-.TH ll_decode_filter_fid 1 "Dec 15, 2010" Lustre "utilities"
+.TH LL_DECODE_FILTER_FID 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
ll_decode_filter_fid \- display Lustre Object ID and MDT parent FID
.SH SYNOPSIS
-.B ll_decode_filter_fid
+.SY ll_decode_filter_fid
.I object_file
.RI [ "object_file ..." ]
-.br
+.YS
.SH DESCRIPTION
.B ll_decode_filter_fid
decodes and prints the Lustre OST object ID and MDT FID and stripe index
.PP
The OST object ID (objid) is useful in case of OST directory corruption,
though normally the
-.SH EXAMPLE
-.fi
-root@oss1# cd /mnt/ost/lost+found
-.fi
-root@oss1# ll_decode_filter_fid #12345[4,5,8]
-.fi
-#123454: objid=690670 seq=0 parent=[0x751c5:0xfce6e605:0x0]
-.fi
-#123455: objid=614725 seq=0 parent=[0x18d11:0xebba84eb:0x1]
-.fi
-#123458: objid=533088 seq=0 parent=[0x21417:0x19734d61:0x0]
+.SH EXAMPLES
+.EX
+.B root@oss1# cd /mnt/ost/lost+found
+.B root@oss1# ll_decode_filter_fid #12345[4,5,8]
+\𞈾: objid=690670 seq=0 parent=[0x751c5:0xfce6e605:0x0]
+\𞈿: objid=614725 seq=0 parent=[0x18d11:0xebba84eb:0x1]
+\𞉂: objid=533088 seq=0 parent=[0x21417:0x19734d61:0x0]
+.EE
.PP
This shows that the 3 files in lost+found have decimal object IDs 690670,
614725, and 533088, respectively. The object sequence number (formerly
.PP
The idx field shows the stripe number of this OST object in the Lustre
RAID-0 striped file.
+.SH AVAILABILITY
+.B ll_decode_filter_fid
+is part of the
+.BR lustre (7)
+filesystem package since release 2.0.0
+.\" Added in commit v1_10_0_43-39-g48ee13b5ee
.SH SEE ALSO
-.BR lustre (7),
+.BR lustre (7)
-.TH ll_decode_linkea 1 "May 25, 2016" Lustre "utilities"
+.TH LL_DECODE_LINKEA 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
ll_decode_linkea \- display parent FIDs and name of a Lustre file
.SH SYNOPSIS
-.B ll_decode_linkea
-.I file
-.RI [ "file ..." ]
-.br
+.SY ll_decode_linkea
+.I FILE
+.RI [ FILE ...]
+.YS
.SH DESCRIPTION
.B ll_decode_linkea
decodes and prints the Lustre parent FIDs and name of a Lustre file, which
recover most of the ldiskfs hierarchy, but it might leave some files or
directories under lost+found. The parent FIDs can be used to determine the
right Lustre paths to move them to.
-.SH EXAMPLE
-.fi
-root@mdt1# cd /mnt/mdt/lost+found
-.fi
-root@mdt1# ll_decode_linkea \#123451 \#123452
-.fi
-#123451: count 2
-.fi
+.SH EXAMPLES
+.EX
+.B root@mdt1# cd /mnt/mdt/lost+found
+.B root@mdt1# ll_decode_linkea \#123451 \#123452
+\𞈻: count 2
0: [0x200034021:0x2:0x0], name "file1"
-.fi
1: [0x200034021:0x1:0x0], name "file1_link"
-.fi
-#123452: count 1
-.fi
+\𞈼: count 1
0: [0x200000007:0x1:0x0], name "file2"
+.EE
.PP
-This shows that the 2 files in lost+found. ll_decode_linkea prints all of the
-parent FIDs of these files. file1 has two hard links, that is why it has two
-parent directories. Command
+This shows that there are 2 files in lost+found.
+.B ll_decode_linkea
+prints all of the parent FIDs of these files.
+file1 has two hard links, that is why it has two parent directories.
.B lfs fid2path
could be used to extract the paths of the parents.
.PP
Since "trusted.link" xattr is accessible on Lustre client too,
.B ll_decode_linkea
could also be used on Lustre client.
-.PP
+.SH AVAILABILITY
+.B ll_decode_linkea
+is part of the
+.BR lustre (7)
+filesystem package since release 2.9.0
+.\" Added in commit v2_8_59_0-28-g11a138cb9d
.SH SEE ALSO
.BR lfs (1),
.BR lustre (7)
-.TH lljobstat 8 "Oct 14, 2022" Lustre "utilities"
-
+.TH LLJOBSTAT 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
lljobstat \- display top jobs and statistics
-
.SH SYNOPSIS
-.B "llobdstat"
+.SY "llobdstat"
.RB [ -c|--count
.IR COUNT ]
.RB [ -i|--interval
.IR PARAM ]
.RB [ --fullname ]
.RB [ --no-fullname ]
-
+.YS
.SH DESCRIPTION
.B lljobstat
reads and parses the job_stats files on local node,
Repeat for some times or forever with given interval.
.P
Type Ctrl-C to stop printing.
-
.SS Abbreviations
-\fBlljobstat\fR displays abbreviations of operation names as listed below:
+.B lljobstat
+displays abbreviations of operation names as listed below:
.P
.nf
ops: total number of operations
qc: quotactl, rd: read, rm: rmdir, sa: setattr, si: set_info
st: statfs, sx: setxattr, sy: sync, ul: unlink, wr: write
.fi
-
-.SH "OPTIONS"
+.SH OPTIONS
.TP
-\fB-c|--count\fR \fICOUNT\fR
+.BR -c ", " --count \ \fICOUNT
how many top jobs are listed. Default 5.
.TP
-\fB-i|--interval\fR \fIINTERVAL\fR
+.BR -i ", " --interval \ \fIINTERVAL
interval in seconds before list top jobs again. Default 10.
.TP
-\fB-n|--repeats\fR \fIREPEATS\fR
+.BR -n ", " --repeats \ \fIREPEATS
how many times to repeat. Default unlimited.
.TP
-\fB-m|--mdt\fR
+.BR -m ", " --mdt
get data from only job_stats of MDTs.
.TP
-\fB-o|--ost\fR
+.BR -o ", " --ost
get data from only job_stats of OSTs.
.TP
-\fB--param\fR \fIPARAM\fR
+.BI --param " PARAM"
get data from only PARAM path. For example, "*.lustre-*.job_stat".
.TP
-\fB--fullname\fR
+.B --fullname
show full name of operations. Default no.
.TP
-\fB--no-fullname\fR
+.B --no-fullname
show abbreviated name of operations.
.TP
-\fB-h|--help\fR
+.BR -h ", " --help
print help message.
-
-.SH EXAMPLE
-.nf
-# lljobstat -n 1 -c 3
+.SH EXAMPLES
+.EX
+.B # lljobstat -n 1 -c 3
---
timestamp: 1665623345
top_jobs:
- touch.500: {ops: 48, op: 8, cl: 8, mn: 8, ga: 8, sa: 16}
- dd.0: {ops: 38, op: 4, cl: 4, mn: 1, ga: 1, sa: 3, gx: 3, wr: 19, pu: 3}
\[char46]..
-.fi
+.EE
+.SH AVAILABILITY
+.B lljobstat
+is part of the
+.BR lustre (7)
+filesystem package since release 2.16.0
+.\" Added in commit v2_15_53-96-ge2812e8773
+.SH SEE ALSO
+.BR llobdstat (8)
-.TH llobdstat 1 "Jul 7, 2008" Lustre "utilities"
+.TH LLOBDSTAT 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
llobdstat \- display OST statistics
.SH SYNOPSIS
-.B "llobdstat ost_name [interval]"
-.br
+.SY llobdstat
+.I ost_name
+.RI [ interval ]
+.YS
.SH DESCRIPTION
.B llobdstat
displays a line of OST statistics for the given
-.I ost_name
+.I OST_NAME
every
-.I interval
+.I INTERVAL
seconds. It should be run directly on an OSS node.
Type control-C to stop statistics printing.
-.SH EXAMPLE
-.nf
-# llobdstat liane-OST0002 1
+.SH EXAMPLES
+.EX
+.B # llobdstat liane-OST0002 1
/usr/bin/llobdstat on liane-OST0002
Processor counters run at 2800.189 MHz
Read: 1.21431e+07, Write: 9.93363e+08, create/destroy: 24/1499, stat: 34, punch: 18
[NOTE: cx: create, dx: destroy, st: statfs, pu: punch ]
-
+\&
Timestamp Read-delta ReadRate Write-delta WriteRate
--------------------------------------------------------
1217026053 0.00MB 0.00MB/s 0.00MB 0.00MB/s
1217026057 0.00MB 0.00MB/s 0.00MB 0.00MB/s
1217026058 0.00MB 0.00MB/s 0.00MB 0.00MB/s
1217026059 0.00MB 0.00MB/s 0.00MB 0.00MB/s st:1
-...
-.fi
+\&...
+.EE
+.SH AVAILABILITY
+.B llobdstat
+is part of the
+.BR lustre (7)
+filesystem package since release 0.8.0
+.\" Added in commit 0.7.1-38-g576c9a8212
+.SH SEE ALSO
+.BR lljobstat (8)
-.TH llog_reader 8 "2009 Apr 02" Lustre "System management commands"
+.TH LLOG_READER 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
llog_reader \- lustre on-disk log parsing utility
.SH SYNOPSIS
-.B "llog_reader filename"
-.br
+.SY llog_reader
+.I FILENAME
+.YS
.SH DESCRIPTION
.B llog_reader
parses the binary format of Lustre's on-disk configuration logs.
-It can only read the logs. Use
+It can only read the logs. Use
.B tunefs.lustre
to write to them.
-.LP
+.SH CAVEATS
+Although they are stored in the CONFIGS directory,
+.I mountdata
+files do not use the config log format and will confuse
+.BR llog_reader .
+.SH EXAMPLES
To examine a log file on a stopped Lustre server, first mount its
backing file system as ldiskfs, then use
.B llog_reader
-to dump the log file's contents, e.g.
-.IP
-.nf
-mount -t ldiskfs /dev/sda /mnt/mgs
+to dump the log file's contents:
+.RS
+.EX
+.B # mount -t ldiskfs /dev/sda /mnt/mgs
llog_reader /mnt/mgs/CONFIGS/tfs-client
-.fi
-.LP
+.EE
+.RE
+.PP
To examine the same log file on a running Lustre server, use the
ldiskfs-enabled debugfs utility (called
.B debug.ldiskfs
-on some distros) to extract the file, e.g.
-.IP
-.nf
-debugfs -c -R 'dump CONFIGS/tfs-client /tmp/tfs-client' /dev/sda
+on some distros) to extract the file:
+.RS
+.EX
+.B # debugfs -c -R 'dump CONFIGS/tfs-client /tmp/tfs-client' /dev/sda
llog_reader /tmp/tfs-client
-.fi
-.LP
+.EE
+.RE
+.PP
To examine Changelog records on a stopped Lustre server, first mount its
backing file system as ldiskfs, then use
.B llog_reader
-to dump the log changelog's contents, e.g.
-.IP
-.nf
-# mount -t ldiskfs /dev/sda /mnt/mgs
-# llog_reader /mnt/mgs/changelog_catalog
+to dump the log changelog's contents:
+.RS
+.EX
+.B # mount -t ldiskfs /dev/sda /mnt/mgs
+.B # llog_reader /mnt/mgs/changelog_catalog
rec #1 type=1064553b len=64 offset 8192
Header size : 8192
Time : Mon Jan 22 23:28:24 2018
Number of records: 1
Target uuid :
-----------------------
-.fi
-#01 (064)id=[0x5:0x1:0x0]:0 path=
-.B "O/1/d5/5"
-
-# llog_reader /mnt/mgs/
-.B O/1/d5/5
-.nf
+\ (064)id=[0x5:0x1:0x0]:0 path=
+.I "O/1/d5/5"
+.B # llog_reader /mnt/mgs/
+.I O/1/d5/5
rec #1 type=10660000 len=136 offset 8192
rec #2 type=10660000 len=136 offset 8328
rec #3 type=10660000 len=128 offset 8464
Number of records: 3
Target uuid :
-----------------------
-#01 (136)changelog record id:0x0 cr_flags:0x9000 cr_type:CREAT(0x1)
+\ (136)changelog record id:0x0 cr_flags:0x9000 cr_type:CREAT(0x1)
date:'14:30:01.370700741 2018.01.22' target:[0x200000402:0x1:0x0]
cr_extra_flags:0x3 user:0:0 nid:10.128.11.159@tcp parent:[0x200000007:0x1:0x0]
name:fileA
-.fi
-.SH CAVEATS
-Although they are stored in the CONFIGS directory, \fImountdata\fR
-files do not use the config log format and will confuse \fBllog_reader\fR.
+.EE
+.RE
+.SH AVAILABILITY
+.B llog_reader
+is part of the
+.BR lustre (7)
+filesystem package since release 1.4.0
+.\" Added in commit 1.3.4-1130-g113303973e
.SH SEE ALSO
-Lustre Operations Manual, Section 21.1, \fITroubleshooting Lustre\fR.
-.br
.BR lustre (7),
.BR tunefs.lustre (8)
+.P
+Lustre Operations Manual,
+.IR "Troubleshooting Lustre" .
-.TH llsom_sync 8 "2018 Jan 10" Lustre "Lustre Filesystem utility"
+.TH LLSOM_SYNC 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
llsom_sync \- Utility to sync file LSOM xattr.
.SH SYNOPSIS
-.br
-.B llsom_sync --mdt|-m <mdt> --user|-u <user id>
-.br
-.B\t\t\t [--daemonize|-d] [--verbose|-v] [--interval|-i]
-.br
-.B\t\t\t [--min-age|-a] [--max-cache|-c] [--sync|-s] <lustre_mount_point>
-.br
-
+.SY llsom_sync
+.RB { --mdt | -m }\c
+.BI = MDT
+.RB { --user | -u }\c
+.BI = USER_ID
+.RB [ --daemonize | -d ]
+.RB [ --verbose | -v ]
+.RB [ --interval | -i ]
+.RB [ --min-age | -a ]
+.RB [ --max-cache | -c ]
+.RB [ --sync | -s ]
+.I LUSTRE_MOUNT_POINT
+.YS
.SH DESCRIPTION
.B llsom_sync
is designed to sync file LSOM xattr in the Lustre Filesystem by
using Lustre MDT changelogs. A changelog user must be registered
(see lctl (8) changelog_register) before using this tool.
-
.SH OPTIONS
-
-.B --mdt=<mdt>
-.br
+.TP
+.BR -m ", " --mdt= \fIMDT
The metadata device which need to be synced the LSOM xattr of files.
A changelog user must be registered for this device.
-
-.B --user=<user id>
-.br
+.TP
+.BR -u ", " --user= \fIUSER_ID
The changelog user id for the above MDT device. See lctl(8) changelog_register.
-
+.TP
.B --daemonize
-.br
-Daemonize the program. In daemon mode, the utility will scan, process the
-changelog records and sync the LSOM xattr for files periodically.
-
+Daemonize the program. In daemon mode, the utility will scan,
+process the changelog records and sync the LSOM xattr for files periodically.
+.TP
.B --verbose
-.br
Produce a verbose output.
-
+.TP
.B --interval
-.br
The time interval to scan the Lustre changelog and process the log record in
daemon mode.
-
+.TP
.B --min-age
-.br
The time that llsom_sync tool will not try to sync the SOM data for any files
-closed less than this many seconds old. The default min-age value is 600s
-(10 minutes).
-
+closed less than this many seconds old.
+The default min-age value is 600s (10 minutes).
+.TP
.B --max-cache
-.br
The total memory used for the FID cache which can be with a suffix [KkGgMm].
The default max-cache value is 256MB. For the parameter value < 100, it is
taken as the percentage of total memory size used for the FID cache instead
of the cache size.
-
+.TP
.B --sync
-.br
Sync file data to make the dirty data out of cache to ensure the blocks count
is correct when update the file LSOM xattr. This option could hurt server
performance significantly if thousands of fsync requests are sent.
-
.SH EXAMPLES
-
-.TP
-Register a changelog consumer for MDT lustre-MDT0000
-$ cl_user=$(ssh root@mds01 lctl --device lustre-MDT0000 changelog_register -n)
-.br
-$ echo $cl_user
-.br
+Register a changelog consumer for MDT lustre-MDT0000:
+.RS
+.EX
+.B # cl_user=$(ssh root@mds01 lctl --device lustre-MDT0000 changelog_register -n)
+.B # echo $cl_user
cl1
-
-.TP
+.EE
+.RE
+.PP
After perform some file operations on the Lustre Filesystem with
mount point '/mnt/lustre' and the filesystem undergoes some changes,
sync file LSOM xattr:
-$ llsom_sync --mdt=lustre-MDT0000 --user=$cl_user \\
-.br
- --verbose /mnt/lustre
-
-.TP
+.RS
+.EX
+.B # llsom_sync --mdt=lustre-MDT0000 --user=$cl_user --verbose /mnt/lustre
+.EE
+.RE
+.PP
To deregister the changelog user (e.g. after this example, or if SOM
updates are no longer needed):
-.br
-$ ssh root@mds01 lctl --device lustre-MDT0000 changelog_deregister $cl_user
-
-.SH AUTHOR
-The llsom_sync command is part of the Lustre filesystem.
-
+.RS
+.EX
+.B # ssh root@mds01 lctl --device lustre-MDT0000 changelog_deregister $cl_user
+.EE
+.RE
+.SH AVAILABILITY
+The
+.B llsom_sync
+command is part of the
+.BR lustre (7)
+filesystem package since release 2.12.0
+.\" Added in commit v2_11_53_0-60-gcaba6b9af0
.SH SEE ALSO
.BR lustre (7),
.BR lctl (8)
-.TH llstat 1 "Jul 7, 2008" Lustre "utilities"
+.TH LLSTAT 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
llstat \- print Lustre statistics
.SH SYNOPSIS
-.B "llstat [-c] [-g] [-i interval] stats_file"
-.br
+.SY llstat
+.RB [ -c ]
+.RB [ -g ]
+.RB [ -i
+.IR INTERVAL ]
+.I STATS_FILE
+.YS
.SH DESCRIPTION
.B llstat
can display statistics from any of several lustre stats files that
-share a common format, updated every \fIinterval\fR seconds.
+share a common format, updated every
+.I INTERVAL
+seconds.
Use control-C to stop statistics printing.
+.SH OPTIONS
.TP
-.I "\-c"
+.B -c
Clear the stats file first.
.TP
-.I "\-i interval"
+.BI -i " INTERVAL"
Polling period in seconds.
.TP
-.I "\-g"
+.B -g
Graphable output format.
.TP
-.I "\-h"
+.B -h
Display help information.
.TP
-.I "stats_file"
+.I STATS_FILE
Either the full path to a stats file, or the shorthand:
-\fImds\fR or \fIost\fR.
-.SH EXAMPLE
-To monitor the OST RPC stats every second:
-.IP
-llstat -i 1 ost
+.I mds
+or
+.IR ost .
.SH FILES
-.nf
+.EX
llite.*.stats
lwp.*.stats
mdc.*.stats
ldlm.services.*.stats
mds.MDS.*.stats
ost.OSS.*.stats
-.fi
+.EE
+.SH EXAMPLES
+To monitor the OST RPC stats every second:
+.RS
+.EX
+.B # llstat -i 1 ost
+.EE
+.RE
+.SH AVAILABILITY
+.B llstat
+is part of the
+.BR lustre (7)
+filesystem package since release 0.8.0
+.\" Added in commit 0.7.1-38-g576c9a8212
+.SH SEE ALSO
+.BR lljobstat (8)
.\" Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved.
.\" This file may be copied under the terms of the GNU Public License, v2.
.\"
-.TH llverdev 8 "2008 Mar 15" Lustre "configuration utilities"
+.TH LLVERDEV 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
llverdev - verify a block device is functioning properly over its full size
.SH SYNOPSIS
-.BI llverdev
+.SY llverdev
.RB [ -c
-.IR chunksize ]
+.IR CHUNKSIZE ]
.RB [ -f "] [" -h ]
.RB [ -o
-.IR offset_kb ]
+.IR OFFSET_KB ]
.RB [ -l "] [" -p "] [" -r ]
.RB [ -s
-.IR size_mb ]
+.IR SIZE_MB ]
.RB [ -t
-.IR timestamp ]
+.IR TIMESTAMP ]
.RB [ -v "] [" -w ]
-.I device
+.I DEVICE
+.YS
.SH DESCRIPTION
Sometimes kernel drivers or hardware devices have bugs that prevent them from
accessing the full device size correctly, or possibly have bad sectors on disk
device is minimally sane before investing the time in a full verification.
.SH OPTIONS
.TP
-.BR -c | --chunksize " \fIchunk_mb"
+.BR -c ", " --chunksize \ \fICHUNK_MB
IO chunk size in megabytes (default=1), with optional KMG suffix.
.TP
-.BR -f | --force
+.BR -f ", " --force
force test to run without confirmation that the device will be overwritten
and all data therein will be permanently destroyed.
.TP
-.BR -h | --help
+.BR -h ", " --help
display a brief help message.
.TP
-.BR -l | --long
+.BR -l ", " --long
Run a full check, writing and then reading and verifying every block on the
disk.
.TP
-.BR -o | --offset " \fIoffset_kb"
+.BR -o ", " --offset \ \fIOFFSET_KB
offset of IO start in kilobytes (default=0), with optional KMGTP suffix.
.TP
-.BR -p | --partial
+.BR -p ", " --partial
Run a partial check, only doing periodic checks across the device (1GB steps).
.TP
-.BR -r | --read
+.BR -r ", " --read
Run test in read (verify) mode only, after having run the test in
.B -w
mode previously.
.TP
-.BR -s | --size " \fIsize_mb"
+.BR -s ", " --size \ \fISIZE_MB
device or file size in megabytes to use for the test, with optional KMGTP
suffix. If unspecified, use the actual device or file size (or write until
an error is hit if 0).
.TP
-.BR -t | --timestamp " \fItimestamp"
+.BR -t ", " --timestamp \ \fITIMESTAMP
Set test start time as printed at the start of a previously interrupted test
to ensure that the validation data is the same across the whole filesystem
(default=current time())
.TP
-.BR -v | --verbose
+.BR -v ", " --verbose
Run test in verbose mode, listing each read and write operation.
.TP
-.BR -w | --write
+.BR -w ", " --write
Run test in write (test-pattern) mode (default run both read and write)
.SH EXAMPLES
-.TP
-Run a partial device verification on \fB/dev/sda\fR:
-.B llverdev -v -p /dev/sda
-.br
+Run a partial device verification on
+.BR /dev/sda :
+.RS
+.EX
+.B # llverdev -v -p /dev/sda
llverdev: permanently overwrite all data on /dev/sda (yes/no)? y
-.br
llverdev: /dev/sda is 4398046511104 bytes (4096.0 GB) in size
-.br
Timestamp: 1009839028
-.br
Current write offset: 4096 kB
-.TP
-Continue an interrupted verification at offset \fB4096\fRkB from the start of the device, using the same timestamp as the previous run:
-.B llverdev -f -v -p --offset=4096 --timestamp=1009839028 /dev/sda
-.br
+.EE
+.RE
+.PP
+Continue an interrupted verification at offset
+.BR 4096 kB
+from the start of the device, using the same timestamp as the previous run:
+.RS
+.EX
+.B # llverdev -f -v -p --offset=4096 --timestamp=1009839028 /dev/sda
llverdev: /dev/sda is 4398046511104 bytes (4096.0 GB) in size
-.br
Timestamp: 1009839028
-.br
write complete
-.br
read complete
+.EE
+.RE
+.SH AVAILABILITY
+.B llverdev
+is part of the
+.BR lustre (7)
+filesystem package since release 1.4.0
+.\" Added in commit 1.3.4-1130-g113303973e
.SH SEE ALSO
.BR llverfs (8)
-.TH llverfs 8 "2023 Apr 23" Lustre "configuration utilities"
+.TH LLVERFS 8 2024-08-28 Lustre "Lustre Configuration Utilities"
.SH NAME
llverfs - filesystem verification tool for ldiskfs based on ext3vt
.SH SYNOPSIS
-.BI llverfs
+.SY llverfs
.RB [ -c
.I CHUNKSIZE_MB
.RB ]
.RB [ -h ]
.RB [ -o
.IR OFFSET_KB ]
-.RB [ -l "] [" -p "] [" -r ]
+.RB [ -l ]
+.RB [ -p ]
+.RB [ -r ]
.RB [ -s
.IR SIZE_MB ]
.RB [ -t
.IR TIMESTAMP ]
-.RB [ -v "] [" -w ]
+.RB [ -v ]
+.RB [ -w ]
.I DEVICE
+.YS
.SH DESCRIPTION
This program tests the correct operation of large filesystems and
the underlying block storage device(s).
This tool has two working modes:
-
+.P
In full mode, the program creates a subdirectory in the test
filesystem, writes n (files_in_dir, default=32) large (4GB) files to
the directory with the test pattern at the start of each 4KB block.
to detect data written to the wrong location on disk). This continues
until the whole filesystem is full and then the tool verifies that the data
in all of the test files is correct.
-
+.P
In partial mode, the tool creates test directories with the
EXT2_TOPDIR_FL flag set (if supported) to spread the directory data
around the block device instead of localizing it in a single place.
that the data in each file is correct.
.SH OPTIONS
.TP
-.BR -c | --chunksize " \fICHUNKSIZE_MB"
+.BR -c ", " --chunksize " \fICHUNKSIZE_MB"
IO chunk size in megabytes (default=1), with optional KMG suffix.
.TP
-.BR -h | --help
+.BR -h ", " --help
Display a brief help message.
.TP
-.BR -l | --long | --full
+.BR -l ", " --long | --full
Run a full check, 4GB files with 4k blocks
.TP
-.BR -n | --num_dirs " \fICOUNT"
-\fICOUNT\fR number of sub-directories to create. (\fICOUNT\fR must be >0 and <= INT_MAX)
+.BR -n ", " --num_dirs " \fICOUNT"
+.I COUNT
+number of sub-directories to create.
+.RI ( COUNT
+must be >0 and <= INT_MAX)
.TP
-.BR -o | --offset " \fIOFFSET_KB"
+.BR -o ", " --offset " \fIOFFSET_KB"
offset of IO start in kilobytes (default=0), with optional KMGTP suffix.
.TP
-.BR -p | --partial
+.BR -p ", " --partial
Run a partial check, with 1MB files
.TP
-.BR -r | --read
+.BR -r ", " --read
Run test in read (verify) mode only, after having run the test in
.B -w
mode previously. The flag
.B --write
run.
.TP
-.BR -s | --filesize " \fISIZE_MB"
+.BR -s ", " --filesize " \fISIZE_MB"
File size in megabytes to use for the test, with optional KMGTP
-suffix. The default is 4096MB.
+suffix. The default is 4096MB.
.TP
-.BR -t | --timestamp " \fITIMESTAMP"
+.BR -t ", " --timestamp " \fITIMESTAMP"
Set test start time as printed at the start of a previously interrupted test.
This is to ensure that read phases (or multiple read phases) after a separate
write phase will be able to check the validation data.
(default=current time())
.TP
-.BR -v | --verbose
+.BR -v ", " --verbose
Run test in verbose mode, listing each read and write operation.
.TP
-.BR -w | --write
+.BR -w ", " --write
Run test in write (test-pattern) mode (default run both read and write)
+.SH AVAILABILITY
+.B llverfs
+is part of the
+.BR lustre (7)
+filesystem package since release 1.4.0
+.\" Added in commit 1.3.4-1130-g113303973e
.SH SEE ALSO
.BR llverdev (8)
git://git.whamcloud.com / fs / lustre-release.git / commitdiff