Investigating Machine Learning Algorithms for Modeling SSD I/O Performance for Container-Based Virtualization

Abstract

One of the cornerstones of the cloud provider business is to reduce hardware resources cost by maximizing their utilization. This is done through smartly sharing processor, memory, network and storage, while fully satisfying SLOs negotiated with customers. For the storage part, while SSDs are increasingly deployed in data centers mainly for their performance and energy efficiency, their internal mechanisms may cause a dramatic SLO violation. In effect, we measured that I/O interference may induce a 10x performance drop. We are building a framework based on autonomic computing which aims to achieve intelligent container placement on storage systems by preventing bad I/O interference scenarios. One prerequisite to such a framework is to design SSD performance models that take into account interactions between running processes/containers, the operating system and the SSD. These interactions are complex. In this paper, we investigate the use of machine learning for building such models in a container based Cloud environment. We have investigated five popular machine learning algorithms along with six different I/O intensive applications and benchmarks. We analyzed the prediction accuracy, the learning curve, the feature importance and the training time of the tested algorithms on four different SSD models. Beyond describing modeling component of our framework, this paper aims to provide insights for cloud providers to implement SLO compliant container placement algorithms on SSDs. Our machine learning-based framework succeeded in modeling I/O interference with a median Normalized Root-Mean-Square Error (NRMSE) of 2.5%.

Predicting file lifetimes for data placement in multi-tiered storage systems for HPC

“…Learning I/O Access Patterns to improve prefetching in SSDs [11] leverages the use of Long Short-Term Memory (LSTM) which are an extension of Recurent Neural Networks to predict at firmware level (inside a SSD) which Logical Block Address will be used. In [14], the authors have evaluated different learning algorithms, this time to model the overall performance of SSDs for Container-based Virtualization to avoid interference that would cause Service Level Objectives violations in a cloud environment Data-Jockey [32] is a data movement scheduling solution that aims to reduce the complexity of moving data across a hierarchy of storage. It uses user-made policies to balance the load across all storage nodes and jobs.…”

Predicting file lifetimes for data placement in multi-tiered storage systems for HPC

“…Learning I/O Access Patterns to improve prefetching in SSDs [11] leverages the use of Long Short-Term Memory (LSTM) which are an extension of Recurent Neural Networks to predict at firmware level (inside a SSD) which Logical Block Address will be used. In [14], the authors have evaluated different learning algorithms, this time to model the overall performance of SSDs for Container-based Virtualization to avoid interference that would cause Service Level Objectives violations in a cloud environment Data-Jockey [32] is a data movement scheduling solution that aims to reduce the complexity of moving data across a hierarchy of storage. It uses user-made policies to balance the load across all storage nodes and jobs.…”

Salamander: a Holistic Scheduling of MapReduce Jobs on Ephemeral Cloud Resources

“…• In Salamander, we considered a fixed capacity for the resources. However, the capacity can be reduced due to interference (e.g., I/O [27]) between co-located workloads. • We did not consider the different storage types such as SSD and HDD for optimizing performance and cost [28], [29].…”