- مبلغ: ۸۶,۰۰۰ تومان
- مبلغ: ۹۱,۰۰۰ تومان
Only little is publicly known about traffic in non-educational data centers. Recent studies made some knowledge available, which gives us the opportunity to create more realistic traffic models for data-center research. We used this knowledge to create the first publicly available traffic generator that produces realistic traffic between hosts in data centers of arbitrary size. We characterize traffic by using six probability distribution functions and concentrate on the generation of traffic on flow-level. The distribution functions are easily exchangeable to enable using up-to-date traffic characteristics whenever new data is available from publications or own experiments. Moreover, in data centers, traffic between hosts in the same rack and hosts in different racks have different properties. We model this phenomenon, making our generated traffic very realistic. We carefully evaluated our approach and conclude that it reproduces these characteristics with accuracy.
The traffic generator DCT2Gen presented in this work creates a Layer 4 traffic schedule for arbitrary sized data centers. When the scheduled payloads are transported using TCP, this produces Layer 2 traffic with properties that can be defined in advance using a set of probability distributions. Our evaluation showed that DCT2Gen reproduces these properties with high accuracy. Solely the generated flow inter-arrival time distribution does not match our chosen target distribution. As DCT2Gen manipulates the inter-arrival time distribution to adjust the amount of flows to the given traffic matrices, this is not surprising. We suspect that this difference will be significantly smaller when using input data of higher quality. Given that DCT2Gen generates a schedule of payload transmissions between all hosts in a data center it is suitable for simulations, network emulations, and testbed experiments. Using our generated traffic schedule combined with a large-scale network emulator such as MaxiNet, novel networking ideas can be evaluated under highly realistic conditions which brings new ideas a step closer to deployment in production environments.