CloneCloud (CC, Distributed Processing)

Question 1

What is the Background (CloneCloud)?

Answer 1

• CloneCloud automatically transforms existing Android applications to leverage cloud resources.
• It uses static and dynamic code analysis.
• At runtime, CloneCloud migrates individual threads from the mobile device to a clone in the cloud.
• It was designed to address the increasing demand for powerful computations on mobile devices with limited energy and processing power, taking advantage of increased connectivity to cloud resources.

Question 2

What is the problem (CloneCloud)?

Answer 2

• Mobile devices have constraints such as slow processing power, short battery life, and storage limitations.
• Ensuring optimal split between client-side and cloud (server-side) computation is a key challenge.
• Existing client/server models often push most computation to the remote server, which may not always be optimal.
• Performance suffers in dynamic conditions like changing network speeds

Question 3

What was the solution to the problem (CloneCloud)?

Answer 3

• CloneCloud provides a flexible architecture for seamless use of ambient computation to augment mobile device applications, making them fast and energy-efficient.
• It uses application-level virtual machines (VMs) to enable fine-grained partitioning of mobile applications for offloading to the cloud.
• CloneCloud employs a combination of static analysis and dynamic profiling to partition applications automatically at a fine granularity.
• It migrates individual threads from the mobile device to a clone in the cloud.
• It uses a migrator thread, a node manager, and a partition database for migration.

Question 4

What is an application (CloneCloud)?

Answer 4

NONE, but assumed to be android mobile devices.

Question 5

Strengeths and Weaknesses (Clonecloud)?

Answer 5

Advantages:
- Mobile devices can achieve up to a 20x execution speed-up and a 20-fold decrease in energy spent for applications with heavy workloads.

• It allows mobile devices to offload only computationally intensive parts of an application to the cloud while remaining sensitive on the mobile side.
• It operates at thread granularity.
• It can use specialized hardware/system facilities on device clones.
• It automatically calculates the cost of thread migration.
• It adapts to different network and workload variations.

Disadvantages:
- Adaptability happens off-line, before execution.

• It doesn’t dynamically adjust to real-time variations.