Trace Description

The three datasets contain the performance metrics of 520,  527 and 547 VMs from a distributed datacenter of Materna. Materna is a full service provider in the premium segment and has been successfully implementing ITC projects for their customers for more than 35 years. Thier client list reads like the “Who’s Who” of German companies and public sector organisations. Throughout Europe there are around 1,700 employees working for Materna, including highly specialized and certified consultants, software developers, software architects and project managers as well as editors and marketing experts - all with well-proven project experience. Materna covers the complete spectrum of services you would expect from a full service provider: from strategy and consulting services through to implementation and operations. The range of services is organised in six Business Lines: IT Factory, Digital Enterprise, Government and Communications, as well as cbs, a SAP business consultancy company.

Each file of teh three datasets contains the performance metrics of a specific VM. These files are organized according by traces: Materna-trace-1, Materna-trace-2 and Materna-trace-3.

The first trace consists of 520 VMs, the second trace consits of 527 VMs and the third trace consits of 547 VMs. The traces were taken in the distributed Materna Data Centers in Dortmund over a timespan of three months. Each trace represents one month of data. Running VMs in the traces are mostly the same in all three datasets. The workloads in the traced VMs are highly critical business applications of internatinaly known companies. All datasets contain a special text-file listing relevant trace-metrics. The traces were taken on a VMware ESX environment using the following physikal resources:

49 Hosts, 69 CPU cores and 6780 GB RAM   

The format of each file is row-based, each row represent an observation of the performance metrics. Each column of a row is separate by “;\t” The format of each row is


  1. Timestamp: number of milliseconds since 1970-01-01.
  2. CPU cores: number of virtual CPU cores provisioned.
  3. CPU capacity provisioned (CPU requested): the capacity of the CPUs in terms of MHZ, it equals to number of cores x speed per core.
  4. CPU usage: in terms of MHZ.
  5. CPU usage: in terms of percentage
  6. Memory provisioned (memory requested): the capacity of the memory of the VM in terms of KB.
  7. Memory usage: the memory that is actively used in terms of KB.
  8. Memory usage: in terms of %
  9. Disk write throughput: in terms of KB/s
  10. Disk size: In terms of GB (total sum of all virtual HDDs)
  11. Network received throughput: in terms of KB/s
  12. Network transmitted throughput: in terms of KB/s

Research work

1. Andreas Kohne, Marc Spohr, Lars Nagel, and Olaf Spinczyk. FederatedCloudSim: A SLA-aware Federated Cloud Simulation Framework. In Proceedings of the 2Nd International Workshop on CrossCloud Systems, CCB '14, pages 3:1 - 3:5, New York, USA, 2014. ACM

2. Andreas Kohne, Damian Pasternak, Lars Nagel, and Olaf Spinczyk. Evaluation of SLA-based Decision Strategies for VM Scheduling in Cloud Data Centers. In Proceedings of the 3rdWorkshop on CrossCloud Infrastructures & Platforms, CrossCloud 2016, London, April 2016. AMC

Trace Download

Copyright note
This trace was graciously provided by Materna GmbH Information & Communications, Dortmund, Germany. To use this traces, you must include an acknowledgement to the source of the data in any published material that refers to the data. Please refer to the two papers above, and please also consider referring to the Grid Workloads Archive in the acknowledgements.

Download the GWA-T13-materna-trace

Side content