Sung Hoon Baek

A Hybrid Flash File System Based on NOR and NAND Flash Memories for Embedded Devices

This paper presents a hybrid flash file system (HFFS) based on both NOR flash and NAND flash memory. In a conventional NAND flash-based flash file system, there is a trade-off between life span and durability in the frequent writing of small amounts of data. Because NAND flash supports only a page-level I/O, at least one page is wasted in the synchronous writing of small amounts of data. The wasting of pages reduces the utilization and life span of the NAND flash. To alleviate the utilization problem, some NAND flash-based flash file systems write small amounts of data asynchronously with RAM buffers, though buffering in RAM decreases the durability of the system. Our HFFS eliminates the trade-off between life span and durability. It synchronously stores data as a log in the NOR flash, whenever we append small amounts of data to a file. The merged logs are then flushed to the NAND flash in a page-aligned fashion. The implementation of our HFFS is based on our previous NAND flash-based file system, called CFFS, The experimental results reveal that our HFFS provides a longer life span than a conventional NAND flash-based synchronous flash file system with a similar level of durability.

Matrix-Stripe-Cache-Based Contiguity Transform for Fragmented Writes in RAID-5

Given that contiguous reads and writes between a cache and a disk outperform fragmented reads and writes, fragmented reads and writes are forcefully transformed into contiguous reads and writes via a proposed matrix-stripe-cache-based contiguity transform (MSC-CT) method which employs a rule of consistency for data integrity at the block level and a rule of performance that ensures no performance degradation. MSC-CT performs for reads and writes, both of which are produced by write requests from a host as a write request from a host employs reads for parity update and writes to disks in a redundant array of independent disks (RAID)-5. MSC-CT is compatible with existing disk technologies. The proposed implementation in a Linux kernel delivers a peak throughput that is 3.2 times higher than a case without MSC-CT on representative workloads. The results demonstrate that MSC-CT is extremely simple to implement, has low overhead, and is ideally suited for RAID controllers not only for random writes but also for sequential writes in various realistic scenarios.

Massive Stripe Cache And Prefetching for Massive File I/O

This paper presents a massive stripe prefetching (MSP) scheme on a massive stripe cache (MSC) structure, which has been implemented in a Linux kernel. Our schemes efficiently perform on massive files such as audio/video files and a bulk of small files. MSC exploits spatial locality. MSP discriminates metadata reads and contiguous data reads in the stripe level, thereby boosting the read performance independently of a file-based prefetching.

Write back routine for JFFS2 efficient i/o

When flash memory is used as a storage in embedded systems, block level translation layer is required between conventional filesystem and flash memory chips due to its physical characteristics. A far more efficient use of it is the design of a filesystem itself without no extra layer of translation. However, since flash filesystem does not use block device layer, it cannot utilize deferred I/O although deferred I/O enhances write latency by delaying the flushing jobs. Linux operating system generally uses the write back routine for deferred I/O using kernel thread, which writes back dirty pages and buffers through the block device layer. In this paper, we design and implement efficient I/O for JFFS2 flash filesystem based on flash memory. For this, we first analyze the write procedure of JFFS2 filesystem in detail, and derive the drawback and overhead. Then, we design the flash write back routine for deferred I/O. We apply it to the Linux JFFS2 by implementing fflush and flash_writeback kernel thread. The designed flash write back routine can reduce average write latency when the kernel buffers are enough to get the users data

An Implementation and Performance Evaluation of a RAID System Based on Embedded Linux

In this article, we present, design, and implement a software and hardware for an embedded RAID system. The merits and drawbacks of our system are presented by performance evaluation. The proposed hardware system consists of three fibre channel controllers for the interface with fibre channel disks and hosts. Embedded Linux in which a RAID software is implemented is ported to the hardware. A SCSI target mode device driver and a target mode SCSI module are designed for that our RAID system is considered as a block device to a host computer. Linux Multi-device is used as RAID functions of this system. A data cache module is implemented for high performance and the interconnection between Linux Multi-device and the target mode SCSI module. The RAID 5 module of Multi-device is modified for improvement of read performance. The benchmark shows that the new RAID 5 module is superior to the original one in overall performance.

R-Barrier: Rapid Barrier for Software RAID Cache Using Hints from Journaling Filesystem

While adopting cache in software RAID brings performance benefit, it can cause data loss at power failure which results in the broken filesystem consistency. Though I/O Barrier can be used to remove the consistency issue, it sacrifices the write performance of software RAID. To mitigate this tradeoff, we suggest R-Barrier for software RAID. The basic idea of R-Barrier is to avoid the synchronization of the entire cache space. Instead, based on given hints from the filesystem layer, R-Barrier makes software RAID cache select one of the following methods when the cache writes a certain write request into the disk: (1)Strictly ordered destaging for the write requests affecting the filesystem consistency, (2)Reordered destaging for the rest of write requests. We show that R-Barrier guarantees same filesystem consistency with I/O Barrier while the performance degradation of R-Barrier is less than that of I/O Barrier.

A Prefetching and Memory Management Policy for Personal Solid State Drives

Traditional technologies that are used to improve the performance of hard disk drives show many negative cases if they are applied to solid state drives (SSD). Access time and block sequence in hard disk drives that consist of mechanical components are very important performance factors. Meanwhile, SSD provides superior random read performance that is not affected by block address sequence due to the characteristics of flash memory. Practically, it is recommended to disable prefetching if a SSD is installed in a personal computer. However, this paper presents a combinational method of a prefetching scheme and a memory management that consider the internal structure of SSD and the characteristics of NAND flash memory. It is important that SSD must concurrently operate multiple flash memory chips. The I/O unit size of NAND flash memory tends to increase and it exceeded the block size of operating systems. Hence, the proposed prefetching scheme performs in an operating unit of SSD. To complement a weak point of the prefetching scheme, the proposed memory management scheme adaptively evicts uselessly prefetched data to maximize the sum of cache hit rate and prefetch hit rate. We implemented the proposed schemes as a Linux kernel module and evaluated them using a commercial SSD. The schemes improved the I/O performance up to 26% in a given experiment.