ocfs2.rst 4.7 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117
  1. .. SPDX-License-Identifier: GPL-2.0
  2. ================
  3. OCFS2 filesystem
  4. ================
  5. OCFS2 is a general purpose extent based shared disk cluster file
  6. system with many similarities to ext3. It supports 64 bit inode
  7. numbers, and has automatically extending metadata groups which may
  8. also make it attractive for non-clustered use.
  9. You'll want to install the ocfs2-tools package in order to at least
  10. get "mount.ocfs2" and "ocfs2_hb_ctl".
  11. Project web page: http://ocfs2.wiki.kernel.org
  12. Tools git tree: https://github.com/markfasheh/ocfs2-tools
  13. OCFS2 mailing lists: https://oss.oracle.com/projects/ocfs2/mailman/
  14. All code copyright 2005 Oracle except when otherwise noted.
  15. Credits
  16. =======
  17. Lots of code taken from ext3 and other projects.
  18. Authors in alphabetical order:
  19. - Joel Becker <joel.becker@oracle.com>
  20. - Zach Brown <zach.brown@oracle.com>
  21. - Mark Fasheh <mfasheh@suse.com>
  22. - Kurt Hackel <kurt.hackel@oracle.com>
  23. - Tao Ma <tao.ma@oracle.com>
  24. - Sunil Mushran <sunil.mushran@oracle.com>
  25. - Manish Singh <manish.singh@oracle.com>
  26. - Tiger Yang <tiger.yang@oracle.com>
  27. Caveats
  28. =======
  29. Features which OCFS2 does not support yet:
  30. - Directory change notification (F_NOTIFY)
  31. - Distributed Caching (F_SETLEASE/F_GETLEASE/break_lease)
  32. Mount options
  33. =============
  34. OCFS2 supports the following mount options:
  35. (*) == default
  36. ======================= ========================================================
  37. barrier=1 This enables/disables barriers. barrier=0 disables it,
  38. barrier=1 enables it.
  39. errors=remount-ro(*) Remount the filesystem read-only on an error.
  40. errors=panic Panic and halt the machine if an error occurs.
  41. intr (*) Allow signals to interrupt cluster operations.
  42. nointr Do not allow signals to interrupt cluster
  43. operations.
  44. noatime Do not update access time.
  45. relatime(*) Update atime if the previous atime is older than
  46. mtime or ctime
  47. strictatime Always update atime, but the minimum update interval
  48. is specified by atime_quantum.
  49. atime_quantum=60(*) OCFS2 will not update atime unless this number
  50. of seconds has passed since the last update.
  51. Set to zero to always update atime. This option need
  52. work with strictatime.
  53. data=ordered (*) All data are forced directly out to the main file
  54. system prior to its metadata being committed to the
  55. journal.
  56. data=writeback Data ordering is not preserved, data may be written
  57. into the main file system after its metadata has been
  58. committed to the journal.
  59. preferred_slot=0(*) During mount, try to use this filesystem slot first. If
  60. it is in use by another node, the first empty one found
  61. will be chosen. Invalid values will be ignored.
  62. commit=nrsec (*) Ocfs2 can be told to sync all its data and metadata
  63. every 'nrsec' seconds. The default value is 5 seconds.
  64. This means that if you lose your power, you will lose
  65. as much as the latest 5 seconds of work (your
  66. filesystem will not be damaged though, thanks to the
  67. journaling). This default value (or any low value)
  68. will hurt performance, but it's good for data-safety.
  69. Setting it to 0 will have the same effect as leaving
  70. it at the default (5 seconds).
  71. Setting it to very large values will improve
  72. performance.
  73. localalloc=8(*) Allows custom localalloc size in MB. If the value is too
  74. large, the fs will silently revert it to the default.
  75. localflocks This disables cluster aware flock.
  76. inode64 Indicates that Ocfs2 is allowed to create inodes at
  77. any location in the filesystem, including those which
  78. will result in inode numbers occupying more than 32
  79. bits of significance.
  80. user_xattr (*) Enables Extended User Attributes.
  81. nouser_xattr Disables Extended User Attributes.
  82. acl Enables POSIX Access Control Lists support.
  83. noacl (*) Disables POSIX Access Control Lists support.
  84. resv_level=2 (*) Set how aggressive allocation reservations will be.
  85. Valid values are between 0 (reservations off) to 8
  86. (maximum space for reservations).
  87. dir_resv_level= (*) By default, directory reservations will scale with file
  88. reservations - users should rarely need to change this
  89. value. If allocation reservations are turned off, this
  90. option will have no effect.
  91. coherency=full (*) Disallow concurrent O_DIRECT writes, cluster inode
  92. lock will be taken to force other nodes drop cache,
  93. therefore full cluster coherency is guaranteed even
  94. for O_DIRECT writes.
  95. coherency=buffered Allow concurrent O_DIRECT writes without EX lock among
  96. nodes, which gains high performance at risk of getting
  97. stale data on other nodes.
  98. journal_async_commit Commit block can be written to disk without waiting
  99. for descriptor blocks. If enabled older kernels cannot
  100. mount the device. This will enable 'journal_checksum'
  101. internally.
  102. ======================= ========================================================