Darko Makreshanski

Surname

Makreshanski

First Name

Darko

Phone

+41 44 632 72 58

Email

darkoma@inf.ethz.ch

Address

CAB E 61.3

 

I am a PhD student in the Systems Group since fall 2012 under the supervision of Prof. Gustavo Alonso. My main research interests revolve around data processing on modern hardware. Some keywords of my research include data analytics, hybrid workloads, shared query execution, hardware accelerators and novel micro-architectural features.

I am currently on the job market. Looking for interesting industry & industry research positions. CV

 

Education:

ETH Zurich (2012-) PhD student in the Systems Group
ETH Zurich (2010-2012) MSc ETH in Computer Science
Jacobs University Bremen (2007-2010) BSc in Electrical Engineering and Computer Science

  

Projects:

SharedDB

SharedDB is a relational database that is designed to handle large and complex transactional and analytic workloads. On modern multicore machines, traditional databases handle such workloads by executing queries one-at-a-time on separate cores which leads to unpredictable resource contention and query interaction.

SharedDB approaches this problem in two ways:

  • To handle large workloads, SharedDB executes batches of queries with aggressive sharing of data and computation across multiple different queries within a batch.
  • To handle complex workloads, SharedDB isolates performance between different query types by distributing the query execution across multiple independent operators.

 

Publications:

Darko Makreshanski, Jana Giceva, Claude Barthels, Gustavo Alonso

BatchDB: Efficient Isolated Execution of Hybrid OLTP+OLAP Workloads, SIGMOD 2017 (to appear)

Darko Makreshanski, Georgios Giannikis, Gustavo Alonso, Donald Kossmann

MQJoin: Efficient Shared Execution of Main-memory Joins, PVLDB 2016 Volume 9, No. 6, February 2016

Darko Makreshanski, Justin Levandoski, Ryan Stutsman

To Lock, Swap, or Elide: On the Interplay of Hardware Transactional Memory and Lock-Free Indexing, PVLDB 2015, Volume 8, No. 11, July 2015

Georgios Giannikis, Darko Makreshanski, Gustavo Alonso, Donald Kossmann

Shared Workload Optimization,  PVLDB 2014, Volume 7, No. 6, February 2014.

Georgios Giannikis, Darko Makreshanski, Gustavo Alonso, Donald Kossmann

DEMO: Workload Optimization using SharedDB,  SIGMOD 2013,  New York, USA, June 22-27, 2013.

Darko Makreshanski 

Shared, Parallel Database Join on Modern Hardware.Systems Group Master's Thesis, no. 53; Department of Computer Science, ETH Zurich, August 2012.

 

Internships:

June-August 2014

MSR Redmond

Topic: Hardware Transactional Memory and Lock-Free Data Structures

June-August 2013 

MSR Redmond

Topic: Data Analytics for Cipherbase 

 June-December 2011

 Google Zurich

Intern in the Search Quality team

 

Teaching

Big Data, Fall 2016

Data Modelling and Data Bases, Sprint 2016

Introduction to Machine Learning, Fall 2014

Advanced Systems Lab, Fall 2013, Fall 2015

Software Design, Spring 2013, Spring 2014