Load-Balance Policy in Two Level-Cluster File System

In this paper, we explored a load-balancing algorithm in a cluster file system contains two levels of metadata-server, primary-level server quickly distributestasks to second-level servers depending on the closest load-balancing information. At the same time, we explored a method which accurately reflect I/O traffic and storage of storage-node： computing the heat-value of file, according to which we realized a more logical storage allocation. According to the experiment result, we conclude that this new algorithm shortens the executing time of tasks and improves the system performance compared with other load algorithm.

A Non-Forced-Write Atomic Commit Protocol for Cluster File Systems

Distributed metadata consistency is one of the critical issues of metadata clusters in distributed file systems. Existing methods to maintain metadata consistency generally need several log forced write operations. Since synchronous disk IO is very inefficient, the average response time of metadata operations is greatly increased. In this paper, an asynchronous atomic commit protocol （ACP） named Dual-Log （DL） is presented. It does not need any log forced write operations. Optimizing for distributed metadata operations involving only two metadata servers, DL mutually records the redo log in counterpart metadata servers by transferring through the low latency network. A crashed metadata server can redo the metadata operation with the redundant redo log. Since the latency of the network is much lower than the latency of disk IO, DL can improve the performance of distributed metadata service significantly. The prototype of DL is implemented based on local journal. The performance is tested by comparing with two widely used protocols, EP and S2PC-MP, and the results show that the average response time of distributed metadata operations is reduced by about 40%-60%, and the recovery time is only I second under 10 thousands uncompleted distributed metadata operations.

Building a Portable File System for Heterogeneous Clusters

Existing in-kernel distributed file systems cannot cope with the higher requirements in well- equipped cluster environments, especially when the system becomes larger and inevitably heterogeneous. TH-CluFS is a cluster file system designed for large heterogeneous systems. TH-CluFS is implemented completely in the user space by emulating the network file system (NFS) V2 server, and is easily portable to other portable operating system interface (POSIX)-compliant platforms with application programming/binary interface API/ABI compliance. In addition, TH-CluFS uses a serverless architecture which flexibly distributes data at file granularity and achieves a consistent file system view from distributed metadata. The global cache makes full use of the aggregated memories and disks in the cluster to optimize system performance. Experimental results suggest that although TH-CluFS is implemented as user-level components, it functions as a portable, single system image, and scalable cluster file system with acceptable performance sacrifices.

Architecture Design of a Variable Length Instruction Set VLIW DSP

The cost of the central register file and the size of the program code limit the scalability of very long instruction word（VLIW） processors with increasing numbers of functional units.This paper presents the architectural design of a six-way VLIW digital signal processor（DSP） with clustered register files.The architecture uses a variable length instruction set and supports dynamic instruction dispatching.The one-level memory system architecture of the processor includes 16-KB instruction and data caches and 16-KB instruction and data on-chip RAM.A compiler based on the Open64 was developed for the system.Evaluations show that the processor is suitable for high performance applications with a high code density and small program code size.