1. 1 File Systems Chapter 6 6.1 Files 6.2 Directories 6.3 File system implementation 6.4 Example file systems

2. 2 Long-term Information Storage 1.Must store large amounts of data 2.Information stored must survive the termination of the process using it 3.Multiple processes must be able to access the information concurrently. In short:

3. 3 Long-term Information Storage Files: Good! No Files: Bad!

4. 4 File Naming Typical file extensions.

5. 5 File Structure Three kinds of files –byte sequence –record sequence –tree

6. 6 File Types: Text and Binary (a) An executable file (b) An archive

7. 7 File Access Sequential access –read all bytes/records from the beginning –cannot jump around, could rewind or back up –convenient when medium was mag tape Random access –bytes/records read in any order –essential for data base systems –read can be … move file marker (seek), then read or … read and then move file marker

8. 8 File Attributes Possible file attributes

9. 9 Example Assume all counters are currently 0. Consider the case when pages 0,2,4, and 5 are referenced between last interrupt.

10. 10 File Operations 1.Create 2.Delete 3.Open 4.Close 5.Read 6.Write 7.Append 8.Seek 9.Get attributes 10.Set Attributes 11.Rename

11. 11 -rwxr-xr-x 1 dickens spcprj 20580000 Nov 16 2003 BM.Contention.Big -rw-r--r-- 1 dickens spcprj 5832 Nov 14 2003 LR.Contention.Low -rwxr-xr-x 1 dickens spcprj 5180000 Nov 14 2003 bitmap.contention.1 -rwxr-xr-x 1 dickens spcprj 12434 Nov 14 2003 Companion.4 -rw-r--r-- 1 dickens spcprj 2767 Nov 14 2003 temp.Swap -rw-r--r-- 1 dickens spcp 2767 Nov 14 2003 temp -rwxr-xr-x 1 dickens spc 16969 Nov 14 2003 ind_calc -rw-r--r-- 1 dickens spcprj 1217 Nov 14 2003 Temp_File.contention

12. 12 An Example Program Using File System Calls (1/2)

13. 13 An Example Program Using File System Calls (2/2)

14. 14 Memory Mapped Files Use system calls like: –map (filename, starting address, size); –unmap (filename, starting address, size) ; Implemented through paging mechanism. Advantages of this: ?

15. 15 Directories Single-Level Directory Systems A single level directory system –contains 4 files –owned by 3 different people, A, B, and C

16. 16 Two-level Directory Systems Letters indicate owners of the directories and files

17. 17 Hierarchical Directory Systems A hierarchical directory system

18. 18 A UNIX directory tree Path Names

19. 19 To Open dict path is: /usr/jim/dict. Path Names

20. 20 Directory Operations 1.Create 2.Delete 3.Opendir 4.Closedir 5.Readdir 6.Rename 7.Link 8.Unlink

21. 21 File System Implementation A possible file system layout

22. 22 Implementing Files (1) (a) Contiguous allocation of disk space for 7 files (b) State of the disk after files D and E have been removed

23. 23 Implementing Files (2) Storing a file as a linked list of disk blocks

24. 24 Implementing Files (3) Linked list allocation using a file allocation table in RAM

25. 25 Entry 4 bytes. Blocks 1K. 20 Million Entries == 80 MB for table.

26. 26 Implementing Files (4) An example i-node

27. 27 An example i-node Double Indirection Triple Indirection Disk block containing addresses of disk blocks containing addresses

28. 28 Implementing Directories (1) (a) A simple directory fixed size entries disk addresses and attributes in directory entry (b) Directory in which each entry just refers to an i-node

29. 29 Implementing Directories (2) Two ways of handling long file names in directory –(a) In-line –(b) In a heap

30. 30 Shared Files (1) File system containing a shared file

31. 31 Shared Files (1) Cyclic Family Tree

32. 32 Links (a) Situation prior to linking (b) After the link is created (c) After the original owner removes the file

33. 33 Symbolic Links Provide the path name of the target file in the linked file

34. 34 Disk Space Management (1) Dark line (left hand scale) gives data rate of a disk Dotted line (right hand scale) gives disk space efficiency All files 2KB Block size

35. 35 Disk Space Management (2) (a) Storing the free list on a linked list (b) A bit map

36. 36 Quotas for keeping track of each user’s disk use

37. 37 Consistency Issues File system can become inconsistent if there is a system crash and recent changes have not all been written to disk. Consider: –inode cached –New block allocated to file –Block filled, written to disk –Free list updated and written to disk –System crash –Now block not on either free list or listed as part of a file

38. 38 Consider: –inode cached –Free list cached –Block 10 deleted from file A. Cached inode updated. –Block 10 now on free list and allocated to file B. –File B closed, inode written to disk. –System crash. –Now block 10 listed as belonging to two files.

39. 39 File System Consistency –Build two tables, each of which counts number of blocks. –Table 1: Number of times block is in a file. –Table 2: Number of times block is in the free list. –Read all inodes. –Increment tables.

40. 40 File System Reliability (3) File system states (a) consistent (b) missing block (c) duplicate block in free list (d) duplicate data block

41. 41 –Figure b- missing block. Adds it back to the free list. –Figure c- block listed twice in free list Rebuilds the free list. –Figure d- Same block belongs to two or more files. Make copy and insert in one of the files.

42. 42 File System Performance (1) The block cache data structures

43. 43 Problem with LRU If frequently used block is inode or directory block, may want to write it out. Modify LRU to account for importance of block. Unix: synch() MS-DOS: Write-through cache

44. 44 Most systems write critical blocks to disk immediately (but not necessarily non-critical blocks). If system crashes, file system likely intact (but users may become unglued). Unix: System call sync that causes all modified blocks to be written to disk. –An update daemon is running in the background and alternates between sleeping and calling synch. Generally every 30 seconds. MS-DOS: Write-through cache. Every change is written through to disk.

45. 45 File System Performance (2) I-nodes placed at the start of the disk Disk divided into cylinder groups –each with its own blocks and i-nodes

46. 46 CP/M 8080 Chip Max 64K RAM 720K Floppy. Size of entire OS: 3584 bytes. Shell: 2K bytes.

47. 47 The CP/M File System (1) Memory layout of CP/M

48. 48 Directory One directory Entries 32 bytes (fixed). After booting, reads in directory and computes free list. Does not save free list. Provides 38 systems calls, 17 File system related. File blocks are 1K.

49. 49 The CP/M File System (2) The CP/M directory entry format

50. 50 MS-DOS Added hierarchical directory structure. Use fixed 32-byte directory entry Added attributes: read-only, archived, hidden, system. Time field: seconds (5 bits), minutes (6 bits), hours (5 bits). Date: day (5 bits), month (4 bits), year (7 bits– only good to 2107).

51. 51 The MS-DOS File System (1) The MS-DOS directory entry

52. 52 File Allocation Table (FAT) Entry 4 bytes. Blocks 1K. 20 Million Entries == 80 MB for table.

53. 53 The MS-DOS File System (2) Maximum partition for different block sizes The empty boxes represent forbidden combinations

54. 54 Windows 98 Used 32-bit FAT. Used the additional 10 bits from DOS entry. Allowed arbitrarily long file names Backward compatible to MS-DOS.

55. 55 The Windows 98 File System (1) The extended MOS-DOS directory entry used in Windows 98 Bytes

56. 56 The Windows 98 File System (2) An entry for (part of) a long file name in Windows 98 Bytes Checksum

57. 57 The Windows 98 File System (3) An example of how a long name is stored in Windows 98

58. 58 The UNIX V7 File System (1) A UNIX V7 directory entry

59. 59 The UNIX V7 File System (1) A UNIX V7 directory entry

60. 60 The UNIX V7 File System (2) A UNIX i-node

61. 61 The UNIX V7 File System (3) The steps in looking up /usr/ast/mbox