fogcached-ros: Hybrid main memory KVS server

• Item type: Symposium(1)
• 公開日: 2021-03-15
• タイトル: fogcached-ros: Hybrid main memory KVS server
• タイトル / 言語: en
• タイトル: fogcached-ros: Hybrid main memory KVS server
• 言語: eng
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_5794
• 資源タイプ: conference paper
• 著者所属: Shibaura Institute of Technology
• 著者所属: VA Linux Systems Japan K K
• 著者所属: Tokai University
• 著者所属: Shibaura Institute of Technology
• 著者所属(英): en
• 著者所属(英): Shibaura Institute of Technology
• 著者所属(英): en
• 著者所属(英): VA Linux Systems Japan K K
• 著者所属(英): en
• 著者所属(英): Tokai University
• 著者所属(英): en
• 著者所属(英): Shibaura Institute of Technology
• 著者名: Koki, Higashi; Yoichi, Ishiwata; Takeshi, Ohkawa; Midori, Sugaya
• 著者名(英): Koki, Higashi; Yoichi, Ishiwata; Takeshi, Ohkawa; Midori, Sugaya
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: In recent years, to process a large amount of sensing data from IoT devices, it is expected to place an edge server located closer than the cloud to reduce the delay from the distance. Edge servers are required to have high responsiveness and reliability. Within the Edge researches, fogcached successfully reduces the response time because of introducing the KVS caches servers, and increase reliability uses not only DRAM but also Non-Volatile-Memory (Intel Optane Data Center Persistent Memory) to store the less access cache. It provides a dual-LRU mechanism that is extended by the memcached as a management for the hybrid memories. However, because fogcached was only evaluated by simulation, it is not clear the mechanism also effective for the practical applications. The purpose of this study is to evaluate the fogcached with the practical robotic application which is necessary to apply the edge computing. These days, robotics application is developed generally on the ROS, and SLAM (Simultaneous Localization and Mapping) application. To apply the fogcached to the ROS and SLAM, we designed an extension of ROS namely fogcached-ros, which executed on ROS to connect to fogcached. It stores and retrieves data in fogcached with API. In preliminary experiments, we thought SLAM as suitable for evaluation of edge computing. We evaluated the API overhead measurement results and found that it was less than 1ms. We concluded these API support many topics.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: In recent years, to process a large amount of sensing data from IoT devices, it is expected to place an edge server located closer than the cloud to reduce the delay from the distance. Edge servers are required to have high responsiveness and reliability. Within the Edge researches, fogcached successfully reduces the response time because of introducing the KVS caches servers, and increase reliability uses not only DRAM but also Non-Volatile-Memory (Intel Optane Data Center Persistent Memory) to store the less access cache. It provides a dual-LRU mechanism that is extended by the memcached as a management for the hybrid memories. However, because fogcached was only evaluated by simulation, it is not clear the mechanism also effective for the practical applications. The purpose of this study is to evaluate the fogcached with the practical robotic application which is necessary to apply the edge computing. These days, robotics application is developed generally on the ROS, and SLAM (Simultaneous Localization and Mapping) application. To apply the fogcached to the ROS and SLAM, we designed an extension of ROS namely fogcached-ros, which executed on ROS to connect to fogcached. It stores and retrieves data in fogcached with API. In preliminary experiments, we thought SLAM as suitable for evaluation of edge computing. We evaluated the API overhead measurement results and found that it was less than 1ms. We concluded these API support many topics.
• 書誌情報: Proceedings of Asia Pacific Conference on Robot IoT System Development and Platform; 巻 2020, p. 39-45, 発行日 2021-03-15
• 出版者 / 言語: ja
• 出版者: 情報処理学会