Arun Rajput Computer Science Notes

Following topics will be covered: BigData stack installation Business case Development 1. Installation: Best way to go for this step is to have Ubuntu. I found this OS to be very friendly. you can go ahead with 14 Long term license for installation. Use Ubuntu site to download OR Note: you can try before actually installing. you can also make bootable USBfrom windows and try this USB -OS first before going for actual installation.   1.1. Which file system ? : HDFS -  1.2 Which data management Layer ? - YARN 2. Business Use case: It is very important to understand that bid data world or hadoop world is for problems which have 5 V's . volume, variety, velocity, value and veracity. More technically speaking if you are not going to have more than 5 datanodes no point using hdfs or less than millions/billions of of records thinks twice to use Hbase. Also check the data generation activity e.g. is it machine or human generated. 3. Development: Eclipse

Ok I understand now's it data on top of REST..... :) I really don't understand the pace at which IT is changing.. It was just few months back we started talking about REST API. Some of the geeks stopped talking about SOAP and started conversation on REST. The potential key point was introduction of mobile and some governance issue around SOA. Many of the sessions and seminars started talking about REST principals and architectural style. Giant companies also started aligning themselves around this concept. Many API tools emerged. Than suddenly people started talking about writing query itself in URL What ???? ...... U mean writing select query in URL itself . Answer is yes. Obviously with little syntax change. So what about writing joins, unions etc.. They all possible. OData  specification supports all this. http://www.odata.org/ Don't miss the bus ... This is important from today's multi-channel application perspective.

coming soon....

VM stands for virtual machines: Name it self tells you that computer hardware has been written as software and provided to you. It has its Operating system and all.... Containers: Linux kernel has some specific system calls like cgroup etc  which can be leveraged o create separate processes called containers. It means same OS and top of it many small processes or lighter VM's running called as containers. Future is moving towards containers. examples are dockers LxC rkt etc since we can have thousands of containers developed and assembled to called as project hence we need tool to orchestrate (e.g. resource management)  this. One of such tools is kebernetes.