Looks like Yahoo is clean to improve their services and make it the best Search Engine in the world, after word got out that Yahoo will be added Yahoo Real Time Results features.
Latest generation of Apache Hadoop Mapreduce 2, a software implementation of processing petabytes of data, will be out this year, said a Yahoo executive told News.com IT.
Todd Papaioannou, Vice President of Architecture Yahoo to Computerworld this week saying that the current iteration Hadoop Mapreduce 2 apply less able to manage resources effectively thousands of servers in a cluster.
So developers are working to improve the use, scheduling and resource management.
For example, a new architecture that will include global Resource Manager that will track the availability of the server scheduling invariant when Application Master operate in clusters and track program semantics for a particular job, the literature of Yahoo developer Arun Murthy in one blog post.
According to Papaioannou, Yahoo currently contributes about 70% of the looping code Hadoop and Hadoop Distributed File System (HDFS).
Earlier this year, Yahoo Hadoop Mapreduce 2 stop distributing its own and began working closely with the community because it can draw the Apache Hadoop open source community to help the developments, said Papaioannou.
Together with Apache, Hadoop MapReduce using iteration, a programming technique that comes from Google, to build parallel programs by using Hadoop, MapReduce allows data processing in parallel.
"The next generation HDFS will be more resilient and reliable. We hope to put it together in a launch some time to come. It is a test of cooperation with the entire development community," said Papaioannou.
Yahoo just launched a new project called H Catalog which is a metadata table management scheme Hadoop.
Latest generation of Apache Hadoop Mapreduce 2, a software implementation of processing petabytes of data, will be out this year, said a Yahoo executive told News.com IT.
Todd Papaioannou, Vice President of Architecture Yahoo to Computerworld this week saying that the current iteration Hadoop Mapreduce 2 apply less able to manage resources effectively thousands of servers in a cluster.
So developers are working to improve the use, scheduling and resource management.
For example, a new architecture that will include global Resource Manager that will track the availability of the server scheduling invariant when Application Master operate in clusters and track program semantics for a particular job, the literature of Yahoo developer Arun Murthy in one blog post.
According to Papaioannou, Yahoo currently contributes about 70% of the looping code Hadoop and Hadoop Distributed File System (HDFS).
Earlier this year, Yahoo Hadoop Mapreduce 2 stop distributing its own and began working closely with the community because it can draw the Apache Hadoop open source community to help the developments, said Papaioannou.
Together with Apache, Hadoop MapReduce using iteration, a programming technique that comes from Google, to build parallel programs by using Hadoop, MapReduce allows data processing in parallel.
"The next generation HDFS will be more resilient and reliable. We hope to put it together in a launch some time to come. It is a test of cooperation with the entire development community," said Papaioannou.
Yahoo just launched a new project called H Catalog which is a metadata table management scheme Hadoop.
0 comments:
Post a Comment
Just Keep on Topic, don't be Spamy!!