Supports the HDFS protocol, allowing components within the Hadoop ecosystem to access and use file systems.
Supports custom export paths for NFS protocol file systems.
Supports adding compute resources to file controllers.
Supports the file management network and the file access network sharing the same VM network, with their IP addresses in the same network segment.
Supports displaying active client connections of file systems.
Provides performance monitoring charts for file storage clusters and file systems.
Supports configuring whether Windows clients can access file storage clusters via the NFSv3 protocol on AOC.
Supports automatic restart of file controllers if their system disks enter read-only mode.
Supports prefetching files.
Sets the metadata disk redundancy policy for the newly created file system to 3 replicas, which improves metadata performance and enhances metadata security.
Reduces I/O pause duration during file controller online/offline operations and high-availability (HA) failovers.
Optimizes the cluster upgrade process to eliminate the need to restart file controllers, further reducing I/O pause duration during upgrades.
Refines the mechanism for adding the file controller's file management IP to the AOC allowlist.
Optimizes the storage engine's block allocation mechanism to improve block reclamation efficiency.
The file management virtual IP (VIP) failed to migrate automatically when the file management network was abnormal. The issue has been resolved by improving the VIP high-availability mechanism.
Leaked socket file descriptors in the NTIRPC component caused NFS requests to be unprocessed. The issue has been resolved in this release.
The NFS backend returned unexpected non-retryable error codes under specific conditions. The issue has been resolved in this release.
The sfs-agent restarted sfs-meta, sfs-data, or sfs-ganesha during a coredump, which interrupted the process. The issue has been resolved in this release.
