0001 .. SPDX-License-Identifier: GPL-2.0
0002
0003 ================
0004 OCFS2 filesystem
0005 ================
0006
0007 OCFS2 is a general purpose extent based shared disk cluster file
0008 system with many similarities to ext3. It supports 64 bit inode
0009 numbers, and has automatically extending metadata groups which may
0010 also make it attractive for non-clustered use.
0011
0012 You'll want to install the ocfs2-tools package in order to at least
0013 get "mount.ocfs2" and "ocfs2_hb_ctl".
0014
0015 Project web page: http://ocfs2.wiki.kernel.org
0016 Tools git tree: https://github.com/markfasheh/ocfs2-tools
0017 OCFS2 mailing lists: https://oss.oracle.com/projects/ocfs2/mailman/
0018
0019 All code copyright 2005 Oracle except when otherwise noted.
0020
0021 Credits
0022 =======
0023
0024 Lots of code taken from ext3 and other projects.
0025
0026 Authors in alphabetical order:
0027
0028 - Joel Becker <joel.becker@oracle.com>
0029 - Zach Brown <zach.brown@oracle.com>
0030 - Mark Fasheh <mfasheh@suse.com>
0031 - Kurt Hackel <kurt.hackel@oracle.com>
0032 - Tao Ma <tao.ma@oracle.com>
0033 - Sunil Mushran <sunil.mushran@oracle.com>
0034 - Manish Singh <manish.singh@oracle.com>
0035 - Tiger Yang <tiger.yang@oracle.com>
0036
0037 Caveats
0038 =======
0039 Features which OCFS2 does not support yet:
0040
0041 - Directory change notification (F_NOTIFY)
0042 - Distributed Caching (F_SETLEASE/F_GETLEASE/break_lease)
0043
0044 Mount options
0045 =============
0046
0047 OCFS2 supports the following mount options:
0048
0049 (*) == default
0050
0051 ======================= ========================================================
0052 barrier=1 This enables/disables barriers. barrier=0 disables it,
0053 barrier=1 enables it.
0054 errors=remount-ro(*) Remount the filesystem read-only on an error.
0055 errors=panic Panic and halt the machine if an error occurs.
0056 intr (*) Allow signals to interrupt cluster operations.
0057 nointr Do not allow signals to interrupt cluster
0058 operations.
0059 noatime Do not update access time.
0060 relatime(*) Update atime if the previous atime is older than
0061 mtime or ctime
0062 strictatime Always update atime, but the minimum update interval
0063 is specified by atime_quantum.
0064 atime_quantum=60(*) OCFS2 will not update atime unless this number
0065 of seconds has passed since the last update.
0066 Set to zero to always update atime. This option need
0067 work with strictatime.
0068 data=ordered (*) All data are forced directly out to the main file
0069 system prior to its metadata being committed to the
0070 journal.
0071 data=writeback Data ordering is not preserved, data may be written
0072 into the main file system after its metadata has been
0073 committed to the journal.
0074 preferred_slot=0(*) During mount, try to use this filesystem slot first. If
0075 it is in use by another node, the first empty one found
0076 will be chosen. Invalid values will be ignored.
0077 commit=nrsec (*) Ocfs2 can be told to sync all its data and metadata
0078 every 'nrsec' seconds. The default value is 5 seconds.
0079 This means that if you lose your power, you will lose
0080 as much as the latest 5 seconds of work (your
0081 filesystem will not be damaged though, thanks to the
0082 journaling). This default value (or any low value)
0083 will hurt performance, but it's good for data-safety.
0084 Setting it to 0 will have the same effect as leaving
0085 it at the default (5 seconds).
0086 Setting it to very large values will improve
0087 performance.
0088 localalloc=8(*) Allows custom localalloc size in MB. If the value is too
0089 large, the fs will silently revert it to the default.
0090 localflocks This disables cluster aware flock.
0091 inode64 Indicates that Ocfs2 is allowed to create inodes at
0092 any location in the filesystem, including those which
0093 will result in inode numbers occupying more than 32
0094 bits of significance.
0095 user_xattr (*) Enables Extended User Attributes.
0096 nouser_xattr Disables Extended User Attributes.
0097 acl Enables POSIX Access Control Lists support.
0098 noacl (*) Disables POSIX Access Control Lists support.
0099 resv_level=2 (*) Set how aggressive allocation reservations will be.
0100 Valid values are between 0 (reservations off) to 8
0101 (maximum space for reservations).
0102 dir_resv_level= (*) By default, directory reservations will scale with file
0103 reservations - users should rarely need to change this
0104 value. If allocation reservations are turned off, this
0105 option will have no effect.
0106 coherency=full (*) Disallow concurrent O_DIRECT writes, cluster inode
0107 lock will be taken to force other nodes drop cache,
0108 therefore full cluster coherency is guaranteed even
0109 for O_DIRECT writes.
0110 coherency=buffered Allow concurrent O_DIRECT writes without EX lock among
0111 nodes, which gains high performance at risk of getting
0112 stale data on other nodes.
0113 journal_async_commit Commit block can be written to disk without waiting
0114 for descriptor blocks. If enabled older kernels cannot
0115 mount the device. This will enable 'journal_checksum'
0116 internally.
0117 ======================= ========================================================
