Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userExploring Scientific Application Performance Using Large Scale Object Storage
223
0
0.0
(
0
)
Added by
Stefano Markidis Prof.
Publication date
2018
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
One of the major performance and scalability bottlenecks in large scientific applications is parallel reading and writing to supercomputer I/O systems. The usage of parallel file systems and consistency requirements of POSIX, that all the traditional HPC parallel I/O interfaces adhere to, pose limitations to the scalability of scientific applications. Object storage is a widely used storage technology in cloud computing and is more frequently proposed for HPC workload to address and improve the current scalability and performance of I/O in scientific applications. While object storage is a promising technology, it is still unclear how scientific applications will use object storage and what the main performance benefits will be. This work addresses these questions, by emulating an object storage used by a traditional scientific application and evaluating potential performance benefits. We show that scientific applications can benefit from the usage of object storage on large scales.
rate research
Read More
Access libraries such as ROOT and HDF5 allow users to interact with datasets using high level abstractions, like coordinate systems and associated slicing operations. Unfortunately, the implementations of access libraries are based on outdated assumptions about storage systems interfaces and are generally unable to fully benefit from modern fast storage devices. The situation is getting worse with rapidly evolving storage devices such as non-volatile memory and ever larger datasets. This project explores distributed dataset mapping infrastructures that can integrate and scale out existing access libraries using Cephs extensible object model, avoiding re-implementation or even modifications of these access libraries as much as possible. These programmable storage extensions coupled with our distributed dataset mapping techniques enable: 1) access library operations to be offloaded to storage system servers, 2) the independent evolution of access libraries and storage systems and 3) fully leveraging of the existing load balancing, elasticity, and failure management of distributed storage systems like Ceph. They also create more opportunities to conduct storage server-local optimizations specific to storage servers. For example, storage servers might include local key/value stores combined with chunk stores that require different optimizations than a local file system. As storage servers evolve to support new storage devices like non-volatile memory, these server-local optimizations can be implemented while minimizing disruptions to applications. We will report progress on the means by which distributed dataset mapping can be abstracted over particular access libraries, including access libraries for ROOT data, and how we address some of the challenges revolving around data partitioning and composability of access operations.
To accommodate the needs of large-scale distributed P2P systems, scalable data management strategies are required, allowing applications to efficiently cope with continuously growing, highly dis tributed data. This paper addresses the problem of efficiently stor ing and accessing very large binary data objects (blobs). It proposesan efficient versioning scheme allowing a large number of clients to concurrently read, write and append data to huge blobs that are fragmented and distributed at a very large scale. Scalability under heavy concurrency is achieved thanks to an original metadata scheme, based on a distributed segment tree built on top of a Distributed Hash Table (DHT). Our approach has been implemented and experimented within our BlobSeer prototype on the Grid5000 testbed, using up to 175 nodes.
Large scale cloud services use Key Performance Indicators (KPIs) for tracking and monitoring performance. They usually have Service Level Objectives (SLOs) baked into the customer agreements which are tied to these KPIs. Dependency failures, code bugs, infrastructure failures, and other problems can cause performance regressions. It is critical to minimize the time and manual effort in diagnosing and triaging such issues to reduce customer impact. Large volume of logs and mixed type of attributes (categorical, continuous) in the logs makes diagnosis of regressions non-trivial. In this paper, we present the design, implementation and experience from building and deploying DeCaf, a system for automated diagnosis and triaging of KPI issues using service logs. It uses machine learning along with pattern mining to help service owners automatically root cause and triage performance issues. We present the learnings and results from case studies on two large scale cloud services in Microsoft where DeCaf successfully diagnosed 10 known and 31 unknown issues. DeCaf also automatically triages the identified issues by leveraging historical data. Our key insights are that for any such diagnosis tool to be effective in practice, it should a) scale to large volumes of service logs and attributes, b) support different types of KPIs and ranking functions, c) be integrated into the DevOps processes.
Erasure codes are an integral part of many distributed storage systems aimed at Big Data, since they provide high fault-tolerance for low overheads. However, traditional erasure codes are inefficient on reading stored data in degraded environments (when nodes might be unavailable), and on replenishing lost data (vital for long term resilience). Consequently, novel codes optimized to cope with distributed storage system nuances are vigorously being researched. In this paper, we take an engineering alternative, exploring the use of simple and mature techniques -juxtaposing a standard erasure code with RAID-4 like parity. We carry out an analytical study to determine the efficacy of this approach over traditional as well as some novel codes. We build upon this study to design CORE, a general storage primitive that we integrate into HDFS. We benchmark this implementation in a proprietary cluster and in EC2. Our experiments show that compared to traditional erasure codes, CORE uses 50% less bandwidth and is up to 75% faster while recovering a single failed node, while the gains are respectively 15% and 60% for double node failures.
Scientific computing sometimes involves computation on sensitive data. Depending on the data and the execution environment, the HPC (high-performance computing) user or data provider may require confidentiality and/or integrity guarantees. To study the applicability of hardware-based trusted execution environments (TEEs) to enable secure scientific computing, we deeply analyze the performance impact of AMD SEV and Intel SGX for diverse HPC benchmarks including traditional scientific computing, machine learning, graph analytics, and emerging scientific computing workloads. We observe three main findings: 1) SEV requires careful memory placement on large scale NUMA machines (1$times$$-$3.4$times$ slowdown without and 1$times$$-$1.15$times$ slowdown with NUMA aware placement), 2) virtualization$-$a prerequisite for SEV$-$results in performance degradation for workloads with irregular memory accesses and large working sets (1$times$$-$4$times$ slowdown compared to native execution for graph applications) and 3) SGX is inappropriate for HPC given its limited secure memory size and inflexible programming model (1.2$times$$-$126$times$ slowdown over unsecure execution). Finally, we discuss forthcoming new TEE designs and their potential impact on scientific computing.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
