Improving Collective I/O Performance with Machine Learning Supported Auto-tuning

Abstract

Collective Input and output (I/O) is an essential approach in high performance computing (HPC) applications. The achievement of effective collective I/O is a nontrivial job due to the complex interdependencies between the layers of I/O stack. These layers provide the best possible I/O performance through a number of tunable parameters. Sadly, the correct combination of parameters depends on diverse applications and HPC platforms. When a configuration space gets larger, it becomes difficult for humans to monitor the interactions between the configuration options. Engineers has no time or experience for exploring good configuration parameters for each problem because of long benchmarking phase. In most cases, the default settings are implemented, often leading to poor I/O efficiency. I/O profiling tools can not tell the optimal default setups without too much effort to analyzing the tracing results. In this case, an auto-tuning solution for optimizing collective I/O requests and providing system administrators or engineers the statistic information is strongly required. In this paper, a study of the machine learning supported collective I/O auto-tuning including the architecture and software stack is performed. Random forest regression model is used to develop a performance predictor model that can capture parallel I/O behavior as a function of application and file system characteristics. The modeling approach can provide insights into the metrics that impact I/O performance significantly.

Applying neural networks to predict HPC-I/O bandwidth over seismic data on lustre file system for ExSeisDat

“…The paper first demonstrates the need for a flexible and efficient optimisation approach through a series of experimental benchmarks. This is similar to much existing research which examines how I/O performance can prove challenging in these scenarios [7][8][9][10][11][12]. There are two key contributions outlined in this paper.…”

“…The approach presented in [10], is the random forest regression modelling, used as the ML technique to predict the I/O bandwidth for only collective write operation in MPI-I/O Library. The accuracy of prediction is very high.…”

“…4. The prediction percentage accuracy has also been used as a metric for measuring the precision of ML models in couple of previous research works [10,12,15]. The Table 4 5, are computed using the following equations:…”

“…Previous research has shown that I/O performance prediction based on the different parameters settings can result in significant benefits [7][8][9][10][11]. Despite the key differences between this and existing research with respect to parameters and ML techniques, the prediction of SEG-Y file I/O performance beforehand is itself innovative.…”

Artificial neural networks based predictions towards the auto-tuning and optimization of parallel IO bandwidth in HPC system

“…The work presented in [22], examines bandwidth predictive modelling for MPI WRITE collective operations via random forest regression. The prediction accuracy values are extremely high, ranging around 82-99%, which depends on depth setting maximum value.…”