Embedded Files
I'm a computer scientist and software engineer at Databricks, where I work on making big data processing fast and reliable as part of the Spark SQL Performance team. Prior to joining Databricks I was a PhD student at Delft University of Technology, where I was advised by Dick Epema. I received my PhD in computer science after defending my thesis entitled "Optimizing the Performance of Data Analytics Frameworks".
I'm a computer scientist and software engineer at Databricks, where I work on making big data processing fast and reliable as part of the Spark SQL Performance team. Prior to joining Databricks I was a PhD student at Delft University of Technology, where I was advised by Dick Epema. I received my PhD in computer science after defending my thesis entitled "Optimizing the Performance of Data Analytics Frameworks".
In my research I designed several scheduling system that optimize the the operation and the performance of large-scale data processing systems. I seek to make big data processing in the cloud a commodity by solving complex problems with simple and intuitive solutions that are widely applicable.
In my research I designed several scheduling system that optimize the the operation and the performance of large-scale data processing systems. I seek to make big data processing in the cloud a commodity by solving complex problems with simple and intuitive solutions that are widely applicable.
PhD Dissertation
PhD Dissertation
Projects
Projects
1. Spot Markets
2. Datacenter Scheduling
3. Random Query Generation
Invited Talks
- Correctness and Performance of Apache Spark SQL, Spark + AI Summit, London, October 2018
- Bid Advising in Cloud Spot Markets, Vrije Universiteit, Amsterdam, June 2018
Selected Publications
- B. Ghit and D. Epema. Better Safe than Sorry: Grappling with Failures of In-Memory Data Analytics Frameworks. ACM HPDC, Washington, June 2017.
- B. Ghit and D. Epema. Tyrex: Size-based Resource Allocation in MapReduce Frameworks. IEEE/ACM CCGrid, Cartagena, May 2016 .
- B. Ghit and D. Epema. Reducing Job Slowdown Variability for Data-Intensive Workloads. IEEE Mascots, Atlanta, October 2015.
- B. Ghit, N. Yigitbasi, A. Iosup, D. Epema. Balanced Resource Allocations Across Multiple Dynamic MapReduce Clusters. ACM Sigmetrics, Austin, July 2014.
- B. Ghit, N. Yigitbasi, D. Epema. Resource Management for Dynamic MapReduce Clusters in Multicluster Systems. Supercomputing MTAGS Workshop, Salt Lake City, November 2012. Best Paper Award.
For the full list of my publications visit my Google Scholar profile.
Chair: IEEE ICAC 2019 (poster and demo)
PC Member: IEEE CCGrid 2019, ISPDC 2017
Reviewer: IEEE TPDS, Elsevier PEVA, Trans. on Cloud Computing, Cluster Computing
Google Sites
Report abuse