On May 20th, openEuler SDS Meetup Beijing Station is coming. This Meetup is jointly sponsored by openEuler sig-SDS & XSKY XSKY , aiming to provide a platform for developers and users to share and exchange distributed storage technology. This event invites experts and developers in the field of openEuler SDS to share their Experiences and cases in the practical application of distributed storage technology, show the latest progress of openEuler SDS and discuss the future development direction, all are welcome to sign up.
- Time: May 20, 2023 (Saturday) 14:30-17:30
- Location: Building C4, Building 7, Yard 3, Fengxiu Road, Haidian District, Beijing - XSKY
Quickly scan the QR code below to sign up for the event poster, sign in and enter the venue with your participation voucher, and see you in Beijing on May 20!
Highlights
Several automated tuning schemes based on Ceph
Teng Lei
Kylin Tuning Expert
Introduction to the topic: This report will introduce a ceph optimization solution based on automatic tuning tools. In the report, we will explain how to use the A-tune automatic tuning tool in ceph tuning practice. In addition, we will also Explain in detail the comparison of the advantages and disadvantages of the intelligent algorithms involved in A-tune, as well as the performance comparison when applied to ceph automatic tuning.
Use of Ceph in Cloud Native Scenarios
Yang Xiaoliang
Tongxin Senior R&D Engineer
Introduction to the topic: This topic mainly explains the various deployment methods of ceph clusters in the cloud-native environment, sorting out and solving common problems, and analyzing actual cases. Combined with actual cloud-native scenarios, a reasonable and reliable ceph deployment combination is recommended.
Common problems and troubleshooting methods in Ceph storage system
Chi Xinze
R&D Director of XSKY Basic Platform
Introduction to the topic: 1. Introduce the common performance positioning methods of Ceph and the main measures of optimization; 2. Common problems that may be triggered in the production environment and the repair plan; 3. The processing method of storage failure caused by hardware failure.
aNOF: An Extremely Low Latency Solution
Liu Qinfei
openEuler sig-SDS maintainer
Introduction to the topic: In the OLTP field, especially in the securities high-frequency trading market (HFT), there is an extremely low latency (ULL) requirement for data writing. The persistence of data written in the data center requires that after the data is written to the primary node, a copy is written to the remote node (fault domain). Therefore, the overall latency of data writing depends on the single-point write latency. We propose an extremely low-latency cross-network writing scheme.
surprise benefits
In addition to a series of technical dry goods, the openEuler community also brings you sincere offline interactive prizes. Developers who participate in the Meetup interactive session can draw prizes on the spot, and lucky developers will be randomly selected to give out surprise gifts!