Thomas Bodner

Research Assistant, PhD Candidate

Email: thomas.bodner(at)hpi.de
Phone: +49 (331) 5509 - 3934
Address: August-Bebel-Str. 88, 14482 Potsdam
Office: Campus II, V 2.05
Office Hours: Just stop by or mail/call ahead for an appointment
Profiles: DBLP, Google Scholar, ResearchGate, GitHub, LinkedIn

I am a Ph.D. student in the Enterprise Platform and Integration Concepts Group at HPI, advised by Professor Hasso Plattner. The goal of my research is to make database systems cheaper and faster through the unique capabilities of cloud environments. I am particularly interested in all aspects around query processing on serverless cloud infrastructure. Before joining HPI, I built database systems at SAP, TU Berlin, UC Irvine, and IBM.

Research Area: Elastic Query Processing

Research Interests:

Cloud infrastructure benchmarking
Query execution
Query optimization

Teaching

Lectures:

Trends and Concepts in the Software Industry I (2021, 2020, 2019, 2018)

Seminars:

Trends and Concepts in the Software Industry II (2020)
Research and Implementation of Database Concepts (2020)
Develop your own Database (2019)

Bachelor's Projects:

Usage Mining for the Enterprise (2018/2019)

Master's Projects:

Building an Elastic Query Engine on Serverless Cloud Infrastructure (2021)
Performance Engineering for Cloud-based Database Systems (2020)

Master's Theses:

Query Compilation for Distributed Execution with Cloud Functions (Julian Menzler, 2021)
Straggler Mitigation in Distributed Query Execution on Cloud Functions (Fabian Engel, 2021)
Elastic Query Execution via Short-lived and Stateless Cloud Functions (Jan Mensch, 2020)
Network Request Handling in Database Systems (Toni Stachewicz, 2019)
Data-dependent Implicit Authorizations for Fine-grained Database Access Control (Dennis Hempfing, 2018)
Pushing Down User-defined Functionality in Distributed Log-centric Big Data Stacks (Josephine Rückert, 2017)

Publications

Bodner, T.: Elastic Query Processing on Function as a Service Platforms.Proceedings of the VLDB PhD Workshop (2020).
[ Abstract ] [ BibTeX ]

Modern analytics workloads are not predictable anymore and require database systems to be able to adapt to their performance demands and cost constraints in an instant. Existing database architectures, however, are incapable of meeting this degree of elastic scalability. For this reason, we propose a novel architecture based on function as a service platforms. This architecture and the concepts surrounding it are being implemented in our research prototype Skyrise.
@inproceedings{bodner2020elastic, abstract = {Modern analytics workloads are not predictable anymore and require database systems to be able to adapt to their performance demands and cost constraints in an instant. Existing database architectures, however, are incapable of meeting this degree of elastic scalability. For this reason, we propose a novel architecture based on function as a service platforms. This architecture and the concepts surrounding it are being implemented in our research prototype Skyrise.}, author = {Bodner, Thomas}, booktitle = {Proceedings of the VLDB PhD Workshop}, keywords = {skyrise}, title = {Elastic Query Processing on Function as a Service Platforms}, year = 2020 }
Goel, A., Pound, J., Auch, N., Bumbulis, P., MacLean, S., Färber, F., Gropengiesser, F., Mathis, C., Bodner, T., Lehner, W.: Towards Scalable Real-time Analytics: An Architecture for Scale-out of OLXP Workloads.Proceedings of the VLDB Endowment. (2015).
[ Abstract ] [ BibTeX ] [ URL ]

We present an overview of our work on the SAP HANA Scale-out Extension, a novel distributed database architecture designed to support large scale analytics over real-time data. This platform permits high performance OLAP with massive scale-out capabilities, while concurrently allowing OLTP workloads. This dual capability enables analytics over real-time changing data and allows fine grained user-specified service level agreements (SLAs) on data freshness. We advocate the decoupling of core database components such as query processing, concurrency control, and persistence, a design choice made possible by advances in high-throughput low-latency networks and storage devices. We provide full ACID guarantees and build on a logical timestamp mechanism to provide MVCC-based snapshot isolation, while not requiring synchronous updates of replicas. Instead, we use asynchronous update propagation guaranteeing consistency with timestamp validation. We provide a view into the design and development of a large scale data management platform for real-time analytics, driven by the needs of modern enterprise customers.
@article{DBLP:journals/pvldb/GoelPABMFGMBL15, abstract = {We present an overview of our work on the SAP HANA Scale-out Extension, a novel distributed database architecture designed to support large scale analytics over real-time data. This platform permits high performance OLAP with massive scale-out capabilities, while concurrently allowing OLTP workloads. This dual capability enables analytics over real-time changing data and allows fine grained user-specified service level agreements (SLAs) on data freshness. We advocate the decoupling of core database components such as query processing, concurrency control, and persistence, a design choice made possible by advances in high-throughput low-latency networks and storage devices. We provide full ACID guarantees and build on a logical timestamp mechanism to provide MVCC-based snapshot isolation, while not requiring synchronous updates of replicas. Instead, we use asynchronous update propagation guaranteeing consistency with timestamp validation. We provide a view into the design and development of a large scale data management platform for real-time analytics, driven by the needs of modern enterprise customers.}, author = {Goel, Anil and Pound, Jeffrey and Auch, Nathan and Bumbulis, Peter and MacLean, Scott and Färber, Franz and Gropengiesser, Francis and Mathis, Christian and Bodner, Thomas and Lehner, Wolfgang}, journal = {Proceedings of the VLDB Endowment}, keywords = {distributed_database htap}, title = {Towards Scalable Real-time Analytics: An Architecture for Scale-out of OLXP Workloads}, year = 2015 }
Alexandrov, A., Schiefer, B., Poelman, J., Ewen, S., Bodner, T., Markl, V.: Myriad - Parallel Data Generation on Shared-nothing Architectures.Proceedings of the 1st Workshop on Architectures and Systems for Big Data. (2011).
[ Abstract ] [ BibTeX ] [ URL ]

The need for efficient data generation for the purposes of testing and benchmarking newly developed massively-parallel data processing systems has increased with the emergence of Big Data problems. As synthetic data model specifications evolve over time, the data generator programs implementing these models have to be adapted continuously -- a task that often becomes more tedious as the set of model constraints grows. In this paper we present Myriad - a new parallel data generation toolkit. Data generators created with the toolkit can quickly produce very large datasets in a shared-nothing parallel execution environment, while at the same time preserve with cross-partition dependencies, correlations and distributions in the generated data. In addition, we report on our efforts towards a benchmark suite for large-scale parallel analysis systems that uses Myriad for the generation of OLAP-style relational datasets.
@article{alexandrov2011myriad, abstract = {The need for efficient data generation for the purposes of testing and benchmarking newly developed massively-parallel data processing systems has increased with the emergence of Big Data problems. As synthetic data model specifications evolve over time, the data generator programs implementing these models have to be adapted continuously -- a task that often becomes more tedious as the set of model constraints grows. In this paper we present Myriad - a new parallel data generation toolkit. Data generators created with the toolkit can quickly produce very large datasets in a shared-nothing parallel execution environment, while at the same time preserve with cross-partition dependencies, correlations and distributions in the generated data. In addition, we report on our efforts towards a benchmark suite for large-scale parallel analysis systems that uses Myriad for the generation of OLAP-style relational datasets.}, author = {Alexandrov, Alexander and Schiefer, Berni and Poelman, John and Ewen, Stephan and Bodner, Thomas and Markl, Volker}, journal = {Proceedings of the 1st Workshop on Architectures and Systems for Big Data}, keywords = {benchmarking distributed_database}, title = {Myriad - Parallel Data Generation on Shared-nothing Architectures}, year = 2011 }

Patents

Berg, G., Wenz, A., Hoeppner, B., Bodner, T., Cherepanova, O., Steffen, L., Siebert, J., Hennemann, D., Schulze, P., Dobler, K., Kahl, K., Beneke, P., Hoberg, P.:
Generation of Bots Based on Observed Behavior.
US 2020.

Bumbulis, P., Pound, J., Auch, N., Goel, A., Ringwald, M., Bodner, T., MacLean, S.:
An Algorithm for Consistent Replication of Log-Structured Data.
EU 2017 and US 2017.

Thomas Bodner

Research Assistant, PhD Candidate

Research Area: Elastic Query Processing

Teaching

Publications

Patents

Open Positions

News

19.11.2020 | Best Paper Award received at CSEE&T 2020

28.07.2020 | "An Experimental Evaluation of Index Selection Algorithms" published in PVLDB

04.05.2020 | "Quantifying TPC-H Choke Points and Their Optimizations" published in PVLDB

01.05.2020 | Change of Chair Representation

03.03.2020 | Best Master Thesis in Education Technologies 2019

Literature

Contact