While we are extremely proud that our abstract came 2nd (out of 107) in the 2014 Hadoop Summit in Amsterdam (see you all there in April 2-3), we will not stop there and our plan is to continue the hard work and we’re looking forward to meet you at 2014 Hadoop Summit in San Jose. We would like to ask for your support by submitting your vote for our session in the largest Hadoop conference in the world.
Please use the following link to vote, or read our abstract below.
Should you have any questions regarding our abstract, the technical solution or implementation detailsm feel free to contact us or check our GitHub page.
Moving to Hadoop 2.0/YARN at SequenceIQ
A showcase of our efforts to bring all our Hadoop based applications under one common cluster management framework – YARN. Our deployment consists of MR2, HBase, Mahout and Hive-all running within one single auto-scaling cluster. We have faced many challenges such as load imbalances, SLA misses, cluster scheduling and VM container deployments – and would like to share our struggle and solution with the community. As a startup, cost savings is important for us – switching to Hadoop 2.0 helped us save significant costs through better utilization of our hardware and cloud VMs. Our decision and investment of moving to YARN has paid off – and opened up new business and technical opportunities.