Which design can organizations use to achieve better results and reduce risks in
running Hadoop on a network?
*Server-to-server and server-to-storage traffic outweigh client-to-server traffic
Multitier network design with degraded bandwidth
One replica of a data node is on one rack, and two copies of two different data
odes are on different racks
Client-to-server and server-to-storage traffic outweigh server-to-server traffic
What is a characteristic of an ideal network?
Latency of intermediate switches at the distribution tier is lower than that of a
TOR switch
Total server communication bandwidth is 2.5 times the inter-rack network bandwidth
*Copies of data replication can be placed into different, unique racks
Data nodes are configured with 20 Gbps connections
Of the four server roles in a Hadoop cluster, which one tracks the location of a
file�s data blocks throughout the cluster?
Client
Job Tracker
Data Node
*Name Node
What is a benefit of the QFabric architecture performance scalability?
Intrarack server communication bandwidth is 20 Gbps
Complexity grows as the size of the cluster grows
*Hadoop can run on most data center networks
Consistently low any-to-any latency
Which technological innovation has contributed to the dramatic increase in data
creation and data gathering?
Structured Query Language (SQL) analysis tools
Information and Communication Technology (ICT)
Hadoop clusters
*Increased use of sensors
What are the major attributes of big data?
Terabytes and petabytes
Social networking and mobile devices
Click streams, data streams, and structured data
*Variety, volume, and velocity
Which healthcare use of big data can help predict and reduce future pandemics?
Medical error prevention
Improved diagnosis and treatment
Predictive analysis
*Disease trend analysis
Which feature of the QFabric system delivers a common window for managing all
components as a single device?
QFabric Node Edge Solution
*QFabric Director
QFabric Interconnect
QFabric Node
What are the quantities of network devices for a midsize Hadoop configuration?
100 QFabric Node devices, 4 QFabric Interconnect devices, and 2 QFabric Director
devices
200 QFabric Node devices, 6 QFabric Interconnect devices, and 4 QFabric Director
devices
*10 QFabric Node devices, 2 QFabric Interconnect devices, and 2 QFabric Director
devices
50 QFabric Node devices, 3 QFabric Interconnect devices, and 2 QFabric Director
devices
How does the QFabric system improve inter-rack bandwidth?
Hadoop clusters can collect data from an FC-based SAN through a converged network
Nodes are part of one switch
*One network hop exists between any-to-any servers
HDFS policy places three copies of data into three unique racks without affecting
the write performance
