HighPerTimer – Timing and Controlled Waiting
A network experiment needs more than a fast connection. It also needs to know when a packet was sent, how long an operation took and when the next action should start. If the measurement program adds too much overhead or resumes late, that timing becomes part of the result.
HighPerTimer is a Linux library for these local timing operations. It provides a common interface for timestamps, interval calculations and controlled waiting, giving applications a consistent way to work with suitable time sources on the host.
© FILA
Timestamps connect events to their time of occurrence. This makes process progression and intervals available for study in an experiment.
The library brings time-source selection and a consistent representation of time values into one interface. The founding paper describes access to hardware counters, with an operating-system time source as an alternative. Which source is suitable depends on the platform.
The C++ interface represents time values as objects with arithmetic and comparison operations. Initialisation selects a suitable time source. This connects platform-dependent counters with a common application interface, allowing timestamps and intervals to be handled consistently within the same program.
For a waiting operation, the interesting value is when the program actually resumes. Evaluations therefore consider deviation from the requested time, call overhead and CPU demand together. The same operation can behave differently under load or inside a virtual machine.
High time resolution is useful, but it is distinct from accuracy and from synchronisation between computers. HighPerTimer addresses local timing; LTest connects such operations with distributed network measurement. Together, they help make the timing assumptions behind a transport experiment explicit.
To use HighPerTimer in your own experiment, discuss the appropriate implementation and documentation with the FILA team. Include your target platform, measurement or transport scenario and required comparison data.
Selected publications
- Irina Fedotova, Eduard Siemens, Hao Hu (2013). A High-precision Time Handling Library. ICNS 2013: The Ninth International Conference on Networking and Services, 193–199 (IARIA).
- Veronika Kirova, Kirill Karpov, Eduard Siemens, Irina Zander, Oksana Vasylenko, Dmitry Kachan, Sergii Maksymov (2019). Impact of Modern Virtualization Methods on Timing Precision and Performance of High-Speed Applications. Future Internet, 11(8), article 179 (MDPI) DOI: 10.3390/fi11080179.
