Time filter

Source Type

Sun Z.,CAS Institute of Computing Technology | Sun Z.,Zhongke Blue Whale Information Technologies Co. | Sun Z.,University of Chinese Academy of Sciences | Guo M.,CAS Institute of Computing Technology | And 10 more authors.
Proceedings - 2015 IEEE 17th International Conference on High Performance Computing and Communications, 2015 IEEE 7th International Symposium on Cyberspace Safety and Security and 2015 IEEE 12th International Conference on Embedded Software and Systems, HPCC-CSS-ICESS 2015 | Year: 2015

An inconsistent update in a parity-based RAID can cause data loss when a disk failure occurs. There are two ways to restore the consistency: 1, Re-calculating parities, which does not work with disk failures. 2, Using transactions to record and replay updated data-contents in the transaction log, which is a single point of failure. It needs much bandwidth to replicate a transaction log of data-contents over network. We proposed the Above-Logging Transaction (ALT), and designed the Redundancy Array of Log-structured Disks (RALD). All updates histories of data-contents were on log-structured disks. ALTs recorded updated data's addresses on those log-structured disks into the transaction log by which ALTs mapped the consistent blocks into the read-only snapshot's space after they finished. To avoid the single point of failure, the RALD copied the ALTs log into those log-structured disks. By using flushes, the RALD utilized write-buffers safely. We had evaluated the RALD on SATA3 HDDs. The RALD can restore consistency from disk failures plus system crashes. The ALT's logs consumes little bandwidth, less than 0.5MB/s per replica. Compared to the Linux MD RAID: on write-dominant traces, the RALD has 20% to 190% more IOPS, on read-dominant traces, the RALD has 30% to 160% more IOPS if its internal caches have effective read-ahead, otherwise it has 10% to 50% less IOPS. © 2015 IEEE. Source

Sun Z.,CAS Institute of Computing Technology | Sun Z.,Zhongke Blue Whale Information Technologies Co. | Sun Z.,University of Chinese Academy of Sciences | Xu L.,CAS Institute of Computing Technology | And 9 more authors.
Jisuanji Yanjiu yu Fazhan/Computer Research and Development | Year: 2015

Storage area network (SAN) is important for network storage. We construct BWRAID, a distributed RAID in a SAN, from commodity hardware. The original version BWRAID has a symmetric architecture. However, it has three problems: Firstly, it reads data to re-calculate parity when it expands, and the re-calculation consumes much IO and time. Secondly, it recovers data to one storage node (SN), and recovery can be more efficient if parallel on multi nodes. Thirdly, its data layout is bad for IO, that makes its internal RAID4 have many costly read-modify-write updates even on sequential writes. To solve these problems, we propose a network declustering BWRAID. It has an asymmetric architecture like a declustering RAID, but it is declustered by equal size virtual disks instead of blocks. It is expanded by moving virtual disks without calculation. It runs multi recovery in parallel as the number of nodes involved in each recovery is less than the total number of nodes in the system. To optimize IO, we change its data layout to express user IO space by internal RAID4 stripes. We also provide algorithms to search suitable virtual disks for system allocation, expansion, or recovery. Experiments show that network declustering BWRAID is better than the original one. It expands the system without calculating parity five to eight times faster, and its parallel recovery is multi times faster, and it increases the IO performance with the new data layout. © 2015, Science Press. All right reserved. Source

Discover hidden collaborations