Site Loader
Rock Street, San Francisco

Bogdan Nicolae, Diana Moise, Luc Bouge  have proposed the efficiency of the hadoop
framework is having a direct interaction with storage layer so the work
demonstrates and usage of  the default
hadoop distributed file system with the another layer, built along with design
principles so they introduced the new file system that is blob seer distributed
file system it can give the application of the huge data, optimized toward
efficient, fine grain access and distributed data access under heavy
concurrency this blob seer distributed file system layer can be sustained in
the throughput of the hadoop and exhibit the concurrent access to the shared
files and this blob seer distributed file system can have the feature of
the  concurrent appends, concurrent
writes at random offsets and versioning.

Viet-Trung Tran, Gabriel Antoniu, Bogdan Nicolae, Luc Bouge,
Osamu Tatebe  have proposed the problem
to address large amount of data volumes at very large scale and special focus
on the applications which manipulate huge data but logically shared data at a
fine grain access under heavy concurrency it provides transparent access
through the globally shared name will greatly simplifies data
management in the applications to transfer the data on various sites and
proposed the architecture based on the grid blob seer distributed file system
this will execute the scalable file access performance in various scenario
where huge files exhibit massive and concurrent access.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

T. Lashmi siva Rama Krishna, T. Ragunathan , Sudheer Kumar
Battula have proposed the algorithm for the improvement of read algorithm based
on speculation and proposed a new read algorithm  for distributed file system and developed
based on the speculative semantics, the above algorithm can read the recent
data which is concurrently updated by the concurrent write, This algorithm is
mainly used in the application of stock exchange system, where availability of
recent data is important so that a user can take the decisions very easily.

Diana Moise, Gabriel Antoniu, Luc Bouge have proposed the
map/reduce framework programming model into the internet services community and
it is vice-versa to increasing application, it mainly focus on the append
operation in this paper append operation is having two functionalities

In introducing append operation for the
applications to be continuously monitored of the log system of read/append to
operation in blob seer distributed file system and also for the deliver of the
multiple clients concurrently append data to same file.

operation is supported by the file system so modified the map/reduce framework
, after modified the hadoop framework reducers can append their data to a
single file instead of writing to a separate file so it was done in the
original version of the hadoop framework and blobseer distributed file system as
uses the storage layer and this optimized hadoop framework can further be used
in the case of the map/reduce framework these applications can be executed in
the pipeline.

Bogdan Nicolae, Gabriel Antoniu, Luc Bouge have proposed the
work for the efficient usage of the versioning for a large object under heavy
access concurrency  and allowing the multiple
clients to execute the operations in concurrently read, write, append data to BLOB
, This algorithm will guarantees 
atomicity for the good data access performance and favour for the
scalability under heavy concurrency and rely on the original metadata based on
the  scheme of distributed segment tree
built on the top of the distributed hash table(DHT), they have extended their
approach on the volatility and failures.

Post Author: admin


I'm Velma!

Would you like to get a custom essay? How about receiving a customized one?

Check it out