The release of the Lustre 2.17 clustered file system has been published, which is used in the majority of the largest Linux clusters containing tens of thousands of nodes. The key components of Lustre are the Metadata Servers (MDS), Management Servers (MGS), Object Storage Servers (OSS), Object Storage Targets (OST, supports operation on ext4 and ZFS), and clients. The project code is distributed under the GPLv2 license.
Key innovations:
- A hybrid I/O performance optimization mechanism is enabled by default, which automatically applies buffered or direct I/O mode depending on the nature of the workload and the size of the transmitted data.
- Support for dynamic configuration of the Nodemap engine, responsible for mapping user identifiers (UID/GID) in 'multi-tenancy' configurations, where different client groups have access to the file system, has been added.
- In the dynamic configuration system for LNet NID (LNet Network Identifier), correct handling of registration attempts for more than 32 NID identifiers has been implemented (previously, such situations led to failures, now NIDs above the limit will be ignored). server for the kernel package from the RHEL 9.7 series (5.14.0-611.13.1.el9_7), and clients for unmodified kernels of RHEL 9.7 (5.14.0-611.13.1.el9_7), RHEL 10.1 (6.12.0-124.20.1.el10_1), SLES15 SP7 (6.4.0-150700.53.22.1), and Ubuntu 24.04 (6.8.0-35).
- Support ensured server The Curl project has eliminated the use of the strcpy function in its code.
Source: opennet.ru
