This question is about Stackable Manipulators. The author talks about using manipulators in section 4.6 and mentions that manipulators manipulate any calls from and replies to the original nfsserv object. The idea behind manipulators is not clear from the paper. Could you comment on how are they useful in bug fixing.
It is said that top level directory is served by auto mounter process and each sub-directory is handled by separate loop-back server... When accessing the directory inside the one handled by loop-back server, a new mount point is created or the old one is used?
Each loop-back server is a separate process, can you compare the performance due to asynchronous IO to the increased number of context switches?
This question is about Stackable Manipulators. The author talks about using manipulators in section 4.6 and mentions that manipulators manipulate any calls from and replies to the original nfsserv object. The idea behind manipulators is not clear from the paper. Could you comment on how are they useful in bug fixing.
ReplyDeleteIs there any future work with respect to toolkit which uses DAFS type of file system instead of NFS due to the advantages DAFS has over NFS?
ReplyDeleteIt is said that top level directory is served by auto mounter process and each sub-directory is handled by separate loop-back server...
ReplyDeleteWhen accessing the directory inside the one handled by loop-back server, a new mount point is created or the old one is used?
Each loop-back server is a separate process, can you compare the performance due to asynchronous IO to the increased number of context switches?
In the very first Line its said that the toolkit is developed in C++. So IS this toolkit platform dependent?
ReplyDeleteIf yes is there any recent research made on it to make it platform independent.?