Summary: The researcher have designed a system to meet their storage needs. Google File System has been replicated in this research which make it possible to run the file system on cheaper commodity software. The files system designed is supporting distributed applications. The system takes in to account the application workload and technological environment. The designed file system can be termed as a step ahead from the previous file system conventions.
Strengths: Google File System is serve that acts as a master and can be accessed by many users and many chunkservers at a single time. Only one machine is needed to serve both. In the researchers use a 64 MB chunk. It is much larger than the traditional chunk size. The large chunk reduces workload because the user then does not need to communicate with the master all the times. It especially helps when program applications use larger files for their read and write operations. The design introduced in the research incorporates a technique that stores the metadata in the memory which makes the master operation quicker.
Weaknesses: The larger chink size also has some disadvantages. In case of smaller files, would get overloaded because of the larger chunk. The other weakness in the research is that there are many assumptions made to design the system. These assumptions doubt the durability of the design system. The authors have also pointed out that they “experienced a variety of issues, some operational and some technical.” In designing and incorporating Google File System.
Questions: How would the design respond to a de-centralized master instead of a centralized one?
How does the designed system responds to inter-server requests?