Metadata Performance Research Articles

Exploiting Flat Namespace to Improve File System Metadata Performance on Ultra-Fast, Byte-Addressable NVMs

The conventional file system provides a hierarchical namespace by structuring it as a directory tree. Tree-based namespace structure leads to inefficient file path walk and expensive namespace tree traversal, underutilizing ultra-low access latency and superior sequential performance provided by non-volatile memories (NVMs). This article proposes FlatFS+, an NVM file system that features a flat namespace architecture while providing a compatible hierarchical namespace view. FlatFS+ incorporates three novel techniques: the direct file path walk model, range-optimized B r tree, and compressed index key design with scan and write dual optimization, to fully exploit flat namespace to improve file system metadata performance on ultra-fast, byte-addressable NVMs. Evaluation results demonstrate that FlatFS+ achieves significant performance improvements for metadata-intensive benchmarks and real-world applications compared to other file systems.

A GPU-Accelerated In-Memory Metadata Management Scheme for Large-Scale Parallel File Systems

Driven by the increasing requirements of high-performance computing applications, supercomputers are prone to containing more and more computing nodes. Applications running on such a large-scale computing system are likely to spawn millions of parallel processes, which usually generate a burst of I/O requests, introducing a great challenge into the metadata management of underlying parallel file systems. The traditional method used to overcome such a challenge is adopting multiple metadata servers in the scale-out manner, which will inevitably confront with serious network and consistence problems. This work instead pursues to enhance the metadata performance in the scale-up manner. Specifically, we propose to improve the performance of each individual metadata server by employing GPU to handle metadata requests in parallel. Our proposal designs a novel metadata server architecture, which employs CPU to interact with file system clients, while offloading the computing tasks about metadata into GPU. To take full advantages of the parallelism existing in GPU, we redesign the in-memory data structure for the name space of file systems. The new data structure can perfectly fit to the memory architecture of GPU, and thus helps to exploit the large number of parallel threads within GPU to serve the bursty metadata requests concurrently. We implement a prototype based on BeeGFS and conduct extensive experiments to evaluate our proposal, and the experimental results demonstrate that our GPU-based solution outperforms the CPU-based scheme by more than 50% under typical metadata operations. The superiority is strengthened further on high concurrent scenarios, e.g., the high-performance computing systems supporting millions of parallel threads.

The Case for Custom Storage Backends in Distributed Storage Systems

For a decade, the Ceph distributed file system followed the conventional wisdom of building its storage backend on top of local file systems. This is a preferred choice for most distributed file systems today, because it allows them to benefit from the convenience and maturity of battle-tested code. Ceph’s experience, however, shows that this comes at a high price. First, developing a zero-overhead transaction mechanism is challenging. Second, metadata performance at the local level can significantly affect performance at the distributed level. Third, supporting emerging storage hardware is painstakingly slow. Ceph addressed these issues with BlueStore, a new backend designed to run directly on raw storage devices. In only two years since its inception, BlueStore outperformed previous established backends and is adopted by 70% of users in production. By running in user space and fully controlling the I/O stack, it has enabled space-efficient metadata and data checksums, fast overwrites of erasure-coded data, inline compression, decreased performance variability, and avoided a series of performance pitfalls of local file systems. Finally, it makes the adoption of backward-incompatible storage hardware possible, an important trait in a changing storage landscape that is learning to embrace hardware diversity.

Scalable Distributed Metadata Server Based on Nonblocking Transactions

Metadata performance scalability is critically important in high-performance computing when accessing many small files from millions of clients. This paper proposes a design of a scalable distributed metadata server, PPMDS, for parallel file systems using multiple key-value servers. In PPMDS, hierarchical namespace of a file system is efficiently managed by multiple servers. Multiple entries can be atomically updated using a nonblocking distributed transaction based on an algorithm of dynamic software transactional memory. This paper also proposes optimizations to further improve the metadata performance by introducing a server-side transaction processing, multiple readers, and a shared lock mode, which reduce the number of remote procedure calls and prevent unnecessary blocking. Performance evaluation shows the scalable performance up to 3 servers, and achieves 62,000 operations per second, which is 2.58x performance improvement compared to a single metadata performance.

Gfarm/BB — Gfarm File System for Node-Local Burst Buffer

Burst buffer has become a major component to meet the I/O performance requirement of HPC bursty traffic. This paper proposes Gfarm/BB that is a file system for a burst buffer efficiently exploiting node-local storage systems. Although node-local storages improve storage performance, they are only available during the job allocation. Gfarm/BB should have better access and metadata performance while it should be constructed on-demand before the job execution. To improve the read and write performance, it exploits the file descriptor passing and remote direct memory access (RDMA). It improves the metadata performance by omitting the persistency and the redundancy since it is a temporal file system. Using RDMA, writes and reads bandwidth are improved by 1.7x and 2.2x compared with IP over InfiniBand (IPoIB), respectively. It achieves 14 700 operations per second in the directory creation performance, which is 13.4x faster than the fully persistent and redundant case. The construction of Gfarm/BB takes 0.31 seconds using 2 nodes. IOR benchmark and ARGOT-IO application I/O benchmark show the scalable performance improvement by exploiting the locality of node-local storages. Compared with BeeOND, Gfarm/BB shows 2.6x and 2.4x better performance in IOR write and read benchmarks, respectively, and it shows 2.5x better performance in ARGOT-IO.

Explicit Data Correlations-Directed Metadata Prefetching Method in Distributed File Systems

Metadata performance in distributed file systems (DFS) is critical, due to the following trends: (a) the growing size of modern storage systems is expected to exceed billions of files and most files are small; (b) over half of the file accesses are metadata operations. In this work, we present SMeta, a metadata prefetching method that is seamlessly integrated into DFS for easy-of-use and significantly scales the metadata performance. Previous prefetching proposals primarily focus on mining groups of files that tend to be accessed together from the access history. Nevertheless, our study discovered that these solutions likely miss a huge number of correlated files whose co-occurrence frequency is not high enough. Unlike access correlations, we take a novel and completely different approach to explore explicit data correlations by understanding the reference relationships between files encoded in some forms of hyperlinks, which naturally exist in many applications. To embrace this new concept, SMeta explores correlations upon files are written via a light-weight pattern matching algorithm, stores correlations in the reserved extended attributes of file metadata to avoid changes in DFS APIs, and collapses multiple I/O rounds for accessing metadata of the target file and its data-correlated files into one round. A cost-efficient adaptive feedback mechanism is introduced to improve prefetching accuracy. We implemented SMeta atop of Ceph and evaluated it using synthetic and real system workloads. Compared to baselines, SMeta provides better metadata performance in terms of latency, throughput and scalability.

Combining Performance Testing and Metadata Models to Support Fault Detection and Diagnostics in Smart Buildings

Performance testing of components and subsystems of buildings is a promising practice for increasing energy efficiency and closing gaps between intended and actual performance of buildings. A typical shortcoming of performance testing is the difficulty of linking a failing test to a faulty or underperforming component. Furthermore, a failing test can also be linked to a wrongly configured performance test. In this paper, we present Building Metadata Performance Testing (BuMPeT), a method that addresses this shortcoming by using building metadata models to extend performance testing with fault detection and diagnostics (FDD) capabilities. We present four different procedures that apply BuMPeT to different data sources and components. We have applied the proposed method to a case study building, located in Denmark, to test its capacity and benefits. Additionally, we use two real case scenarios to showcase examples of failing performance tests in the building, as well as discovery of causes of underperformance. Finally, to examine the limits to the benefits of the applied procedure, a detailed elaboration of a hypothetical scenario is presented. Our findings demonstrate that the method has potential and it can serve to increase the energy efficiency of a wide range of buildings.

Extended Query Model for MOOC Education Resource Metadata Based on Big data

In traditional education resource metadata extended query methods, data recall is poor. In order to solve this problem, an extended query model for MOOC education resource metadata based on big data information fusion clustering scheduling is proposed. The semantic ontology model is adopted to analyse the storage structure of MOOC education resource metadata and extract binary semantic feature quantity of MOOC education resource metadata to construct a binary semantic decision model for extended query of MOOC education resource metadata; the integrated information processing technology of big data is adopted for extended query and adaptive scheduling of MOOC education resource metadata to improve the ability to retrieve and identify metadata. The simulation results show that the proposed method can provide an average precision ratio of 0.976 in extended query of MOOC education resource metadata and good data recall performance.

Distributed Metadata Management of Mass Storage System in High Energy Physics

In this contribution, we design and implement a dynamic and scalable distributed metadata management system(StarFS) to High Energy Physics (HEP) mass storage system. Particularly, we discuss the key technologies of the distributed metadata management system. We propose a new algorithm named Adaptive Directory Sub-tree Partition(ADSP) for metadata distribution. ADSP divides the filesystem namespace into sub-trees with directory granularity. Sub-trees will be stored on storage devices in flat structure, whose location and file attributes are recorded as extended attributes. The placement of sub-tree is adjusted adaptively according to the workload of metadata cluster so that the load balance could be improved and metadata cluster could be extended dynamically. Experiments show that ADSP achieves higher metadata performance and scalability compared to Gluster and Lustre. We also propose a new algorithm called Distributed Unified Layout(DULA) to improve dynamic scalability and efficiency of data positioning. A system with DULA could provide uniform data distribution and efficient data positioning. DULA is an improved consistent hashing algorithm which is able to locate data in O(1) without the help of routing information. Experiments prove that the better uniform data distribution and efficient data access can be achieved by DULA. This work is validated in YangBaJing Cosmic Ray(YBJ) experiment.

Direct data access protocols benchmarking on DPM

The Disk Pool Manager is an example of a multi-protocol, multi-VO system for data access on the Grid that went though a considerable technical evolution in the last years. Among other features, its architecture offers the opportunity of testing its different data access frontends under exactly the same conditions, including hardware and backend software. This characteristic inspired the idea of collecting monitoring information from various testbeds in order to benchmark the behaviour of the HTTP and Xrootd protocols for the use case of data analysis, batch or interactive. A source of information is the set of continuous tests that are run towards the worldwide endpoints belonging to the DPM Collaboration, which accumulated relevant statistics in its first year of activity. On top of that, the DPM releases are based on multiple levels of automated testing that include performance benchmarks of various kinds, executed regularly every day. At the same time, the recent releases of DPM can report monitoring information about any data access protocol to the same monitoring infrastructure that is used to monitor the Xrootd deployments. Our goal is to evaluate under which circumstances the HTTP-based protocols can be good enough for batch or interactive data access. In this contribution we show and discuss the results that our test systems have collected under the circumstances that include ROOT analyses using TTreeCache and stress tests on the metadata performance.

Efficient and Scalable Metadata Management in EB-Scale File Systems

Efficient and scalable distributed metadata management is critically important to overall system performance in large-scale distributed file systems, especially in the EB-scale era. Hash-based mapping and subtree partitioning are state-of-the-art distributed metadata management schemes. Hash-based mapping evenly distributes workload among metadata servers, but it eliminates all hierarchical locality of metadata. Subtree partitioning does not uniformly distribute workload among metadata servers, and metadata needs to be migrated to keep the load balanced roughly. Distributed metadata management is relatively difficult since it has to guarantee metadata consistency. Meanwhile, scaling metadata performance is more complicated than scaling raw I/O performance. The complexity further rises with distributed metadata. It results in a primary goal that is to improve metadata management scalability while paying attention to metadata consistency. In this paper, we present a ring-based metadata management mechanism named Dynamic Ring Online Partitioning (DROP). It can preserve metadata locality using locality-preserving hashing, keep metadata consistency, as well as dynamically distribute metadata among metadata server cluster to keep load balancing. By conducting performance evaluation through extensive trace-driven simulations and a prototype implementation, experimental results demonstrate the efficiency and scalability of DROP.

Dynamic file system semantics to enable metadata optimizations in PVFS

AbstractModern file systems maintain extensive metadata about stored files. While metadata typically is useful, there are situations when the additional overhead of such a design becomes a problem in terms of performance. This is especially true for parallel and cluster file systems, where every metadata operation is even more expensive due to their architecture. In this paper several changes made to the parallel cluster file system Parallel Virtual File System (PVFS) are presented. The changes target at the optimization of workloads with large numbers of small files. To improve the metadata performance, PVFS was modified such that unnecessary metadata is not managed anymore. Several tests with a large quantity of files were performed to measure the benefits of these changes. The tests have shown that common file system operations can be sped up by a factor of two even with relatively few changes. Copyright © 2009 John Wiley & Sons, Ltd.

HFS

Two oft-cited file systems, the Fast File System (FFS) and the Log-Structured File System (LFS), adopt two sharply different update strategies--- update-in-place and update-out-of-place . This paper introduces the design and implementation of a hybrid file system called hFS, which combines the strengths of FFS and LFS while avoiding their weaknesses. This is accomplished by distributing file system data into two partitions based on their size and type. In hFS, data blocks of large regular files are stored in a data partition arranged in a FFS-like fashion, while metadata and small files are stored in a separate log partition organized in the spirit of LFS but without incurring any cleaning overhead . This segregation makes it possible to use more appropriate layouts for different data than would otherwise be possible. In particular, hFS has the ability to perform clustered I/O on all kinds of data---including small files, metadata, and large files. We have implemented a prototype of hFS on FreeBSD and have compared its performance against three file systems, including FFS with Soft Updates, a port of NetBSD's LFS, and our lightweight journaling file system called yFS. Results on a number of benchmarks show that hFS has excellent small file and metadata performance. For example, hFS beats FFS with Soft Updates in the range from 53% to 63% in the PostMark benchmark.