NFS-GANESHA is a user-space NFSv2, NFSv3, and NFSv4 server. It runs on Linux, BSD variants, and POSIX-compliant Unixes.
|Operating Systems||POSIX Linux Unix|
Release Notes: FSALs (filesystem backends) are now loadable shared objects. The server can support multiple backends at runtime. NFSv4.1 pNFS is supported. DBus is now the administration tool. All the significant bugfixes from the 1.5.x branch have been forward ported or re-implemented in the 2.x codebase. The server passes all of the cthonv4 and pynfs 4.0 tests. Most subsystems/modules have had refactor/rewrite work. NFSv2 support has been deprecated. The build process has been converted to Cmake. The codebase has been reformatted to conform to the Linux kernel coding style.
Release Notes: The TIRPC library was significantly refurbished. Multiple RPC dispatch is now supported. Cache inode readdir logic is now based on callbacks. State Management has been unified and improved. Client ID management support has been added. 9p.2000L support has been added. File handle support was improved. Memory allocation has been refactored. TC-Malloc or Jemalloc can be use with a fallback to libc's malloc/free. Kerberos support was improved. Compatibility with Microsoft AD Kerberos is improved. FSAL_SNMP is now officially deprecated.
Release Notes: A memory leak has been fixed in RPC's DRC.Bad computation of rbt_value in RPC's DRC's hashtable has been fixed. New FSAL_CEPH (dedicated to the ceph filesystem). pNFS code has been refurbished with CEPH-specific code and layout file support. All RPCs are now based on TIRPC. In nfs_Write, for clean handling of umask, the server should allow the owner of a file to write to it, even if the file is read-only. Bad memory padding in fsal_handle_t and fsal_cookie_t have been fixed.
Release Notes: This release adds Lustre FSAL, exporting several Lustre filesystems with the same server instance. A bug was found (during Bake-A-Thon) and fixed in OP4_READDIRPLUS (leading to a missing directory's entries). The directory content cache is now managed as a tree. Formerly used dir_chain were removed. Lock support has been added for FSAL_LUSTRE. For FSAL_LUSTRE, FSAL_XFS, FSAL_FUSELIKE, and FSAL_ZFS, a regression in readdir has been detected and fixed. The export list now supports IP ranges using the CIDR format.
Release Notes: This release will call cache_inode_close after deleting share state. It adds support for ERR_FSAL_DELAY in the upper layer. It has a new SAL (State Abstraction Layer) to manage locks (NFSv4/NLM) and share reservations. Symbolic links are managed via a pool to save memory in Cache_Inode. Early 9P support has been added. Lustre FSAL supports exporting several Lustre filesystems with the same server instance.