Quote: Cedar files designed for simplicity and performance; only from file servers, immutable, atomic updates, send & cache whole files [»giffDK3_1988]

Quote: in Cedar, all remote files are immutable; replaces cache consistency problem with version control [»giffDK3_1988]

Quote: Cedar atomically adds a file to a file server (transfer, new version id, and directory update); implemented with mutual exclusion [»giffDK3_1988]

Quote: Cedar can retrieve a file with a specified creation time; searches all versions [»giffDK3_1988]

Quote: with Cedar, can attach local names to remote files that are cached on disk; allows a large set of locally named files [»giffDK3_1988]

Quote: Modeler uses immutable objects with unique names; allows name to be used in place of text [»lampBW6_1983]

Quote: Google uses BigFiles with 64-bit offsets, multiple file systems, compression, and allocation/deallocation [»brinS4_1998]

Quote: a GFS file consists of fixed-size chunks identified by a globally unique 64-bit handle; a chunk is a Linux file [»gherS10_2003]

Quote: V reads and writes blocks without a file position; idempotent transactions allows kernel to ignore reply messages [»cherDR4_1984]

Quote: ZFS provides a transactional store of objects with 2^64 sets of 2^64 objects, each 2^64 bytes; not designed for 32-bit architectures [»browD9_2007]

Quote: a backing store server manages persistent objects with internal names; allocates and frees storage space [»birrAD9_1980]

Quote: a file server stores and retrieves bulk data for a local area network [»birrAD9_1980]

Quote: a filing system is a client of the backing store server; applications have access to either [»birrAD9_1980]

Quote: objects in a backing store server are either publicly accessible segments or private indices [»birrAD9_1980]

Quote: the Elephant file system saves all important versions; easy undo of recent changes [»santDS12_1999]

Quote: the Elephant file system uses Keep-One for unimportant and easily recreated files; like standard file system [»santDS12_1999]

Quote: Keep-Safe guarantees that a file version is retained for an interval; deleted after a younger version or permanent deletion [»santDS12_1999]

Quote: versioned file system does not affect write disk caching; little data rewritten to versioned files [»santDS12_1999]

Quote: file system implemented in Multics [»daleRC_1965]

Quote: UNIX file system described by a replicated, immutable superblock [»mckuMK8_1984]

Quote: UNIX uses at least 4K file blocks; allows 4 gigabyte files with two levels of indirection [»mckuMK8_1984]

Quote: description of new file system for UNIX 4.2bsd [»mckuMK8_1984]

Quote: a Unix file exists independently of its directory entries; deleted when last entry is deleted [»ritcDM7_1978a]

Quote: Unix files accessed by their i-number which indexes their i-node in the devices i-list [»ritcDM7_1978a]

Quote: except for Inames of mounted volumes, Inames are only preserved in indices of the same volume [»birrAD9_1980]

Quote: the Unix i-list gives each file a short, unique name; is reliable and simple, with easy access to file information [»ritcDM7_1978a]

Quote: forked, Unix process share the same open file table entry; multiple opens of a file share the same i-node table entry [»thomK7_1978]

Quote: disk fragmentation slows down the BSD Fast File System by 5-15%; segment cleaner overhead slows down a log-structured system by 33% [»seltM1_1995]

Quote: the BSD Fast File System takes up to six distinct I/O operations to create a new one-kilobyte file [»seltM1_1995]

Quote: the BSD Fast File System does not attempt to preserve large free extents

Quote: Multics treats files as segments in virtual memory; avoids file copies and overlay management, needs 'attach' mechanism [»dennPJ_1980]

Quote: an Oberon writer initially creates an anonymous file; extension of memory since the disk space is reclaimed on system boot [»wirtN9_1989]

Quote: Pilot's managers implement file management and complex virtual memory operations

Quote: Pilot's kernel manages virtual memory access to special files and to preexisting, fixed-size normal files [»redeDD2_1980]

Quote: Amoeba files are replicated on two disks; the create-file operation can return when 0, 1 or 2 copies written to disk [»vanrR10_1988]

Quote: Amoeba's bullet server keeps the i-node table in memory; records the file's cache address if any [»taneAS12_1990]

Quote: Amoeba's bullet file server reads 2-3 times faster than SUN NFS; large write bandwidth is 4 times higher; small file create is slower [»vanrR10_1988]

Quote: Amoeba file reads are 2-3 times faster than Sun NFS; large file writes are faster; capability overhead is constant [»taneAS12_1990]

Quote: Amoeba has separate file and directory servers; one handles storage, the other naming and protection; works well [»taneAS12_1990]

Quote: Amoeba's file server performs garbage collection by touching every file every k days and deleting a file if not accessed in n days (n>>k) [»taneAS12_1990]

Quote: a file's capability includes its i-node index; Amoeba's memory-resident i- node table gives the disk and cache address; at most 1 disk seek [»vanrR10_1988]

Quote: Inferno applications use a fixed set of files for devices and system services [»dorwSM1_1997]

Quote: in Inferno and Plan 9, all resources have a hierarchical file name in a forest of file systems and in a private name space [»dorwSM1_1997]

Quote: represent device drivers with a data file and a control file