• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.4
• /
• pp.17-22
• /
• 2022

With the spread of web services, data types are becoming more diversified. In addition to the form of storing data such as images, videos, and texts, the number and form of properties and metadata expressing the data are different for each data. In order to efficiently process such unstructured data, a key-value store is widely used for state-of-the-art applications. LSM-Tree (Log Structured Merge Tree) is the core data structure of various commercial key-value stores. LSM-Tree is optimized to provide high performance for small writes by recording all write and delete operations in a log manner. However, there is a problem in that the delay time and processing speed of user requests are lowered as batches of deletion operations for expired data are inserted into the LSM-Tree as special key-value data. This paper presents a Filtered LSM-Tree (FLSM-Tree) that solves the above problem by separating the deleted key from the main tree structure while maintaining all the advantages of the existing LSM-Tree. The proposed method is implemented in LevelDB, a commercial key-value store and it shows that the read performance is improved by up to 47% in performance evaluation.

• Mycobiology
• /
• v.43 no.2
• /
• pp.131-136
• /
• 2015

The basidiomycete Laetiporus sulphureus var. miniatus belongs to the Aphyllophorales, Polyporaceae, and grows on the needleleaf tree. The fruiting bodies of Laetiporus species are known to produce N-methylated tyramine derivatives, polysaccharides, and various lanostane triterpenoids. As part of our ongoing effort to discover biologically active compounds from wood-rotting fungi, an anti-inflammatory triterpene, LSM-H7, has been isolated from the fruiting body of L. sulphureus var. miniatus and identified as acetyl eburicoic acid. LSM-H7 dose-dependently inhibited the NO production in RAW 264.7 cells without any cytotoxicity at the tested concentrations. Furthermore it suppressed the production of proinflammatory cytokines, mainly inducible nitric oxide synthase, cyclooxygenase-2, interleukin (IL)-$1{\beta}$, IL-6 and tumor necrosis factor ${\alpha}$, when compared with glyceraldehyde 3-phosphate dehydrogenase. These data suggest that LSM-H7 is a crucial component for the anti-inflammatory activity of L. sulphureus var. miniatus.

• Annual Conference of KIPS
• /
• 2017.11a
• /
• pp.709-712
• /
• 2017

RocksDB는 Facebook에서 LevelDB를 기반으로 개발한 임베디드 key-value 스토리지 엔진이다. Log structured tree(LSM-tree)를 기본구조로 사용하는 RocksDB는 데이터를 레벨단위로 저장한다. 지속적인 데이터 입력으로 인하여 레벨의 크기를 초과하게 되면 하위 레벨의 SST 파일과 병합을 통해 하위레벨로 내려 보낸다. 이 과정에서 디바이스의 부가적인 쓰기가 발생한다. 본 논문에서는 RocksDB의 디스크영역에 있는 SST 파일의 크기가 디바이스의 쓰기 증폭에 미치는 영향을 분석하였다. SST 파일크기변화에 따른 디바이스의 쓰기 증폭과 성능변화를 측정하고 비교하였다. 실험결과를 통해 SST의 크기가 작을수록 쓰기 증폭이 줄었지만 디바이스의 쓰기와 읽기 성능이 감소하는 것을 확인하였다. 결과적으로 쓰기 증폭을 줄이고 성능을 최대화 하기 위해서는 이 둘의 트레이드오프를 파악하고 분석하여 시스템에 맞는 최적의 SST 파일 크기를 찾아야한다.

• Annual Conference of KIPS
• /
• 2020.11a
• /
• pp.69-72
• /
• 2020

Log Structured Merged Tree(LSM-Tree)구조를 사용하여 빠른 데이터 쓰기 성능을 보유한 RocksDB에는 쓰기 증폭과 공간 증폭 현상이 발생한다. 쓰기 증폭은 과도한 쓰기 연산을 유발하여 데이터 처리 성능 저하와 플래시 메모리 기반 장치의 수명 저하를 초래하며, 공간 증폭은 데이터 저장 공간 점유로 인한 저장 공간 부족 문제를 야기한다. 본 논문에서는 쓰기 증폭과 공간 증폭 완화를 위해 RocksDB 의 성능에 영향 주는 주요 파라미터를 추출하고, 기계학습 기법인 랜덤 포레스트를 사용하여 추출한 파라미터가 쓰기 증폭과 공간 증폭에 미치는 영향을 분석하였다. 실험결과 쓰기 증폭과 공간 증폭에 영향을 많이 주는 주요 요소를 선별하였고 다른 파라미터에 대비해서 성능 격차가 61.7% 더 나타낸 것을 발견하였다.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.3
• /
• pp.111-116
• /
• 2023

The WiscKey database, which optimizes overhead by compaction of the LSM tree-based Key-Value database, stores the value in a separate file, and stores only the key and value addresses in the database. Each time an fsync system call function is used to ensure data integrity in the process of storing values. In previously conducted studies, workload performance was reduced by up to 5.8 times as a result of performing the workload without calling the fsync system call function. However, it is difficult to ensure the data integrity of the database without using the fsync system call function. In this paper, to reduce the overhead of the fsync system call function while performing workloads on the WiscKey database, we use NVDIMM caching techniques to ensure data integrity while improving the performance of the WiscKey database.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.5
• /
• pp.107-112
• /
• 2020

Recently, unstructured data is rapidly being produced based on web-based services. NoSQL systems and key value stores that process unstructured data as key and value pairs are widely used in various applications. In this paper, a study was conducted on a skip list used for in-memory data management in an LSM-tree based key value store. The skip list used in the key value store is an insertion-based skip list that does not allow overwriting and processes all changes only by inserting. This behavior can support Multi-Version Concurrency Control (MVCC), which can simultaneously process multiple read/write requests through snapshot isolation. However, since duplicate keys exist in the skip list, the performance significantly degrades due to unnecessary node visits during a list traverse. In particular, serious overhead occurs when a range query or scan operation that collectively searches a specific range of data occurs. This paper proposes a newly designed Stride SkipList to reduce this overhead. The stride skip list additionally maintains an indexing pointer for the last node of the same key to avoid unnecessary node visits. The proposed scheme is implemented using RocksDB's in-memory component, and the performance evaluation shows that the performance of SCAN operation improves by up to 350 times compared to the existing skip list for various workloads.

• Journal of KIISE:Information Networking
• /
• v.30 no.3
• /
• pp.407-415
• /
• 2003

As services requesting data transfer from a sender to a multiple number of receivers become popular, efficient group communication mechanisms like multicast get much attention. Since multicast is more efficient than unicast in terms of bandwidth usage and group management for group communication, many multicast protocols providing scalability and reliability have been proposed. Recently, router-supported reliable multicast protocols have been proposed because routers have the knowledge of the physical multicast tree structure and, in this scheme, repliers which retransmit lost packets are selected by routers. Repliers are selected dynamically based on the network situation, therefore, any receiver within a multicast group can become a replier Hence, all receivers within a group maintains a buffer for loss recovery within which received packets are stored. It is an overhead for all group receivers to store unnecessary packets. Therefore, in this paper, we propose a new scheme which reduces resource usage by discarding packets unnecessary for loss recovery from the receiver buffer. Our scheme performs the replier selection and the loss recovery of lost packets based on the LSM [1] model, and discards unnecessary packets determined by ACKs from erasers which represent local groups.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.3
• /
• pp.803-809
• /
• 2023

This paper analyzes and evaluates to optimally build a data collection system essential for factory energy management systems on an edge-based lightweight platform. A "Sensor/OPC-UA simulator" was developed based on sensors in an actual food factory and used to evaluate the storage engine of edge devices. The performance of storage engines in edge devices was evaluated to suggest the optimal storage engine. The experimental results show that when using the RocksDB storage engine, it has less than half the memory and database size compared to using InnoDB, and has a 3.01 times faster processing time. This study enables the selection of advantageous storage engines for managing time-series data on devices with limited resources and contributes to further research in this field through the sensor/OPC simulator.

• Journal of Platform Technology
• /
• v.11 no.5
• /
• pp.49-58
• /
• 2023

Factory energy management system is rapidly growing and evolving due to factors such as the 3rd Basic Energy Plan and global energy cost increases, as well as environmental issues. However, implementing an essential data collection system for energy management in factory settings, which have limited space and unique characteristics, presents spatial, environmental, and energy-related challenges. This paper endeavors to mitigate these challenges by devising a data collection system implemented through an edge-based lightweight platform. A comparison and evaluation of database operation on edge devices are conducted. To conduct the evaluation, a benchmarking tool called CDI Benchmark is developed, utilizing the characteristics of existing factories involved in practical applications. The evaluation results revealed that RDBMS systems like MySQL encountered errors in the database due to high data insertion loads, making them inoperable. On the other hand, InfluxDB, thanks to its highly efficient compression algorithm, demonstrated compression rates about 6 times higher than MyRocks according to the evaluation. However, it was observed that MyRocks outperformed InfluxDB by a significant margin, recording a maximum processing time approximately 80 times faster compared to InfluxDB.

• The Journal of the Korea Contents Association
• /
• v.21 no.7
• /
• pp.127-133
• /
• 2021

In industrial IoT environments, sensors generate data for their detection targets and deliver the data to IoT gateways. Therefore, managing large amounts of real-time sensor data is an essential feature for IoT gateways, and key-value storage engines are widely used to manage these sensor data. However, key-value storage engines used in IoT gateways do not take into account the characteristics of sensor data generated in industrial IoT environments, and this limits the performance of key-value storage engines. In this paper, we optimize the key-value storage engine by utilizing the features of sensor data in industrial IoT environments. The proposed optimization technique is to analyze the key, which is the input of a key-value storage engine, for further indexing. This reduces excessive write amplification and improves performance. We implement our optimization scheme in LevelDB and use the workload of the TPCx-IoT benchmark to evaluate our proposed scheme. From experimental results we show that our proposed technique achieves up to 21 times better than the existing scheme, and this shows that the proposed technique can perform high-speed data ingestion in industrial IoT environments.