9fa168b01a
Rather then combining hardlink creation for the geom(8) binary with shared library build, move libraries to src/lib/geom so they are built and installed normally. Create a common Makefile.classes which is included by both lib/geom/Makefile and sbin/geom/Makefile so the symlink and libraries stay in sync. The relocation of libraries allows libraries to be build for 32-bit compat. This also reduces the number of non-standard builds in the system. This commit is not sufficent to run a 32-bit /sbin/geom on a 64-bit system out of the box as it will look in the wrong place for libraries unless GEOM_LIBRARY_PATH is set appropriatly in the environment. Reviewed by: bdrewery Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D15360
347 lines
9.7 KiB
Groff
347 lines
9.7 KiB
Groff
.\" Copyright (c) 2006-2009 Pawel Jakub Dawidek <pjd@FreeBSD.org>
|
|
.\" All rights reserved.
|
|
.\"
|
|
.\" Redistribution and use in source and binary forms, with or without
|
|
.\" modification, are permitted provided that the following conditions
|
|
.\" are met:
|
|
.\" 1. Redistributions of source code must retain the above copyright
|
|
.\" notice, this list of conditions and the following disclaimer.
|
|
.\" 2. Redistributions in binary form must reproduce the above copyright
|
|
.\" notice, this list of conditions and the following disclaimer in the
|
|
.\" documentation and/or other materials provided with the distribution.
|
|
.\"
|
|
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHORS AND CONTRIBUTORS ``AS IS'' AND
|
|
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE LIABLE
|
|
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
.\" SUCH DAMAGE.
|
|
.\"
|
|
.\" $FreeBSD$
|
|
.\"
|
|
.Dd February 17, 2009
|
|
.Dt GJOURNAL 8
|
|
.Os
|
|
.Sh NAME
|
|
.Nm gjournal
|
|
.Nd "control utility for journaled devices"
|
|
.Sh SYNOPSIS
|
|
.Nm
|
|
.Cm label
|
|
.Op Fl cfhv
|
|
.Op Fl s Ar jsize
|
|
.Ar dataprov
|
|
.Op Ar jprov
|
|
.Nm
|
|
.Cm stop
|
|
.Op Fl fv
|
|
.Ar name ...
|
|
.Nm
|
|
.Cm sync
|
|
.Op Fl v
|
|
.Nm
|
|
.Cm clear
|
|
.Op Fl v
|
|
.Ar prov ...
|
|
.Nm
|
|
.Cm dump
|
|
.Ar prov ...
|
|
.Nm
|
|
.Cm list
|
|
.Nm
|
|
.Cm status
|
|
.Nm
|
|
.Cm load
|
|
.Nm
|
|
.Cm unload
|
|
.Sh DESCRIPTION
|
|
The
|
|
.Nm
|
|
utility is used for journal configuration on the given GEOM provider.
|
|
The Journal and data may be stored on the same provider or on two separate
|
|
providers.
|
|
This is block level journaling, not file system level journaling, which means
|
|
everything gets logged, e.g.\& for file systems, it journals both data and
|
|
metadata.
|
|
The
|
|
.Nm
|
|
GEOM class can talk to file systems, which allows the use of
|
|
.Nm
|
|
for file system journaling and to keep file systems in a consistent state.
|
|
At this time, only UFS file system is supported.
|
|
.Pp
|
|
To configure journaling on the UFS file system using
|
|
.Nm ,
|
|
one should first create a
|
|
.Nm
|
|
provider using the
|
|
.Nm
|
|
utility, then run
|
|
.Xr newfs 8
|
|
or
|
|
.Xr tunefs 8
|
|
on it with the
|
|
.Fl J
|
|
flag which instructs UFS to cooperate with the
|
|
.Nm
|
|
provider below.
|
|
There are important differences in how journaled UFS works.
|
|
The most important one is that
|
|
.Xr sync 2
|
|
and
|
|
.Xr fsync 2
|
|
system calls do not work as expected anymore.
|
|
To ensure that data is stored on the data provider, the
|
|
.Nm Cm sync
|
|
command should be used after calling
|
|
.Xr sync 2 .
|
|
For the best performance possible, soft-updates should be disabled when
|
|
.Nm
|
|
is used.
|
|
It is also safe and recommended to use the
|
|
.Cm async
|
|
.Xr mount 8
|
|
option.
|
|
.Pp
|
|
When
|
|
.Nm
|
|
is configured on top of
|
|
.Xr gmirror 8
|
|
or
|
|
.Xr graid3 8
|
|
providers, it also keeps them in a consistent state, thus
|
|
automatic synchronization on power failure or system crash may be disabled
|
|
on those providers.
|
|
.Pp
|
|
The
|
|
.Nm
|
|
utility uses on-disk metadata, stored in the provider's last sector,
|
|
to store all needed information.
|
|
This could be a problem when an existing file system is converted to use
|
|
.Nm .
|
|
.Pp
|
|
The first argument to
|
|
.Nm
|
|
indicates an action to be performed:
|
|
.Bl -tag -width ".Cm status"
|
|
.It Cm label
|
|
Configures
|
|
.Nm
|
|
on the given provider(s).
|
|
If only one provider is given, both data and journal are stored on the same
|
|
provider.
|
|
If two providers are given, the first one will be used as data provider and the
|
|
second will be used as the journal provider.
|
|
.Pp
|
|
Additional options include:
|
|
.Bl -tag -width ".Fl s Ar jsize"
|
|
.It Fl c
|
|
Checksum journal records.
|
|
.It Fl f
|
|
May be used to convert an existing file system to use
|
|
.Nm ,
|
|
but only if the journal will be configured on a separate provider and if the
|
|
last sector in the data provider is not used by the existing file system.
|
|
If
|
|
.Nm
|
|
detects that the last sector is used, it will refuse to overwrite it
|
|
and return an error.
|
|
This behavior may be forced by using the
|
|
.Fl f
|
|
flag, which will force
|
|
.Nm
|
|
to overwrite the last sector.
|
|
.It Fl h
|
|
Hardcode provider names in metadata.
|
|
.It Fl s Ar jsize
|
|
Specifies size of the journal if only one provider is used for both data and
|
|
journal.
|
|
The default is one gigabyte.
|
|
Size should be chosen based on provider's load, and not on its size;
|
|
recommended minimum is twice the size of the physical memory installed.
|
|
It is not recommended to use
|
|
.Nm
|
|
for small file systems (e.g.: only few gigabytes big).
|
|
.El
|
|
.It Cm clear
|
|
Clear metadata on the given providers.
|
|
.It Cm stop
|
|
Stop the given provider.
|
|
.Pp
|
|
Additional options include:
|
|
.Bl -tag -width ".Fl f"
|
|
.It Fl f
|
|
Stop the given provider even if it is opened.
|
|
.El
|
|
.It Cm sync
|
|
Trigger journal switch and enforce sending data to the data provider.
|
|
.It Cm dump
|
|
Dump metadata stored on the given providers.
|
|
.It Cm list
|
|
See
|
|
.Xr geom 8 .
|
|
.It Cm status
|
|
See
|
|
.Xr geom 8 .
|
|
.It Cm load
|
|
See
|
|
.Xr geom 8 .
|
|
.It Cm unload
|
|
See
|
|
.Xr geom 8 .
|
|
.El
|
|
.Pp
|
|
Additional options include:
|
|
.Bl -tag -width ".Fl v"
|
|
.It Fl v
|
|
Be more verbose.
|
|
.El
|
|
.Sh EXIT STATUS
|
|
Exit status is 0 on success, and 1 if the command fails.
|
|
.Sh EXAMPLES
|
|
Create a
|
|
.Nm
|
|
based UFS file system and mount it:
|
|
.Bd -literal -offset indent
|
|
gjournal load
|
|
gjournal label da0
|
|
newfs -J /dev/da0.journal
|
|
mount -o async /dev/da0.journal /mnt
|
|
.Ed
|
|
.Pp
|
|
Configure journaling on an existing file system, but only if
|
|
.Nm
|
|
allows this (i.e., if the last sector is not already used by the file system):
|
|
.Bd -literal -offset indent
|
|
umount /dev/da0s1d
|
|
gjournal label da0s1d da0s1e && \e
|
|
tunefs -J enable -n disable da0s1d.journal && \e
|
|
mount -o async /dev/da0s1d.journal /mnt || \e
|
|
mount /dev/da0s1d /mnt
|
|
.Ed
|
|
.Sh SYSCTLS
|
|
Gjournal adds the sysctl level kern.geom.journal.
|
|
The string and integer information available is detailed below.
|
|
The changeable column shows whether a process with appropriate privilege may
|
|
change the value.
|
|
.Bl -column "accept_immediatelyXXXXXX" integerXXX -offset indent
|
|
.It Sy "sysctl name Type Changeable"
|
|
.It "debug integer yes"
|
|
.It "switch_time integer yes"
|
|
.It "force_switch integer yes"
|
|
.It "parallel_flushes integer yes"
|
|
.It "accept_immediately integer yes"
|
|
.It "parallel_copies integer yes"
|
|
.It "record_entries integer yes"
|
|
.It "optimize integer yes"
|
|
.El
|
|
.Bl -tag -width 6n
|
|
.It Li debug
|
|
Setting a non-zero value enables debugging at various levels.
|
|
Debug level 1 will record actions at a journal level, relating to journal
|
|
switches, metadata updates, etc.
|
|
Debug level 2 will record actions at a higher level, relating to the numbers of
|
|
entries in journals, access requests, etc.
|
|
Debug level 3 will record verbose detail, including insertion of I/Os to the
|
|
journal.
|
|
.It Li switch_time
|
|
The maximum number of seconds a journal is allowed to remain open before
|
|
switching to a new journal.
|
|
.It Li force_switch
|
|
Force a journal switch when the journal uses more than N% of the free journal
|
|
space.
|
|
.It Li parallel_flushes
|
|
The number of flush I/O requests to be sent in parallel when flushing the
|
|
journal to the data provider.
|
|
.It Li accept_immediately
|
|
The maximum number of I/O requests accepted at the same time.
|
|
.It Li parallel_copies
|
|
The number of copy I/O requests to send in parallel.
|
|
.It Li record_entries
|
|
The maximum number of record entries to allow in a single journal.
|
|
.It Li optimize
|
|
Controls whether entries in a journal will be optimized by combining overlapping
|
|
I/Os into a single I/O and reordering the entries in a journal.
|
|
This can be disabled by setting the sysctl to 0.
|
|
.El
|
|
.Ss cache
|
|
The string and integer information available for the cache level
|
|
is detailed below.
|
|
The changeable column shows whether a process with appropriate
|
|
privilege may change the value.
|
|
.Bl -column "alloc_failuresXXXXXX" integerXXX -offset indent
|
|
.It Sy "sysctl name Type Changeable"
|
|
.It "used integer no"
|
|
.It "limit integer yes"
|
|
.It "divisor integer no"
|
|
.It "switch integer yes"
|
|
.It "misses integer yes"
|
|
.It "alloc_failures integer yes"
|
|
.El
|
|
.Bl -tag -width 6n
|
|
.It Li used
|
|
The number of bytes currently allocated to the cache.
|
|
.It Li limit
|
|
The maximum number of bytes to be allocated to the cache.
|
|
.It Li divisor
|
|
Sets the cache size to be used as a proportion of kmem_size.
|
|
A value of 2 (the default) will cause the cache size to be set to 1/2 of the
|
|
kmem_size.
|
|
.It Li switch
|
|
Force a journal switch when this percentage of cache has been used.
|
|
.It Li misses
|
|
The number of cache misses, when data has been read, but was not found in the
|
|
cache.
|
|
.It Li alloc_failures
|
|
The number of times memory failed to be allocated to the cache because the cache
|
|
limit was hit.
|
|
.El
|
|
.Ss stats
|
|
The string and integer information available for the statistics level
|
|
is detailed below.
|
|
The changeable column shows whether a process with appropriate
|
|
privilege may change the value.
|
|
.Bl -column "skipped_bytesXXXXXX" integerXXX -offset indent
|
|
.It Sy "sysctl name Type Changeable"
|
|
.It "skipped_bytes integer yes"
|
|
.It "combined_ios integer yes"
|
|
.It "switches integer yes"
|
|
.It "wait_for_copy integer yes"
|
|
.It "journal_full integer yes"
|
|
.It "low_mem integer yes"
|
|
.El
|
|
.Bl -tag -width 6n
|
|
.It Li skipped_bytes
|
|
The number of bytes skipped.
|
|
.It Li combined_ios
|
|
The number of I/Os which were combined by journal optimization.
|
|
.It Li switches
|
|
The number of journal switches.
|
|
.It Li wait_for_copy
|
|
The number of times the journal switch process had to wait for the previous
|
|
journal copy to complete.
|
|
.It Li journal_full
|
|
The number of times the journal was almost full, forcing a journal switch.
|
|
.It Li low_mem
|
|
The number of times the low_mem hook was called.
|
|
.El
|
|
.Sh SEE ALSO
|
|
.Xr geom 4 ,
|
|
.Xr geom 8 ,
|
|
.Xr mount 8 ,
|
|
.Xr newfs 8 ,
|
|
.Xr tunefs 8 ,
|
|
.Xr umount 8
|
|
.Sh HISTORY
|
|
The
|
|
.Nm
|
|
utility appeared in
|
|
.Fx 7.0 .
|
|
.Sh AUTHORS
|
|
.An Pawel Jakub Dawidek Aq Mt pjd@FreeBSD.org
|