[manpages] Rewrite man pages in a simpler text format.

Makefile still needs work, but I wanted to commit to save progress.
This commit is contained in:
Joe Thornber 2017-09-20 11:15:00 +01:00
parent e48516914d
commit dd8848f088
41 changed files with 1182 additions and 1173 deletions

4
.gitignore vendored
View File

@ -1,4 +1,5 @@
*~ *~
*.tmp
*.swp *.swp
*.swo *.swo
*.o *.o
@ -7,12 +8,15 @@
*.gmo *.gmo
*_t *_t
*.d *.d
*.8
*.d.[0-9]* *.d.[0-9]*
test.data test.data
cachegrind.* cachegrind.*
\#*\# \#*\#
core core
googletest/
bin/pdata_tools bin/pdata_tools
thin_check thin_check
thin_dump thin_dump

View File

@ -193,7 +193,7 @@ else
TEST_INCLUDES= TEST_INCLUDES=
endif endif
.SUFFIXES: .d .SUFFIXES: .d .txt .8
%.o: %.cc %.o: %.cc
@echo " [CXX] $<" @echo " [CXX] $<"
@ -203,6 +203,10 @@ endif
sed 's,\([^ :]*\)\.o[ :]*,\1.o \1.gmo $* : Makefile ,g' < $*.$$$$ > $*.d; \ sed 's,\([^ :]*\)\.o[ :]*,\1.o \1.gmo $* : Makefile ,g' < $*.$$$$ > $*.d; \
$(RM) $*.$$$$ $(RM) $*.$$$$
%.8: %.txt bin/txt2man
@echo " [txt2man] $<"
$(V) bin/txt2man -p -t $(basename $(notdir $<)) $< > $@
#---------------------------------------------------------------- #----------------------------------------------------------------
lib/libpdata.a: $(OBJECTS) $(EMITTERS) lib/libpdata.a: $(OBJECTS) $(EMITTERS)
@ -232,7 +236,36 @@ clean:
distclean: clean distclean: clean
$(RM) config.cache config.log config.status configure.h version.h Makefile unit-tests/Makefile $(RM) config.cache config.log config.status configure.h version.h Makefile unit-tests/Makefile
install: bin/pdata_tools TOOLS:=\
cache_check \
cache_dump \
cache_metadata_size \
cache_repair \
cache_restore \
cache_writeback \
thin_check \
thin_delta \
thin_dump \
thin_ls \
thin_repair \
thin_restore \
thin_rmap \
thin_metadata_size \
thin_trim \
era_check \
era_dump \
era_invalidate \
era_restore \
ifeq ("@DEVTOOLS", "yes")
thin_show_duplicates \
thin_ll_dump \
thin_generate_metadata \
thin_scan
endif
MANPAGES:=$(patsubst %,man8/%.8,$(TOOLS))
install: bin/pdata_tools $(MANPAGES)
$(INSTALL_DIR) $(BINDIR) $(INSTALL_DIR) $(BINDIR)
$(INSTALL_PROGRAM) bin/pdata_tools $(BINDIR) $(INSTALL_PROGRAM) bin/pdata_tools $(BINDIR)
$(STRIP) $(BINDIR)/pdata_tools $(STRIP) $(BINDIR)/pdata_tools

365
bin/txt2man Executable file
View File

@ -0,0 +1,365 @@
#!/bin/sh
test "$HOME" = ~ || exec ksh $0 "$@" # try ksh if sh too old (not yet POSIX)
# Copyright (C) 2001, 2002, 2003 Marc Vertes
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2, or (at your option)
# any later version.
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
# 02111-1307, USA.
# txt2man-1.5
usage()
{
cat << EOT
NAME
txt2man - convert flat ASCII text to man page format
SYNOPSIS
txt2man [-hpTX] [-t mytitle] [-P pname] [-r rel] [-s sect]
[-v vol] [-I txt] [-B txt] [ifile]
DESCRIPTION
txt2man converts the input text into nroff/troff standard man(7)
macros used to format Unix manual pages. Nice pages can be generated
specially for commands (section 1 or 8) or for C functions reference
(sections 2, 3), with the ability to recognize and format command and
function names, flags, types and arguments.
txt2man is also able to recognize and format sections, paragraphs,
lists (standard, numbered, description, nested), cross references and
literal display blocks.
If input file ifile is omitted, standard input is used. Result is
displayed on standard output.
Here is how text patterns are recognized and processed:
Sections These headers are defined by a line in upper case, starting
column 1. If there is one or more leading spaces, a
sub-section will be generated instead.
Paragraphs They must be separated by a blank line, and left aligned.
Tag list The item definition is separated from the item description
by at least 2 blank spaces, even before a new line, if
definition is too long. Definition will be emphasized
by default.
Bullet list
Bullet list items are defined by the first word being "-"
or "*" or "o".
Enumerated list
The first word must be a number followed by a dot.
Literal display blocks
This paragraph type is used to display unmodified text,
for example source code. It must be separated by a blank
line, and be indented. It is primarily used to format
unmodified source code. It will be printed using fixed font
whenever possible (troff).
Cross references
A cross reference (another man page) is defined by a word
followed by a number in parenthesis.
Special sections:
NAME The function or command name and short description are set in
this section.
SYNOPSIS This section receives a special treatment to identify command
name, flags and arguments, and propagate corresponding
attributes later in the text. If a C like function is recognized
(word immediately followed by an open parenthesis), txt2man will
print function name in bold font, types in normal font, and
variables in italic font. The whole section will be printed using
a fixed font family (courier) whenever possible (troff).
It is a good practice to embed documentation into source code, by using
comments or constant text variables. txt2man allows to do that, keeping
the document source readable, usable even without further formatting
(i.e. for online help) and easy to write. The result is high quality
and standard complying document.
OPTIONS
-h The option -h displays help.
-P pname Set pname as project name in header. Default to uname -s.
-p Probe title, section name and volume.
-t mytitle Set mytitle as title of generated man page.
-r rel Set rel as project name and release.
-s sect Set sect as section in heading, ususally a value from 1 to 8.
-v vol Set vol as volume name, i.e. "Unix user 's manual".
-I txt Italicize txt in output. Can be specified more than once.
-B txt Emphasize (bold) txt in output. Can be specified more than once.
-T Text result previewing using PAGER, usually more(1).
-X X11 result previewing using gxditview(1).
ENVIRONMENT
PAGER name of paging command, usually more(1), or less(1). If not set
falls back to more(1).
EXAMPLE
Try this command to format this text itself:
$ txt2man -h 2>&1 | txt2man -T
HINTS
To obtain an overall good formating of output document, keep paragraphs
indented correctly. If you have unwanted bold sections, search for
multiple spaces between words, which are used to identify a tag list
(term followed by a description). Choose also carefully the name of
command line or function parameters, as they will be emphasized each
time they are encountered in the document.
SEE ALSO
man(1), mandoc(7), rman(1), groff(1), more(1), gxditview(1), troff(1).
BUGS
- Automatic probe (-p option) works only if input is a regular file (i.e.
not stdin).
AUTHOR
Marc Vertes <mvertes@free.fr>
EOT
}
sys=$(uname -s)
rel=
volume=
section=
title=untitled
doprobe=
itxt=
btxt=
post=cat
while getopts :hpTXr:s:t:v:P:I:B: opt
do
case $opt in
r) rel=$OPTARG;;
t) title=$OPTARG;;
s) section=$OPTARG;;
v) volume=$OPTARG;;
P) sys=$OPTARG;;
p) doprobe=1;;
I) itxt="$OPTARG§$itxt";;
B) btxt=$OPTARG;;
T) post="groff -mandoc -Tlatin1 | ${PAGER:-more}";;
X) post="groff -mandoc -X";;
*) usage; exit;;
esac
done
shift $(($OPTIND - 1))
if test "$doprobe"
then
title=${1##*/}; title=${title%.txt}
section="8"
volume="System Manager's Manual"
# get release from path
#rel=$(pwd | sed 's:/.*[^0-9]/::g; s:/.*::g')
rel="Device Mapper Tools"
fi
head=".\\\" Text automatically generated by txt2man
.TH $title $section \"$rel\" \"$volume\""
# All tabs converted to spaces
expand $* |
# gawk is needed because use of non standard regexp
gawk --re-interval -v head="$head" -v itxt="$itxt" -v btxt="$btxt" '
BEGIN {
print head
avar[1] = btxt; avar[2] = itxt
for (k in avar) {
mark = (k == 1) ? "\\fB" : "\\fI"
split(avar[k], tt, "§")
for (i in tt)
if (tt[i] != "")
subwords["\\<" tt[i] "\\>"] = mark tt[i] "\\fP"
for (i in tt)
delete tt[i]
}
for (k in avar)
delete avar[k]
}
{
# to avoid some side effects in regexp
sub(/\.\.\./, "\\.\\.\\.")
# remove spaces in empty lines
sub(/^ +$/,"")
}
/^[[:upper:][:space:]]+$/ {
# Section header
if ((in_bd + 0) == 1) {
in_bd = 0
print ".fam T\n.fi"
}
if (section == "SYNOPSIS") {
print ".fam T\n.fi"
type["SYNOPSIS"] = ""
}
if ($0 ~/^[^[:space:]]/)
print ".SH " $0
else
print ".SS" $0
sub(/^ +/, "")
section = $0
if (section == "SYNOPSIS")
print ".nf\n.fam C"
ls = 0 # line start index
pls = 0 # previous line start index
pnzls = 0 # previous non zero line start index
ni = 0 # indent level
ind[0] = 0 # indent offset table
prevblankline = 0
next
}
{
# Compute line start index, handle start of example display block
pls = ls
if (ls != 0)
pnzls = ls
match($0, /[^ ]/)
ls = RSTART
if (pls == 0 && pnzls > 0 && ls > pnzls && $1 !~ /^[0-9\-\*\o]\.*$/) {
# example display block
if (prevblankline == 1) {
print ".PP"
prevblankline = 0
}
print ".nf\n.fam C"
in_bd = 1
eoff = ls
}
if (ls > 0 && ind[0] == 0)
ind[0] = ls
}
(in_bd + 0) == 1 {
# In example display block
if (ls != 0 && ls < eoff) {
# End of litteral display block
in_bd = 0
print ".fam T\n.fi"
} else { print; next }
}
section == "NAME" {
$1 = "\\fB" $1
sub(/ \- /, " \\fP- ")
}
section == "SYNOPSIS" {
# Identify arguments of fcts and cmds
if (type["SYNOPSIS"] == "") {
if (index($0, "(") == 0 && index($0, ")") == 0 &&
index($0, "#include") == 0)
type["SYNOPSIS"] = "cmd"
else
type["SYNOPSIS"] = "fct"
}
if (type["SYNOPSIS"] == "cmd") {
# Line is a command line
if ($1 !~ /^\[/) {
b = $1
sub(/^\*/, "", b)
subwords["\\<" b "\\>"] = "\\fB" b "\\fP"
}
for (i = 2; i <= NF; i++) {
a = $i
gsub(/[\[\]\|]/, "", a)
if (a ~ /^[^\-]/)
subwords["\\<" a "\\>"] = "\\fI" a "\\fP"
}
} else {
# Line is a C function definition
if ($1 == "typedef")
subwords["\\<" $2 "\\>"] = "\\fI" $2 "\\fP"
else if ($1 == "#define")
subwords["\\<" $2 "\\>"] = "\\fI" $2 "\\fP"
for (i = 1; i <= NF; i++) {
if ($i ~ /[\,\)]/) {
a = $i
sub(/.*\(/, "", a)
gsub(/\W/, "", a)
if (a !~ /^void$/)
subwords["\\<" a "\\>"] = "\\fI" a "\\fP"
}
}
}
}
{
# protect dots inside words
while ($0 ~ /\w\.\w/)
sub(/\./, "_dOt_")
# identify func calls and cross refs
for (i = 1; i <= NF; i++) {
b = $i
sub(/^\*/, "", b)
if ((a = index(b, ")(")) > 3) {
w = substr(b, 3, a - 3)
subwords["\\<" w "\\>"] = "\\fI" w "\\fP"
}
if ((a = index(b, "(")) > 1) {
w = substr(b, 1, a - 1)
subwords["\\<" w "\\("] = "\\fB" w "\\fP("
}
}
# word attributes
for (i in subwords)
gsub(i, subwords[i])
# shell options
gsub(/\B\-+\w+(\-\w+)*/, "\\fB&\\fP")
# unprotect dots inside words
gsub(/_dOt_/, ".")
if (match($0, /[^ ] +/) > 0) {
# tag list item
adjust_indent()
tag = substr($0, 1, RSTART)
sub(/^ */, "", tag)
if (RSTART+RLENGTH < length())
$0 = substr($0, RSTART + RLENGTH)
else
$0 = ""
print ".TP\n.B"
print tag
prevblankline = 0
if (NF == 0)
next
} else if ($1 == "-"||$1 == "o"||$1 == "*") {
# bullet list item
adjust_indent()
print ".IP \\(bu 3"
prevblankline = 0
$1 = ""
} else if ($1 ~ /^[0-9]+[\).]$/) {
# enum list item
adjust_indent()
print ".IP " $1 " 4"
prevblankline = 0
$1 = ""
} else if (pls == 0) {
# new paragraph
adjust_indent()
} else if (NF == 0) {
# blank line
prevblankline = 1
next
} else
prevblankline = 0
# flush vertical space
if (prevblankline == 1) {
print ".PP"
prevblankline = 0
}
if (section != "SYNOPSIS" || $0 ~ /^ {1,4}/)
sub(/ */,"")
print
}
function adjust_indent()
{
if (ls > ind[ni]) {
ind[++ni] = ls
print ".RS"
} else if (ls < ind[ni]) {
while (ls < ind[ni]) {
ni--
print ".RE"
}
}
}
' | eval $post

View File

@ -1,68 +0,0 @@
.TH CACHE_CHECK 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
cache_check \- repair cache metadata on device or file
.SH SYNOPSIS
.B cache_check
.RB [ options ]
.I {device|file}
.SH DESCRIPTION
.B cache_check
checks cache metadata created by
the device-mapper cache target on a
.I device
or
.I file.
This tool cannot be run on live metadata.
.SH OPTIONS
.IP "\fB\-q, \-\-quiet\fP"
Suppress output messages, return only exit code.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.IP "\fB\-\-super\-block\-only\fP"
Only check the superblock is present.
.IP "\fB\-\-skip-mappings\fP"
Skip checking of the block mappings which make up the bulk of the
metadata.
.IP "\fB\-\-skip-hints\fP"
Skip checking of the policy hint values metadata.
.IP "\fB\-\-skip-discards\fP"
Skip checking of the discard bits in the metadata.
.IP "\fB\-\-clear\-needs\-check\-flag\fP"
The kernel may set a flag to force the pool to be checked before it's next
activated. Set this switch to clear the flag if the check is successful.
.SH EXAMPLE
Analyses and repairs cache metadata on logical volume
/dev/vg/metadata:
.sp
.B cache_check /dev/vg/metadata
The device may not be actively used by the target
when running.
.SH DIAGNOSTICS
.B cache_check
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B cache_dump(8)
.B cache_repair(8)
.B cache_restore(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

40
man8/cache_check.txt Normal file
View File

@ -0,0 +1,40 @@
NAME
cache_check - validates cache metadata on a device or file.
SYNOPSIS
cache_check [options] {device|file}
DESCRIPTION
cache_check checks cache metadata created by the device-mapper cache target
on a device or file.
This tool cannot be run on live metadata.
OPTIONS
-q, --quiet Suppress output messages, return only exit code.
-h, --help Print help and exit.
-V, --version Output version information and exit.
--super-block-only Only check the superblock.
--skip-hints Skip checking of the policy hint values metadata.
--skip-discards Skip checking of the discard bits in the metadata.
--clear-needs-check-flag Clears the 'needs_check' flag in the superblock.
The kernel may set a flag to force the pool to be checked before it's next
activated. Set this switch to clear the flag if the check is successful.
EXAMPLE
Analyses and repairs cache metadata on logical volume /dev/vg/metadata:
$ cache_check /dev/vg/metadata
The device may not be actively used by the target when running.
DIAGNOSTICS
cache_check returns an exit code of 0 for success or 1 for error.
SEE ALSO
cache_dump(8), cache_repair(8), cache_restore(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,64 +0,0 @@
.TH CACHE_DUMP 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
cache_dump \- dump cache metadata from device or file to standard output
.SH SYNOPSIS
.B cache_dump
.RB [options]
.I {device|file}
.SH DESCRIPTION
.B cache_dump
dumps binary cache metadata created by
the device-mapper cache target on a
.I device
or
.I file
to standard output for
analysis or postprocessing in XML format.
XML formated metadata can be fed into cache_restore (see
.BR cache_restore(8) )
in order to put it back onto a metadata
.I device
(to process by the device-mapper target) or
.I file.
This tool cannot be run on live metadata.
.IP "\fB\-r, \-\-repair\fP".
Repair the metadata whilst dumping it.
.IP "\fB\-h, \-\-help\fP".
Print help and exit.
.IP "\fB-o \<xml file\>\fP".
Specify an output file for the xml, rather than printing to stdout.
.IP "\fB\-V, \-\-version\fP".
Output version information and exit.
.SH EXAMPLES
Dumps the cache metadata on logical volume /dev/vg/metadata
to standard output in XML format:
.sp
.B cache_dump /dev/vg/metadata
.SH EXAMPLES
Dumps the cache metadata on logical volume /dev/vg/metadata
whilst repairing it to standard output in XML format:
.sp
.B cache_dump --repair /dev/vg/metadata
.SH DIAGNOSTICS
.B cache_dump
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B cache_check(8)
.B cache_repair(8)
.B cache_restore(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

41
man8/cache_dump.txt Normal file
View File

@ -0,0 +1,41 @@
NAME
cache_dump - dump cache metadata from device or file to standard output.
SYNOPSIS
cache_dump [options] {device|file}
DESCRIPTION
cache_dump dumps binary cache metadata created by the device-mapper cache
target on a device or file to standard output for analysis or postprocessing
in XML format. XML formatted metadata can be fed into cache_restore in order
to put it back onto a metadata device (to process by the device-mapper
target), or file.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-r, --repair Repair the metadata whilst dumping it.
-o {xml file> Specify an output file for the xml, rather than printing to stdout.
EXAMPLES
Dumps the cache metadata on logical volume /dev/vg/metadata to standard
output in XML format:
$ cache_dump /dev/vg/metadata
Dumps the cache metadata on logical volume /dev/vg/metadata whilst repairing
it to standard output in XML format: .sp
$ cache_dump --repair /dev/vg/metadata
DIAGNOSTICS
cache_dump returns an exit code of 0 for success or 1 for error.
SEE ALSO
cache_check(8), cache_repair(8), cache_restore(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,43 +0,0 @@
.TH CACHE_METADATA_SIZE 8 "Thin Provisioning Tools" "Red Hat, Inc."
.SH NAME
cache_metadata_size \- Estimate the size of the metadata device needed for a given configuration
.SH SYNOPSIS
.B cache_metadata_size
.RB [ options ]
.SH DESCRIPTION
.B cache_metadata_size
estimates the size of the metadata. Either \-\-nr\-blocks, or \-\-block\-size
and \-\-device\-size must be specified.
.SH OPTIONS
.IP "\fB\-\-nr\-blocks\fP \<natural\>"
Specify the number of cache blocks.
.IP "\fB\-\-block-size\fP \<sectors\>"
Specify the size of each cache block in 512 byte sectors.
.IP "\fB\-\-device\-size\fP \<sectors\>"
Specify total size of the fast device used in the cache. In 512 byte sectors.
.IP "\fB\-\-max\-hint\-width\fP \<nr bytes\>"
Cache policies use a per block 'hint' to record extra info (for instance hit
counts). At the moment all policies use a 4 byte hint width. If you want to use
a different hint width specify it with this switch.
.SH EXAMPLE
.B cache_metadata_size --nr-blocks 10240
.sp
.B cache_metadata_size --block-size 128 --device-size 1024000
.SH SEE ALSO
.B cache_dump(8)
.B cache_repair(8)
.B cache_restore(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <heinzm@redhat.com>

View File

@ -0,0 +1,31 @@
NAME
cache_metadata_size - Estimate the size of the metadata device needed for a given configuration.
SYNOPSIS
cache_metadata_size [options]
DESCRIPTION
cache_metadata_size estimates the size of the metadata. Either --nr-blocks,
or --block-size and --device-size must be specified.
OPTIONS
--nr-blocks {natural} Specify the number of cache blocks.
--block-size {sectors} Specify the size of each cache block in 512 byte sectors.
--device-size {sectors} Specify total size of the fast device used in the cache. In 512 byte sectors.
--max-hint-width {nr bytes} Specify hint width.
Cache policies use a per block 'hint' to record extra info (for instance
hit counts). At the moment all policies use a 4 byte hint
width. If you want to use a different hint width specify it with this
switch.
EXAMPLE
$ cache_metadata_size --nr-blocks 10240
$ cache_metadata_size --block-size 128 --device-size 1024000
SEE ALSO
cache_dump(8), cache_repair(8), cache_restore(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <heinzm@redhat.com>

View File

@ -1,64 +0,0 @@
.TH CACHE_REPAIR 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
cache_repair \- repair cache binary metadata from device/file to device/file
.SH SYNOPSIS
.B cache_repair
.RB [ options ]
.RB -i
.I {device|file}
.RB -o
.I {device|file}
.SH DESCRIPTION
.B cache_repair
reads binary cache metadata created by the
respective device-mapper target from one
.I device
or
.I file
, repairs it and writes it to another
.I device
or
.I file.
If written to a metadata
.I device
, the metadata can be processed
by the device-mapper target.
This tool cannot be run on live metadata.
.IP "\fB\-i, \-\-input\fP \fI{device|file}\fP"
Input file or device with binary metadata.
.IP "\fB\-o, \-\-output\fP \fI{device|file}\fP"
Output file or device for repaired binary metadata. If a file is used
then it must be preallocated, and large enough to hold the metadata.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH EXAMPLE
Reads the binary cache metadata from file
.B metadata
, repairs it and writes it to logical volume /dev/vg/metadata
for further processing by the respective device-mapper target:
.sp
.B cache_repair -i metadata -o /dev/vg/metadata
.SH DIAGNOSTICS
.B cache_repair
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B cache_dump(8)
.B cache_check(8)
.B cache_restore(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

38
man8/cache_repair.txt Normal file
View File

@ -0,0 +1,38 @@
NAME
cache_repair - repair cache binary metadata from device/file to device/file.
SYNOPSIS
cache_repair [options] -i {device|file} -o {device|file}
DESCRIPTION
cache_repair reads binary cache metadata created by the respective
device-mapper target from one device or file, repairs it and writes it to
another device or file. If written to a metadata device, the metadata can
be processed by the device-mapper target.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-i, --input {device|file} Input file or device containing binary metadata.
-o, --output {device|file} Output file or device for repaired binary metadata.
If a file is then it must be preallocated, and large enough to hold the
metadata.
EXAMPLE
Reads the binary cache metadata from file metadata, repairs it and writes it
to logical volume /dev/vg/metadata for further processing by the respective
device-mapper target:
$ cache_repair -i metadata -o /dev/vg/metadata
DIAGNOSTICS
cache_repair returns an exit code of 0 for success or 1 for error.
SEE ALSO
cache_dump(8), cache_check(8), cache_restore(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,78 +0,0 @@
.TH CACHE_RESTORE 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
cache_restore \- restore cache metadata file to device or file
.SH SYNOPSIS
.B cache_restore
.RB [ options ]
.RB -i
.I {device|file}
.RB -o
.I {device|file}
.SH DESCRIPTION
.B cache_restore
restores cache metadata created by the respective
device-mapper target dumped into an XML formatted (see
.BR cache_dump(8) )
.I file
, which optionally can be preprocessed before the restore to another
.I device
or
.I file.
If restored to a metadata
.I device
, the metadata can be processed by the device-mapper target.
This tool cannot be run on live metadata.
.IP "\fB\-i, \-\-input\fP \fI{device|file}\fP"
Input file or device with metadata.
.IP "\fB\-o, \-\-output\fP \fI{device|file}\fP"
Output file or device for repaired binary metadata. If a file is used
then it must be preallocated, and large enough to hold the metadata.
.IP "\fB\-\-metadata\-version\fP \fI{1|2}\fP"
Choose metadata version.
.IP "\fB\-q, \-\-quiet\fP"
Don't print any output. Check the exit code to test for success.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.IP "\fB\-\-debug-override-metadata-version\fP \fI<integer>\fP"
ONLY FOR DEBUGGING PURPOSES:
.br
override the version stored in the metadata.
.IP "\fB\-\-omit\-clean\-shutdown\fP"
ONLY FOR DEBUGGING PURPOSES:
.br
Don't set the clean shutdown flag.
.SH EXAMPLE
Restores the XML formatted cache metadata on file
.B metadata
to logical volume /dev/vg/metadata for further processing by the
respective device-mapper target:
.sp
.B cache_restore -i metadata -o /dev/vg/metadata
.SH DIAGNOSTICS
.B cache_restore
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B cache_dump(8)
.B cache_check(8)
.B cache_repair(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

47
man8/cache_restore.txt Normal file
View File

@ -0,0 +1,47 @@
NAME
cache_restore - restore cache metadata file to device or file.
SYNOPSIS
cache_restore [options] -i {xml file} -o {device|file}
DESCRIPTION
cache_restore restores cache metadata created by the respective device-mapper
target dumped into an XML formatted (see cache_dump(8)) file, which
optionally can be preprocessed before the restore to another device or file.
If restored to a metadata device, the metadata can be processed by the
device-mapper target.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-q, --quiet Don't print any output. Check the exit code to test for success.
-i, --input {xml file} Input xml.
-o, --output {device|file} Output file or device for restored binary metadata.
If a file is used thin it must be preallocated, and large enough to hold
the metadata.
--metadata-version {1|2} Choose a metadata version.
DEBUGGING OPTIONS
--debug-override-metadata-version {integer} Override the version stored in the metadata.
--omit-clean-shutdown Don't set the clean shutdown flag.
EXAMPLE
Restores the XML formatted cache metadata on file .B metadata to logical
volume /dev/vg/metadata for further processing by the respective
device-mapper target:
$ cache_restore -i metadata -o /dev/vg/metadata
DIAGNOSTICS
cache_restore returns an exit code of 0 for success or 1 for error.
SEE ALSO
cache_dump(8), cache_check(8), cache_repair(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,60 +0,0 @@
.TH CACHE_WRITEBACK 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
cache_writeback \- writeback dirty blocks to the origin device.
.SH SYNOPSIS
.B cache_writeback
.RB [ options ]
.RB --metadata-device
.I {device|file}
.RB --origin-device
.I {device|file}
.RB --fast-device
.I {device|file}
.SH DESCRIPTION
.B cache_writeback
An offline tool that writesback dirty data to the data device
(origin). Intended for use in recovery scenarios when the SSD is
giving IO errors.
This tool cannot be run on a live cache.
.SH OPTIONS
.IP "\fB\\-\-metadata\-device\fP \fI{device|file}\fP"
Location of cache metadata.
.IP "\fB\-\-origin\-device\fP \fI{device|file}\fP"
Slow device being cached.
.IP "\fB\-\-fast\-device\fP \fI{device|file}\fP"
Fast device containing the data that needs to be written back.
.IP "\fB\-\-no\-metadata\-update\fP"
Do not update the metadata to clear the dirty flags for written back
data. You may not want to do this if you're decommissioning the
cache.
.IP "\fB\-\-buffer\-size\-meg \fI{size}\fP"
Specify the size for the metadata cache. Defaults to 16 Gig, a larger size may
improve performance.
.IP "\fB\-\-list\-failed\-blocks\fP
List any blocks that failed the writeback process.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH SEE ALSO
.B cache_dump(8)
.B cache_check(8)
.B cache_repair(8)
.B cache_restore(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>

34
man8/cache_writeback.txt Normal file
View File

@ -0,0 +1,34 @@
NAME
cache_writeback - writeback dirty blocks to the origin device.
SYNOPSIS
cache_writeback [options] --metadata-device {device|file} --origin-device {device|file} --fast-device {device|file}
DESCRIPTION
cache_writeback is an offline tool that writes back dirty data to the data
device (origin). Intended for use in recovery scenarios when the SSD is
giving IO errors.
This tool cannot be run on a live cache.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
--metadata-device {device|file} Location of cache metadata.
--origin-device {device|file} Slow device being cached.
--fast-device {device|file} Fast device containing the data that needs to be written back.
--no-metadata-update Do not update the metadata to clear the dirty flags.
You may want to use this flag if you're decommissioning the cache.
--buffer-size-meg {size} Specify the size for the metadata cache.
Defaults to 16 Gig, a larger size may improve performance.
--list-failed-blocks List any blocks that failed the writeback process.
SEE ALSO
cache_dump(8), cache_check(8), cache_repair(8), cache_restore(8)
AUTHOR
Joe Thornber <ejt@redhat.com>

View File

@ -1,59 +0,0 @@
.TH ERA_CHECK 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
era_check \- validate era metadata on device or file
.SH SYNOPSIS
.B era_check
.RB [ options ]
.I {device|file}
.SH DESCRIPTION
.B era_check
checks era metadata created by
the device-mapper era target on a
.I device
or
.I file.
This tool cannot be run on live metadata.
.SH OPTIONS
.IP "\fB\-q, \-\-quiet\fP"
Suppress output messages, return only exit code.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.IP "\fB\-\-super\-block\-only\fP"
Only check the superblock is present.
.B era_check
will return a non-zero exit code if it finds a fatal
error. If any errors are discovered use
.B era_repair
to correct.
.SH EXAMPLE
Analyse thin provisioning metadata on logical volume
/dev/vg/metadata:
.sp
.B era_check /dev/vg/metadata
The device may not be actively used by the target
when running.
.SH DIAGNOSTICS
.B era_check
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B era_dump(8)
.B era_repair(8)
.B era_restore(8)
.B era_invalidate(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>

33
man8/era_check.txt Normal file
View File

@ -0,0 +1,33 @@
NAME
era_check - validate era metadata on device or file.
SYNOPSIS
era_check [options] {device|file}
DESCRIPTION
era_check checks era metadata created by the device-mapper era target on a
device or file.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-q, --quiet Suppress output messages, return only exit code.
--super-block-only Only check the superblock is present.
EXAMPLE
Analyse thin provisioning metadata on logical volume /dev/vg/metadata:
$ era_check /dev/vg/metadata
The device may not be actively used by the target when running.
DIAGNOSTICS
era_check returns an exit code of 0 for success or 1 for error.
SEE ALSO
era_dump(8), era_repair(8), era_restore(8), era_invalidate(8)
AUTHOR
Joe Thornber <ejt@redhat.com>

View File

@ -1,61 +0,0 @@
.TH ERA_DUMP 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
era_dump \- dump era metadata from device or file to standard output
.SH SYNOPSIS
.B era_dump
.RB [options]
.I {device|file}
.SH DESCRIPTION
.B era_dump
dumps binary era metadata created by the device-mapper
era target on a
.I device
or
.I file
to standard output for
analysis or postprocessing in XML format.
XML formated metadata can be fed into era_restore (see
.BR era_restore(8) )
in order to put it back onto a metadata
.I device
(to process by the device-mapper target) or
.I file.
This tool cannot be run on live metadata.
.IP "\fB\-\-repair\fP".
Repair the metadata whilst dumping it.
.IP "\fB\-h, \-\-help\fP".
Print help and exit.
.IP "\fB\-V, \-\-version\fP".
Output version information and exit.
.IP "\fB\-\-logical\fP".
Fold any unprocessed write sets into the final era array. You
probably want to do this if you're intending to process the results as
it simplifies the XML.
.IP "\fB-o \fI{xml file}\fP"
Specify a file for the output rather than writing to stdout.
.SH EXAMPLES
Dumps era metadata on logical volume /dev/vg/metadata
to standard output in XML format:
.sp
.B era_dump /dev/vg/metadata
.SH DIAGNOSTICS
.B era_dump
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B era_check(8)
.B era_repair(8)
.B era_restore(8)
.B era_invalidate(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>

40
man8/era_dump.txt Normal file
View File

@ -0,0 +1,40 @@
NAME
era_dump - dump era metadata from device or file to standard output.
SYNOPSIS
era_dump [options] {device|file}
DESCRIPTION
era_dump dumps binary era metadata created by the device-mapper era target on
a device or file to standard output for analysis or postprocessing in
XML format. XML formated metadata can be fed into era_restore (see
era_restore(8)) in order to put it back onto a metadata device (to process by
the device-mapper target) or file.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
--repair Repair the metadata whilst dumping it.
--logical Fold any unprocessed write sets into the final era array.
You probably want to do this if you're intending to process the results as
it simplifies the XML.
-o {xml file} Specify a file for the output rather than writeing to stdout.
EXAMPLES
Dumps era metadata on logical volume /dev/vg/metadata to standard output in
XML format:
$ era_dump /dev/vg/metadata
DIAGNOSTICS
era_dump returns an exit code of 0 for success or 1 for error.
SEE ALSO
era_check(8), era_repair(8), era_restore(8), era_invalidate(8)
AUTHOR
Joe Thornber <ejt@redhat.com>

View File

@ -1,54 +0,0 @@
.TH ERA_INVALIDATE 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
era_invalidate \- Provide a list of blocks that have changed since a particular era.
.SH SYNOPSIS
.B era_invalidate
.RB [ options ]
.I {device|file}
.SH DESCRIPTION
.B era_invalidate
Examines era metadata and lists blocks that may have changed since a given era.
This tool cannot be run on live metadata unless the \fB\-\-metadata\-snap\fP option is used.
.SH OPTIONS
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.IP "\fB\-o <output file>\fP"
Write output to a file rather than
.B stdout
.
.IP "\fB\-\-metadata\-snapshot\fP"
Use the metadata snapshot rather than the current superblock.
.IP "\fB\-\-written-since {era}\fP"
Blocks written since the given era will be listed.
.SH EXAMPLE
List the blocks that may have been written since the beginning of era
13 on the metadata device /dev/vg/metadata.
.sp
.B era_invalidate --written-since 13 /dev/vg/metadata
The device may not be actively used by the target
when running.
.SH DIAGNOSTICS
.B era_invalidate
returns an exit code of 0 for success or 1 for error (eg, metadata corruption).
.SH SEE ALSO
.B era_check(8),
.B era_dump(8),
.B era_repair(8),
.B era_restore(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>

37
man8/era_invalidate.txt Normal file
View File

@ -0,0 +1,37 @@
NAME
era_invalidate - Provide a list of blocks that have changed since a particular era.
SYNOPSIS
era_invalidate [options] {device|file}
DESCRIPTION
era_invalidate examines era metadata and lists blocks that may have changed
since a given era.
This tool cannot be run on live metadata unless the --metadata-snap option is
used.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-o {output file} Write output to a file rather than stdout.
--metadata-snapshot Use the metadata snapshot rather than the current superblock.
--written-since Blocks written since the given era will be listed.
EXAMPLE
List the blocks that may have been written since the beginning of era 13 on the
metadata device /dev/vg/metadata:
$ era_invalidate --written-since 13 /dev/vg/metadata
The device may not be actively used by the target when running.
DIAGNOSTICS
era_invalidate returns an exit code of 0 for success or 1 for error (eg,
metadata corruption).
SEE ALSO
era_check(8), era_dump(8), era_repair(8), era_restore(8)
AUTHOR
Joe Thornber <ejt@redhat.com>

View File

@ -1,64 +0,0 @@
.TH ERA_RESTORE 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
era_restore \- restore era metadata file to device or file
.SH SYNOPSIS
.B era_restore
.RB [ options ]
.RB -i
.I {device|file}
.RB -o
.I {device|file}
.SH DESCRIPTION
.B era_restore
restores era metadata created by the respective
device-mapper target dumped into an XML formatted (see
.BR era_dump(8) )
.I file
, which optionally can be preprocessed before the restore to another
.I device
or
.I file.
If restored to a metadata
.I device
, the metadata can be processed by the device-mapper target.
This tool cannot be run on live metadata.
.IP "\fB\-i, \-\-input\fP \fI{device|file}\fP"
Input file or device with metadata.
.IP "\fB\-o, \-\-output\fP \fI{device|file}\fP"
Output file or device for repaired binary metadata. If a file is used
then it must be preallocated, and large enough to hold the metadata.
.IP "\fB\-q, \-\-quiet\fP"
Don't print any output. Check the exit code to test for success.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH EXAMPLE
Restores the XML formatted era metadata on file
.B metadata
to logical volume /dev/vg/metadata for further processing by the
respective device-mapper target:
.sp
.B era_restore -i metadata -o /dev/vg/metadata
.SH DIAGNOSTICS
.B era_restore
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B era_dump(8)
.B era_check(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

40
man8/era_restore.txt Normal file
View File

@ -0,0 +1,40 @@
NAME
era_restore - restore era metadata file to device or file.
SYNOPSIS
era_restore [options] -i {xml file} -o {device|file}
DESCRIPTION
era_restore restores era metadata created by the respective device-mapper
target dumped into an XML formatted (see .BR era_dump(8)) file, which
optionally can be preprocessed before the restore to another device or
file. If restored to a metadata device, the metadata can be processed by
the device-mapper target.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-q, --quiet Don't print any output. Check the exit code to test for success.
-i, --input {xml file} Specify input file containing xml metadata.
-o, --output {device|file} Output device or file for restored binary metadata.
If a file is used, then it must be preallocated, and large enough to hold
the metadata.
EXAMPLE
Restores the XML formatted era metadata on file metadata to logical volume
/dev/vg/metadata for further processing by the respective device-mapper
target:
$ era_restore -i metadata -o /dev/vg/metadata
DIAGNOSTICS
era_restore returns an exit code of 0 for success or 1 for error.
SEE ALSO
era_dump(8), era_check(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,67 +0,0 @@
.TH THIN_CHECK 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_check \- validate thin provisioning metadata on device or file
.SH SYNOPSIS
.B thin_check
.RB [ options ]
.I {device|file}
.SH DESCRIPTION
.B thin_check
checks thin provisioning metadata created by
the device-mapper thin provisioning target on a
.I device
or
.I file.
This tool cannot be run on live metadata.
.SH OPTIONS
.IP "\fB\-q, \-\-quiet\fP"
Suppress output messages, return only exit code.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.IP "\fB\-\-super\-block\-only\fP"
Only check the superblock is present.
.IP "\fB\-\-skip-mappings\fP"
Skip checking of the block mappings which make up the bulk of the
metadata.
.IP "\fB\-\-ignore\-non\-fatal\-errors\fP"
.B thin_check
will only return a non-zero exit code if it finds a fatal
error. An example of a non fatal error is an incorrect data block
reference count causing a block to be considered allocated when it in
fact isn't. Ignoring errors for a long time is not advised, you
really should be using thin_repair to fix them.
.SH EXAMPLE
Analyses thin provisioning metadata on logical volume
/dev/vg/metadata:
.sp
.B thin_check /dev/vg/metadata
The device must not be actively used by the target when running.
.SH DIAGNOSTICS
.B thin_check
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B thin_dump(8)
.B thin_repair(8)
.B thin_restore(8)
.B thin_rmap(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

43
man8/thin_check.txt Normal file
View File

@ -0,0 +1,43 @@
NAME
thin_check - validates thin provisioning metadata on a device or file
SYNOPSIS
thin_check [options] {device|file}
DESCRIPTION
thin_check checks thin provisioning metadata created by the device-mapper
thin provisioning target on a device or file.
The tool cannot be run on live metadata.
OPTIONS
-q, --quiet Suppress output messages, return only exit code.
-h, --help Print help and exit.
-V, --version Output version information and exit.
--super-block-only Only check the superblock.
--skip-mappings Skip checking of the block mappings which make up the bulk of the metadata.
--ignore-non-fatal-errors Will only return a non-zero exit code if it finds a fatal error.
An example of a nonfatal error is an incorrect data block reference count
causing a block to be considered allocated when it in fact isn't. Ignoring
errors for a long time is not advised, you really should be using
thin_repair to fix them.
EXAMPLE
Analyses thin provisioning metadata on logical volume /dev/vg/metadata:
$ thin_check /dev/vg/metadata
The device must not be actively used by the target when running.
DIAGNOSTICS
thin_check returns an exit code of 0 for success or 1 for error.
SEE ALSO
thin_dump(8), thin_repair(8), thin_restore(8), thin_rmap(8), thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <heinzm@redhat.com>

View File

@ -1,48 +0,0 @@
.TH THIN_DELTA 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_delta \- Print the differences in the mappings between two thin devices.
.SH SYNOPSIS
.B thin_delta
.RB [ options ]
.I {device|file}
.SH DESCRIPTION
.B thin_delta
allows you to compare the mappings in two thin volumes (snapshots allow common blocks between thin volumes).
This tool cannot be run on live metadata unless the \fB\-\-metadata\-snap\fP option is used.
.SH OPTIONS
.IP "\fB\-\-thin1, \-\-snap1\fP"
The numeric identifier for the first thin volume to diff.
.IP "\fB\-\-thin2, \-\-snap2\fP"
The numeric identifier for the second thin volume to diff.
.IP "\fB\-m, \-\-metadata\-snap\fP [block#]"
If you want to get information out of a live pool then you will need
to take a metadata snapshot and use this switch. In order for the
information to be meaningful you need to ensure the thin volumes
you're examining are not changing (eg, do not activate those thins).
.IP "\fB\-\-verbose"
Provide extra information on the mappings.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH SEE ALSO
.B thin_dump(8)
.B thin_repair(8)
.B thin_restore(8)
.B thin_rmap(8)
.B thin_trim(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>

33
man8/thin_delta.txt Normal file
View File

@ -0,0 +1,33 @@
NAME
thin_delta - Print the differences in the mappings between two thin devices.
SYNOPSIS
thin_delta [options] {device|file}
DESCRIPTION
thin_delta allows you to compare the mappings in two thin volumes (snapshots
allow common blocks between thin volumes).
This tool cannot be run on live metadata unless the --metadata-snap option is
used.
OPTIONS
--thin1, --snap1 {natural} The numeric identifier for the first thin volume to diff.
--thin2, --snap2 {natural} The numeric identifier for the second thin volume to diff.
--metadata-snap [block nr] Use a metadata snapshot.
If you want to get information out of a live pool then you will need to
take a metadata snapshot and use this switch. In order for the information
to be meaningful, you need to ensure the thin volumes you're examining are
not changing (ie, do not activate those thins).
--verbose Provide extra information on the mappings.
-h, --help Print help and exit.
-V, --version Output version information and exit.
SEE ALSO
thin_dump(8), thin_repair(8), thin_restore(8), thin_rmap(8), thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <heinzm@redhat.com>

View File

@ -1,94 +0,0 @@
.TH THIN_DUMP 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_dump \- dump thin provisioning metadata from device or file to standard output
.SH SYNOPSIS
.B thin_dump
.RB [options]
.I {device|file}
.SH DESCRIPTION
.B thin_dump
dumps binary thin provisioning metadata (optionally from alternate block;
see option \-\-metadata-snap) created by the device-mapper
thin provisioning target on a
.I device
or
.I file
to standard output for
analysis or postprocessing in either XML or human readable format.
XML formated metadata can be fed into thin_restore (see
.BR thin_restore(8) )
in order to put it back onto a metadata
.I device
(to process by the device-mapper target) or
.I file.
This tool cannot be run on live metadata unless the \fB\-\-metadata\-snap\fP option is used.
.IP "\fB\-f, \-\-format\fP \fI{xml|human_readable|custom}\fP".
Print output in XML or human readable format. Custom formats are
supported via shared library plugins. They should be specified as in
this example:
.sp
.B thin_dump --format custom=mylib.so /dev/sda
.IP "\fB\-r, \-\-repair\fP".
Repair the metadata whilst dumping it.
.IP "\fB\-m, \-\-metadata\-snap\fP [block#]".
Dump metadata snapshot created by device-mapper thin provisioning target.
If block is not provided, access the default metadata snapshot created by
the thin provisioning device-mapper target, else try the one at block#.
See the thin provisioning target documentation on how to create or release
a metadata snapshot and retrieve the block number from the kernel.
.IP "\fB\-\-dev\-id\fP <dev-id>".
Dump the specified device. This option may be specified multiple
times to select more than one thin device.
.IP "\fB\-\-skip\-mappings".
Do not dump the mappings.
.IP "\fB-o \fI{xml file}\fP"
Specify a file for the output rather than writing to stdout.
.IP "\fB\-h, \-\-help\fP".
Print help and exit.
.IP "\fB\-V, \-\-version\fP".
Output version information and exit.
.SH EXAMPLES
Dumps the thin provisioning metadata on logical volume /dev/vg/metadata
to standard output in human readable format:
.sp
.B thin_dump -f human_readable /dev/vg/metadata
Dumps the thin provisioning metadata on logical volume /dev/vg/metadata
to standard output in XML format:
.sp
.B thin_dump /dev/vg/metadata
Dumps the thin provisioning metadata snapshot on logical volume /dev/vg/metadata
to standard output in human readable format (not processable by
.B thin_restore(8)
):
.sp
.B thin_dump --format human_readable --metadata-snap /dev/vg/metadata
.SH DIAGNOSTICS
.B thin_dump
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B thin_check(8)
.B thin_repair(8)
.B thin_restore(8)
.B thin_rmap(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

68
man8/thin_dump.txt Normal file
View File

@ -0,0 +1,68 @@
NAME
thin_dump - dump thin provisioning metadata from device or file to standard output.
SYNOPSIS
thin_dump [options] {device|file}
DESCRIPTION
thin_dump dumps binary thin provisioning metadata (optionally from alternate
block; see option --metadata-snap) created by the
device-mapper thin provisioning target on a device or file to standard
output for analysis or postprocessing in either XML or human readable format.
XML formated metadata can be fed into thin_restore (see thin_restore(8))
in order to put it back onto a metadata device (to process by the
device-mapper target) or file.
This tool cannot be run on live metadata unless the --metadata-snap
option is used.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-f, --format {xml|human_readable|custom} Choose output format.
Custom formats are supported via shared library plugins. They should be
specified as in this example:
$ thin_dump --format custom=mylib.so /dev/sda
-r, --repair Repair the metadata whilst dumping it.
-m, --metadata-snap {block nr} Dump metadata snapshot.
If block is not provided, access the default metadata snapshot created by
the thin provisioning device-mapper target, else try the one at block nr.
See the thin provisioning target documentation on how to create or release
a metadata snapshot and retrieve the block number from the kernel.
--dev-id {natural} Dump the specified device.
This option may be specified multiple times to select more than one thin
device.
--skip-mappings Do not dump the mappings.
-o {xml file} Specify a file for the output rather than writing to stdout.
EXAMPLES
Dumps the thin provisioning metadata on logical volume /dev/vg/metadata to
standard output in human readable format:
$ thin_dump -f human_readable /dev/vg/metadata
Dumps the thin provisioning metadata on logical volume /dev/vg/metadata to
standard output in XML format:
$ thin_dump /dev/vg/metadata
Dumps the thin provisioning metadata snapshot on logical volume /dev/vg/metadata
to standard output in human readable format (not processable by thin_restore(8)):
$ thin_dump --format human_readable --metadata-snap /dev/vg/metadata
DIAGNOSTICS
thin_dump returns an exit code of 0 for success or 1 for error.
SEE ALSO
thin_check(8) .B thin_repair(8) .B thin_restore(8) .B thin_rmap(8) .B thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,48 +0,0 @@
.TH THIN_LS 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_ls \- List thin volumes within a pool.
.SH SYNOPSIS
.B thin_ls
.RB [ options ]
.I {metadata device}
.SH DESCRIPTION
.B thin_ls
Displays infomation about thin volumes in a pool. Pass the metadata device on the command line, not the
pool device.
This tool cannot be run on live metadata unless the \fB\-\-metadata\-snap\fP option is used.
.SH OPTIONS
.IP "\fB\-o, \-\-format\fP"
Give a comma separated list of the fields to be output. Valid fields are:
DEV, MAPPED_BLOCKS, EXCLUSIVE_BLOCKS, SHARED_BLOCKS, MAPPED_SECTORS,
EXCLUSIVE_SECTORS, SHARED_SECTORS, MAPPED_BYTES, EXCLUSIVE_BYTES,
SHARED_BYTES, MAPPED, EXCLUSIVE, SHARED, TRANSACTION, CREATE_TIME, SNAP_TIME
.IP "\fB\-\-no\-headers\fP"
Don't output headers.
.IP "\fB\-m, \-\-metadata\-snap\fP"
If you want to get information out of a live pool then you will need
to take a metadata snapshot and use this switch.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH SEE ALSO
.B thin_dump(8)
.B thin_repair(8)
.B thin_restore(8)
.B thin_rmap(8)
.B thin_trim(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>

36
man8/thin_ls.txt Normal file
View File

@ -0,0 +1,36 @@
NAME
thin_ls - List thin volumes within a pool.
SYNOPSIS
thin_ls [options] {device|file}
DESCRIPTION
thin_ls displays infomation about thin volumes in a pool. Pass the metadata
device on the command line, not the pool device.
This tool cannot be run on live metadata unless the --metadata-snap
option is used.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-o, --format Give a comma separated list of fields to be output.
Valid fields are:
DEV, MAPPED_BLOCKS, EXCLUSIVE_BLOCKS, SHARED_BLOCKS, MAPPED_SECTORS,
EXCLUSIVE_SECTORS, SHARED_SECTORS, MAPPED_BYTES, EXCLUSIVE_BYTES,
SHARED_BYTES, MAPPED, EXCLUSIVE, SHARED, TRANSACTION, CREATE_TIME,
SNAP_TIME
--no-headers Don't output headers.
-m, --metadata-snap Use metadata snapshot.
If you want to get information out of a live pool then you will need to
take a metadata snapshot and use this switch.
SEE ALSO
thin_dump(8), thin_repair(8), thin_restore(8), thin_rmap(8), thin_trim(8),
thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>

View File

@ -1,79 +0,0 @@
.TH THIN_METADATA_SIZE 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_metadata_size \- thin provisioning metadata device/file size calculator.
.SH SYNOPSIS
.B thin_metadata_size
.RB [ options ]
.SH DESCRIPTION
.B thin_metadata_size
calculates the size of the thin provisioning metadata based on the block size
of the thin provisioned devices, the size of the thin provisioning pool and
the maximum number of all thin prisioned devices and snapshots.
Because thin provisioning pools are holding widely variable contents,
this tool is needed to provide sensible initial default size.
.IP "\fB\-b, \-\-block-size\fP \fIBLOCKSIZE[bskKmMgGtTpPeEzZyY]\fP"
Block size of thin provisioned devices in units of bytes, sectors, kibibytes, kilobytes, ... respectively.
Default is in sectors without a block size unit specifier.
Size/number option arguments can be followed by unit specifiers in short one character
and long form (eg. -b1m or -b1mebibytes).
.IP "\fB\-s, \-\-pool-size\fP \fIPOOLSIZE[bskKmMgGtTpPeEzZyY]\fP"
Thin provisioning pool size in units of bytes, sectors, kibibytes, kilobytes, ... respectively.
Default is in sectors without a pool size unit specifier.
.IP "\fB\-m, \-\-max-thins\fP \fI#[bskKmMgGtTpPeEzZyY]\fP"
Maximum sum of all thin provisioned devices and snapshots.
Unit identifier supported to allow for convenient entry of large quantities, eg. 1000000 = 1M.
Default is absolute quantity without a number unit specifier.
.IP "\fB\-u, \-\-unit\fP \fI{bskKmMgGtTpPeEzZyY}\fP"
Output unit specifier in units of bytes, sectors, kibibytes, kilobytes, ... respectively.
Default is in sectors without an output unit specifier.
.IP "\fB\-n, \-\-numeric-only [short|long]\fP"
Limit output to just the size number with the optional unit specifier character/string.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH EXAMPLES
Calculates the thin provisioning metadata device size for block size 64 kibibytes,
pool size 1 tebibytes and maximum number of thin provisioned devices and snapshots of 1000
in units of sectors with long output:
.sp
.B thin_metadata_size -b64k -s1t -m1000
Or (using the long options instead) for block size 1 gibibyte, pool size 1 petabyte and maximum number of thin provisioned devices
and snapshots of 1 million with numeric-only output in units of gigabytes:
.sp
.B thin_metadata_size --block-size=1g --pool-size=1P --max-thins=1M --unit=G --numeric-only
Same as before (1g, 1P, 1M, numeric-only) but with unit specifier character appended:
.sp
.B thin_metadata_size --block-size=1gibi --pool-size=1petabytes --max-thins=1mega --unit=G --numeric-only=short
Or with unit specifier string appended:
.sp
.B thin_metadata_size --block-size=1gibi --pool-size=1petabytes --max-thins=1mega --unit=G -nlong
.SH DIAGNOSTICS
.B thin_metadata_size
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B thin_dump(8)
.B thin_check(8)
.B thin_repair(8)
.B thin_restore(8)
.B thin_rmap(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -0,0 +1,75 @@
NAME
thin_metadata_size - thin provisioning metadata device/file size calculator.
SYNOPSIS
thin_metadata_size [options]
DESCRIPTION
thin_metadata_size calculates the size of the thin provisioning metadata
based on the block size of the thin provisioned devices, the size of the thin
provisioning pool and the maximum number of all thin provisioned devices and
snapshots. Because thin provisioning pools are holding widely variable
contents, this tool is needed to provide sensible initial default size.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-b, --block-size {BLOCKSIZE[bskKmMgGtTpPeEzZyY]} Set block size.
Block size of thin provisioned devices in units of bytes, sectors,
kibibytes, kilobytes, ... respectively. Default is in sectors without a
block size unit specifier. Size/number option arguments can be followed by
unit specifiers in short one character and long form (eg. -b1m or
-b1mebibytes).
-s, --pool-size {POOLSIZE[bskKmMgGtTpPeEzZyY]} Set pool size.
Thin provisioning pool size in units of bytes, sectors, kibibytes,
kilobytes, ... respectively. Default is in sectors without a pool size
unit specifier.
-m, --max-thins {count[bskKmMgGtTpPeEzZyY]} Set max thins.
Maximum sum of all thin provisioned devices and snapshots. Unit identifier
supported to allow for convenient entry of large quantities, eg. 1000000 =
1M. Default is absolute quantity without a number unit specifier.
-u, --unit {bskKmMgGtTpPeEzZyY}
Output unit specifier in units of bytes, sectors, kibibytes, kilobytes, ...
respectively. Default is in sectors without an output unit specifier.
-n, --numeric-only {short|long} Limit output to just the size number.
EXAMPLES
Calculates the thin provisioning metadata device size for block size 64
kibibytes, pool size 1 tebibytes and maximum number of thin provisioned
devices and snapshots of 1000 in units of sectors with long output:
$ thin_metadata_size -b64k -s1t -m1000
Or (using the long options instead) for block size 1 gibibyte, pool size 1
petabyte and maximum number of thin provisioned devices and snapshots of 1
million with numeric-only output in units of gigabytes:
$ thin_metadata_size --block-size=1g --pool-size=1P --max-thins=1M --unit=G --numeric-only
Same as before (1g, 1P, 1M, numeric-only) but with unit specifier character
appended:
$ thin_metadata_size --block-size=1gibi --pool-size=1petabytes --max-thins=1mega --unit=G --numeric-only=short
Or with unit specifier string appended:
$ thin_metadata_size --block-size=1gibi --pool-size=1petabytes --max-thins=1mega --unit=G -nlong
DIAGNOSTICS
thin_metadata_size returns an exit code of 0 for success or 1 for error.
SEE ALSO
thin_dump(8), thin_check(8), thin_repair(8), thin_restore(8), thin_rmap(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,66 +0,0 @@
.TH THIN_REPAIR 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_repair \- repair thin provisioning binary metadata from device/file to device/file
.SH SYNOPSIS
.B thin_repair
.RB [ options ]
.RB -i
.I {device|file}
.RB -o
.I {device|file}
.SH DESCRIPTION
.B thin_repair
reads binary thin provisioning metadata created by the
respective device-mapper target from one
.I device
or
.I file
, repairs it and writes it to another
.I device
or
.I file.
If written to a metadata
.I device
, the metadata can be processed
by the device-mapper target.
This tool cannot be run on live metadata.
.IP "\fB\-i, \-\-input\fP \fI{device|file}\fP"
Input file or device with binary metadata.
.IP "\fB\-o, \-\-output\fP \fI{device|file}\fP"
Output file or device for repaired binary metadata. If a file is used
then it must be preallocated, and large enough to hold the metadata.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH EXAMPLE
Reads the binary thin provisioning metadata from file
.B metadata
, repairs it and writes it to logical volume /dev/vg/metadata
for further processing by the respective device-mapper target:
.sp
.B thin_repair -i metadata -o /dev/vg/metadata
.SH DIAGNOSTICS
.B thin_repair
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B thin_dump(8)
.B thin_check(8)
.B thin_restore(8)
.B thin_rmap(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

40
man8/thin_repair.txt Normal file
View File

@ -0,0 +1,40 @@
NAME
thin_repair - repair thin provisioning binary metadata.
SYNOPSIS
thin_repair [options] -i {device|file} -o {device|file}
DESCRIPTION
thin_repair reads binary thin provisioning metadata created by the respective
device-mapper target from one device or file, repairs it and writes it
to different device or file. If written to a metadata device, the metadata
can be processed by the device-mapper target.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-i, --input {device|file} Input file or device with binary data.
-o, --output {device|file} Output file or device for binary data.
If a file is used for output, then it must be preallocated, and large
enough to hold the metadata.
EXAMPLE
Reads the binary thin provisioning metadata from file metadata, repairs
it and writes it to logical volume /dev/vg/metadata for further processing by
the respective device-mapper target:
$ thin_repair -i metadata -o /dev/vg/metadata
DIAGNOSTICS
thin_repair returns an exit code of 0 for success or 1 for error.
SEE ALSO
thin_dump(8), thin_check(8), thin_restore(8), thin_rmap(8), thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,67 +0,0 @@
.TH THIN_RESTORE 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_restore \- restore thin provisioning metadata file to device or file
.SH SYNOPSIS
.B thin_restore
.RB [ options ]
.RB -i
.I {device|file}
.RB -o
.I {device|file}
.SH DESCRIPTION
.B thin_restore
restores thin provisioning metadata created by the
respective device-mapper target dumped into an XML formatted (see
.BR thin_dump(8) )
.I file
, which optionally can be preprocessed before the restore to another
.I device
or
.I file.
If restored to a metadata
.I device
, the metadata can be processed by the device-mapper target.
This tool cannot be run on live metadata.
.IP "\fB\-q, \-\-quiet\fP"
Suppress output messages, return only exit code.
.IP "\fB\-i, \-\-input\fP \fI{device|file}\fP"
Input file or device with metadata.
.IP "\fB\-o, \-\-output\fP \fI{device|file}\fP"
Output file or device for repaired binary metadata. If a file is used
then it must be preallocated, and large enough to hold the metadata.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH EXAMPLE
Restores the XML formatted thin provisioning metadata on file
.B metadata
to logical volume /dev/vg/metadata for further processing by the
respective device-mapper target:
.sp
.B thin_restore -i metadata -o /dev/vg/metadata
.SH DIAGNOSTICS
.B thin_restore
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B thin_dump(8)
.B thin_check(8)
.B thin_repair(8)
.B thin_rmap(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

42
man8/thin_restore.txt Normal file
View File

@ -0,0 +1,42 @@
NAME
thin_restore - restore thin provisioning metadata file to device or file.
SYNOPSIS
thin_restore [options] -i {xml file} -o {device|file}
DESCRIPTION
thin_restore restores thin provisioning metadata created by the respective
device-mapper target dumped into an XML formatted (see thin_dump(8)) file,
which optionally can be preprocessed before the restore to another device or
file. If restored to a metadata device, the metadata can be processed by the
device-mapper target.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
-q, --quiet Suppress output messages, return only exit code.
-i, --input {xml file} Input file containing XML metadata.
-o, --output {device|file} Output file or device for restored binary metadata.
If a file is used for output, then it must be preallocated, and large
enough to hold the metadata.
EXAMPLE
Restores the XML formatted thin provisioning metadata on file metadata to
logical volume /dev/vg/metadata for further processing by the respective
device-mapper target:
$ thin_restore -i metadata -o /dev/vg/metadata
DIAGNOSTICS
thin_restore returns an exit code of 0 for success or 1 for error.
SEE ALSO
thin_dump(8), thin_check(8), thin_repair(8), thin_rmap(8), thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,50 +0,0 @@
.TH THIN_RMAP(g 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_rmap \- output reverse map of a thin provisioned region of blocks from metadata device or file
.SH SYNOPSIS
.B thin_rmap
.RB [options]
.I {device|file}
.SH DESCRIPTION
.B thin_rmap
outputs the reverse mapping stored in the metadata on a
.I device
or
.I file
between a region of thin provisioned pool blocks
and the associated thin provisioned devices.
This tool cannot be run on live metadata.
.IP "\fB\\-\-region\fP \fI<block range>\fP".
output reverse map
.IP "\fB\-h, \-\-help\fP".
Print help and exit.
.IP "\fB\-V, \-\-version\fP".
Output version information and exit.
.SH EXAMPLES
output reverse map for pool blocks 5..45 (denotes blocks 5 to 44 inclusive, but not block 45)
.sp
.B thin_rmap --region 5..45 /dev/pool-metadata
.SH DIAGNOSTICS
.B thin_rmap
returns an exit code of 0 for success or 1 for error.
.SH SEE ALSO
.B thin_check(8)
.B thin_dump(8)
.B thin_repair(8)
.B thin_restore(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>
.br
Heinz Mauelshagen <HeinzM@RedHat.com>

37
man8/thin_rmap.txt Normal file
View File

@ -0,0 +1,37 @@
NAME
thin_rmap - output reverse map of a thin provisioned region of blocks from
metadata device or file.
SYNOPSIS
thin_rmap [options] {device|file}
DESCRIPTION
thin_rmap outputs the reverse mapping stored in the metadata on a device or
file between a region of thin provisioned pool blocks and the associated thin
provisioned devices.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
--region {block range} Specify range of blocks on the data device.
At least one region must be specified. Multiple regions may be specified.
The range takes the format <begin>..<one past the end>. For example,
"5..45" specifies data blocks 5 to 44 inclusive, but not 45.
EXAMPLES
$ thin_rmap --region 5..45 /dev/pool-metadata
DIAGNOSTICS
thin_rmap returns an exit code of 0 for success or 1 for error.
SEE ALSO
thin_check(8), thin_dump(8), thin_repair(8), thin_restore(8), thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>, Heinz Mauelshagen <HeinzM@RedHat.com>

View File

@ -1,37 +0,0 @@
.TH THIN_TRIM 8 "Thin Provisioning Tools" "Red Hat, Inc." \" -*- nroff -*-
.SH NAME
thin_trim \- Issue discard requests for free pool space (offline tool).
.SH SYNOPSIS
.B thin_trim
.RB [ options ]
.B --metadata-dev
.I {device|file}
.B --data-dev
.I {device|file}
.SH DESCRIPTION
.B thin_trim
sends discard requests to the pool device for unprovisioned areas.
This tool cannot be run on live metadata.
.SH OPTIONS
.IP "\fB\-\-pool-inactive\fP"
Indicates you are aware the pool should be inactive. Suppresses a warning message and prompt.
.IP "\fB\-h, \-\-help\fP"
Print help and exit.
.IP "\fB\-V, \-\-version\fP"
Output version information and exit.
.SH SEE ALSO
.B thin_dump(8)
.B thin_repair(8)
.B thin_restore(8)
.B thin_rmap(8)
.B thin_metadata_size(8)
.SH AUTHOR
Joe Thornber <ejt@redhat.com>

23
man8/thin_trim.txt Normal file
View File

@ -0,0 +1,23 @@
NAME
thin_trim - Issue discard requests for free pool space (offline tool).
SYNOPSIS
thin_trim [options] --metadata-dev {device|file} --data-dev {device|file}
DESCRIPTION
thin_trim sends discard requests to the pool device for unprovisioned areas.
This tool cannot be run on live metadata.
OPTIONS
-h, --help Print help and exit.
-V, --version Print version information and exit.
--pool-inactive Indicates you are aware the pool should be inactive.
Suppresses a warning message and prompt.
SEE ALSO
thin_dump(8), thin_repair(8), thin_restore(8), thin_rmap(8), thin_metadata_size(8)
AUTHOR
Joe Thornber <ejt@redhat.com>