Hadoop Admin

We are looking for a Hadoop Admin for a contract that can last for years.


Position : Hadoop Admin


Location : Phoenix, AZ


Duration : 6 Months plus

This will be purely Hadoop administration NOT development with exposure to the Map-R distribution system and about 60% proficiency in Linux administration. Here is a little bit more about the position:

Top Three Skills:


1. Hadoop Administration (NOT hadoop development) and Map-R (map-reduce)


2. 50-60% Proficiency in Linux Administration


Nice to Have: automation skills

1. Responsible for implementation and support of the Enterprise Hadoop environment.
2. Involves designing, capacity planning, cluster set up, monitoring, structure planning, scaling and administration of Hadoop components ((YARN, MapReduce, HDFS, HBase, Zookeeper, Storm, Kafka, Spark, Pig and Hive)
3. Work closely with infrastructure, network, database, business intelligence and application teams to ensure business applications are highly available and performing within agreed on service levels.
4. Accountable for performance tuning and resource management of Hadoop clusters and MapReduce routines.
5. Strong Experience with LINUX based systems & scripting (either of Shell, Perl or Python).
6. Experience with configuration management tools like puppet, chef or salt.
7. Strong Experience with Configuring Security in Hadoop using Kerberos or PAM.
8. Good knowledge of directory services like LDAP & ADS and Monitoring tools like Nagios or Icinga2.
9. Strong troubleshooting skills of Hive, Pig, Hbase and JAVA Mapreduce codes/jobs.
10. Evaluate technical aspects of any change requests pertaining to the Cluster.
11. Research, identify and recommend technical and operational improvements resulting in improved reliability efficiencies in developing the Cluster.

For this role, they will be digging in on how much the hadoop admins knows from an architecture perspective, and asking if they're hands on.. this is not a code monkey job. The team will be trying to see what the candidates understand, and what they've actually.


If they have Map-R that's a plus, or any other distribution, if they understands the distribution architecture they will be fine


This role will involve a lot of hands on scripting, at least shell scripting and good linux skills and configuration and performance tuning, and integration, and Hadoop eco system tools.