Increased overhead due to agent migration and state maintenance.
-
Requires infrastructure support for agent-based systems.
Use Case:
-
Logistics or delivery systems where agents (like delivery personnel) are constantly moving but require consistent access to backend systems.
0 Comments