1 .\" Hey, Emacs, edit this file in -*- nroff-fill -*- mode
3 .\" Copyright (c) 1997, 1998
4 .\" Nan Yang Computer Services Limited. All rights reserved.
6 .\" This software is distributed under the so-called ``Berkeley
9 .\" Redistribution and use in source and binary forms, with or without
10 .\" modification, are permitted provided that the following conditions
12 .\" 1. Redistributions of source code must retain the above copyright
13 .\" notice, this list of conditions and the following disclaimer.
14 .\" 2. Redistributions in binary form must reproduce the above copyright
15 .\" notice, this list of conditions and the following disclaimer in the
16 .\" documentation and/or other materials provided with the distribution.
17 .\" 3. All advertising materials mentioning features or use of this software
18 .\" must display the following acknowledgement:
19 .\" This product includes software developed by Nan Yang Computer
21 .\" 4. Neither the name of the Company nor the names of its contributors
22 .\" may be used to endorse or promote products derived from this software
23 .\" without specific prior written permission.
25 .\" This software is provided ``as is'', and any express or implied
26 .\" warranties, including, but not limited to, the implied warranties of
27 .\" merchantability and fitness for a particular purpose are disclaimed.
28 .\" In no event shall the company or contributors be liable for any
29 .\" direct, indirect, incidental, special, exemplary, or consequential
30 .\" damages (including, but not limited to, procurement of substitute
31 .\" goods or services; loss of use, data, or profits; or business
32 .\" interruption) however caused and on any theory of liability, whether
33 .\" in contract, strict liability, or tort (including negligence or
34 .\" otherwise) arising in any way out of the use of this software, even if
35 .\" advised of the possibility of such damage.
37 .\" $Id: vinum.8,v 1.13 2000/01/03 03:12:56 grog Exp grog $
45 .Nd Logical Volume Manager control program
51 .Cd attach Ar plex Ar volume
53 .Cd attach Ar subdisk Ar plex Ar [offset]
56 Attach a plex to a volume, or a subdisk to a plex.
58 .Cd checkparity Ar plex Op Fl f
61 Check the parity blocks of a RAID-4 or RAID-5 plex.
69 Create a concatenated volume from the specified drives.
75 Create a volume as described in
80 Cause the volume manager to enter the kernel debugger.
91 Detach a plex or subdisk from the volume or plex to which it is attached.
96 List the configuration information stored on the specified drives, or all drives
97 in the system if no drive names are specified.
103 List information about volume manager state.
111 Initialize the contents of a subdisk or all the subdisks of a plex to all zeros.
116 Create a volume label
123 .Op volume | plex | subdisk
125 List information about specified objects
132 .Op volume | plex | subdisk
134 List information about specified objects (alternative to
145 List information about drives
154 List information about subdisks
163 List information about plexes
172 List information about volumes
176 Remake the device nodes in
186 Create a mirrored volume from the specified drives.
188 .Cd mv Fl f Ar drive Ar object ...
190 .Cd move Fl f Ar drive Ar object ...
192 Move the object(s) to the specified drive.
197 Write a copy of the current configuration to
204 program when running in interactive mode. Normally this would be done by
214 configuration from the specified disks.
217 .Ar [ drive | subdisk | plex | volume ]
220 Change the name of the specified object.
228 Move all the subdisks from the specified drive onto the new drive.
231 .Cd rebuildparity Ar plex Op Fl f
235 Rebuild the parity blocks of a RAID-4 or RAID-5 plex.
245 .Op volume | plex | subdisk
247 Reset statistisc counters for the specified objects, or for all objects if none
253 .Ar volume | plex | subdisk
261 configuration to disk.
268 .Ar volume | plex | subdisk | disk
270 Set the state of the object to \fIstate\fP\|
276 Set dæmon configuration.
280 .Op Ar volume | plex | subdisk | drive
282 Set state without influencing other objects, for diagnostic purposes only.
286 Read configuration from all vinum drives.
292 volume | plex | subdisk
294 Allow the system to access the objects
298 .Op volume | plex | subdisk
300 Terminate access to the objects, or stop
302 if no parameters are specified.
310 Create a striped volume from the specified drives.
314 is a utility program to communicate with the \fBVinum\fP\| logical volume
317 for more information about the volume manager.
319 is designed either for interactive use, when started without command line
320 arguments, or to execute a single command if the command is supplied on the
321 command line. In interactive mode,
323 maintains a command line history.
326 commands may optionally be followed by an option. Any of the following options
327 may be specified with any command, but in some cases they do not make any
328 difference: cases, the options are ignored. For example, the
341 option overrides safety checks. Use with extreme care. This option is for
342 emergency use only. For example, the command
343 .Bd -unfilled -offset indent
349 even if it is open. Any subsequent access to the volume will almost certainly
351 .It Fl i Ar millisecs
358 milliseconds between copying each block. This lowers the load on the system.
362 option to specify a volume name to the simplified configuration commands
370 .if t (``recursive'')
372 option is used by the list commands to display information not
373 only about the specified objects, but also about subordinate objects. For
374 example, in conjnction with the
378 option will also show information about the plexes and subdisks belonging to the
383 .if t (``statistics'')
385 option is used by the list commands to display statistical information. The
387 command also uses this option to specify that it should create striped plexes.
391 option specifies the transfer size for the
401 option can be used to request more detailed information.
405 .if t (``Very verbose'')
406 .if n ("Very verbose")
407 option can be used to request more detailed information than the
417 to wait for completion of commands which normally run in the background, such as
421 .Ss COMMANDS IN DETAIL
424 commands perform the following functions:
426 .It Nm attach Ar plex Ar volume
430 .It Nm attach Ar subdisk Ar plex Ar [offset]
435 inserts the specified plex or subdisk in a volume or plex. In the case of a
436 subdisk, an offset in the plex may be specified. If it is not, the subdisk will
437 be attached at the first possible location. After attaching a plex to a
440 reintegrates the plex.
446 renames the object (and in the case of a plex, any subordinate subdisks) to fit
451 A number of considerations apply to attaching subdisks:
454 Subdisks can normally only be attached to concatenated plexes.
456 If a striped or RAID-5 plex is missing a subdisk (for example after drive
457 failure), it should be replaced by a subdisk of the same size only.
459 In order to add further subdisks to a striped or RAID-5 plex, use the
461 (force) option. This will corrupt the data in the plex.
462 .\"No other attachment of
463 .\"subdisks is currently allowed for striped and RAID-5 plexes.
465 For concatenated plexes, the
467 parameter specifies the offset in blocks from the beginning of the plex. For
468 striped and RAID-5 plexes, it specifies the offset of the first block of the
469 subdisk: in other words, the offset is the numerical position of the subdisk
470 multiplied by the stripe size. For example, in a plex of block size 256k, the
471 first subdisk will have offset 0, the second offset 256k, the third 512k, etc.
472 This calculation ignores parity blocks in RAID-5 plexes.
479 Check the parity blocks on the specified RAID-4 or RAID-5 plex. This operation
480 maintains a pointer in the plex, so it can be stopped and later restarted from
481 the same position if desired. In addition, this pointer is used by the
483 command, so rebuilding the parity blocks need only start at the location where
484 the first parity problem has been detected.
490 starts checking at the beginning of the plex. If the
494 prints a running progress report.
503 command provides a simplified alternative to the
505 command for creating volumes with a single concatenated plex. The largest
506 contiguous space available on each drive is used to create the subdisks for the
511 command creates an arbitrary name for the volume and its components. The name
512 is composed of the text
514 and a small integer, for example
516 You can override this with the
518 option, which assigns the name specified to the volume. The plexes and subdisks
519 are named after the volume in the default manner.
521 There is no choice of name for the drives. If the drives have already been
524 drives, the name remains. Otherwise the drives are given names starting with
527 and a small integer, for example
533 option can be used to specify that a previous name should be overwritten. The
535 is used to specify verbose output.
537 See the section SIMPLIFIED CONFIGURATION below for some examples of this
539 .It Nm create Op Fl f Ar description-file
543 is used to create any object. In view of the relatively complicated
544 relationship and the potential dangers involved in creating a
546 object, there is no interactive interface to this function. If you do not
549 starts an editor on a temporary file. If the environment variable
553 starts this editor. If not, it defaults to
555 See the section CONFIGURATION FILE below for more information on the format of
561 function is additive: if you run it multiple times, you will create multiple
562 copies of all unnamed objects.
566 command will not change the names of existing
568 drives, in order to avoid accidentally erasing them. The correct way to dispose
571 drives is to reset the configuration with the
573 command. In some cases, however, it may be necessary to create new data on
575 drives which can no longer be started. In this case, use the
582 is used to enter the remote kernel debugger. It is only activated if
586 option. This option will stop the execution of the operating system until the
587 kernel debugger is exited. If remote debugging is set and there is no remote
588 connection for a kernel debugger, it will be necessary to reset the system and
589 reboot in order to leave the debugger.
593 Set a bit mask of internal debugging flags. These will change without warning
594 as the product matures; to be certain, read the header file
595 .Pa sys/dev/vinumvar.h .
596 The bit mask is composed of the following values:
598 .It DEBUG_ADDRESSES (1)
600 Show buffer information during requests
602 .It DEBUG_NUMOUTPUT (2)
611 .It DEBUG_LASTREQS (8)
613 Keep a circular buffer of last requests.
614 .It DEBUG_REVIVECONFLICT (16)
616 Print info about revive conflicts.
617 .It DEBUG_EOFINFO (32)
619 Print information about internal state when returning an EOF on a striped plex.
620 .It DEBUG_MEMFREE (64)
622 Maintain a circular list of the last memory areas freed by the memory allocator.
623 .It DEBUG_REMOTEGDB (256)
630 .It DEBUG_WARNINGS (512)
632 Print some warnings about minor problems in the implementation.
634 .It Nm detach Op Fl f
638 .It Nm detach Op Fl f
643 removes the specified plex or subdisk from the volume or plex to which it is
644 attached. If removing the object would impair the data integrity of the volume,
645 the operation will fail unless the
647 option is specified. If the object is named after the object above it (for
648 example, subdisk vol1.p7.s0 attached to plex vol1.p7), the name will be changed
649 by prepending the text
652 (for example, ex-vol1.p7.s0). If necessary, the name will be truncated in the
656 does not reduce the number of subdisks in a striped or RAID-5 plex. Instead,
657 the subdisk is marked absent, and can later be replaced with the
660 .It Nm dumpconfig Op Ar drive ...
664 shows the configuration information stored on the specified drives. If no drive
667 searches all drives on the system for Vinum partitions and dumps the
668 information. If configuration updates are disabled, it is possible that this
669 information is not the same as the information returned by the
671 command. This command is used primarily for maintenance and debugging.
676 displays information about
678 memory usage. This is intended primarily for debugging. With the
680 option, it will give detailed information about the memory areas in use.
686 displays information about the last up to 64 I/O requests handled by the
688 driver. This information is only collected if debug flag 8 is set. The format
694 Total of 38 blocks malloced, total memory: 16460
695 Maximum allocs: 56, malloc table at 0xf0f72dbc
697 Time Event Buf Dev Offset Bytes SD SDoff Doffset Goffset
699 14:40:00.637758 1VS Write 0xf2361f40 91.3 0x10 16384
700 14:40:00.639280 2LR Write 0xf2361f40 91.3 0x10 16384
701 14:40:00.639294 3RQ Read 0xf2361f40 4.39 0x104109 8192 19 0 0 0
702 14:40:00.639455 3RQ Read 0xf2361f40 4.23 0xd2109 8192 17 0 0 0
703 14:40:00.639529 3RQ Read 0xf2361f40 4.15 0x6e109 8192 16 0 0 0
704 14:40:00.652978 4DN Read 0xf2361f40 4.39 0x104109 8192 19 0 0 0
705 14:40:00.667040 4DN Read 0xf2361f40 4.15 0x6e109 8192 16 0 0 0
706 14:40:00.668556 4DN Read 0xf2361f40 4.23 0xd2109 8192 17 0 0 0
707 14:40:00.669777 6RP Write 0xf2361f40 4.39 0x104109 8192 19 0 0 0
708 14:40:00.685547 4DN Write 0xf2361f40 4.39 0x104109 8192 19 0 0 0
709 11:11:14.975184 Lock 0xc2374210 2 0x1f8001
710 11:11:15.018400 7VS Write 0xc2374210 0x7c0 32768 10
711 11:11:15.018456 8LR Write 0xc2374210 13.39 0xcc0c9 32768
712 11:11:15.046229 Unlock 0xc2374210 2 0x1f8001
717 field always contains the address of the user buffer header. This can be used
718 to identify the requests associated with a user request, though this is not 100%
719 reliable: theoretically two requests in sequence could use the same buffer
720 header, though this is not common. The beginning of a request can be identified
725 The first example above shows the requests involved in a user request. The
726 second is a subdisk I/O request with locking.
730 field contains information related to the sequence of events in the request
735 indicates the approximate sequence of events, and the two-letter abbreviation is
736 a mnemonic for the location
739 (vinumstrategy) shows information about the user request on entry to
741 The device number is the
743 device, and offset and length are the user parameters. This is always the
744 beginning of a request sequence.
746 (launch_requests) shows the user request just prior to launching the low-level
748 requests in the function
749 .Fn launch_requests .
750 The parameters should be the same as in the
754 In the following requests,
756 is the device number of the associated disk partition,
758 is the offset from the beginning of the partition,
760 is the subdisk index in
763 is the offset from the beginning of the subdisk,
765 is the offset of the associated data request, and
767 is the offset of the associated group request, where applicable.
769 (request) shows one of possibly several low-level
771 requests which are launched to satisfy the high-level request. This information
773 .Fn launch_requests .
775 (done) is called from
777 showing the completion of a request. This completion should match a request
778 launched either at stage
781 .Fn launch_requests ,
783 .Fn complete_raid5_write
789 (RAID-5 data) is called from
790 .Fn complete_raid5_write
791 and represents the data written to a RAID-5 data stripe after calculating
794 (RAID-5 parity) is called from
795 .Fn complete_raid5_write
796 and represents the data written to a RAID-5 parity stripe after calculating
799 shows a subdisk I/O request. These requests are usually internal to
801 for operations like initialization or rebuilding plexes.
803 shows the low--level operation generated for a subdisk I/O request.
805 specifies that the process is waiting for a range lock. The parameters are the
806 buffer header associated with the request, the plex number and the block number.
807 For internal reasons the block number is one higher than the address of the
808 beginning of the stripe.
810 specifies that a range lock has been obtained. The parameters are the same as
813 specifies that a range lock has been released. The parameters are the same as
824 initializes a subdisk by writing zeroes to it. You can initialize all subdisks
825 in a plex by specifying the plex name. This is the only way to ensure
826 consistent data in a plex. You must perform this initialization before using a
827 RAID-5 plex. It is also recommended for other new plexes.
829 initializes all subdisks of a plex in parallel. Since this operation can take a
830 long time, it is normally performed in the background. If you want to wait for
831 completion of the command, use the
837 option if you want to write blocks of a different size from the default value of
840 prints a console message when the initialization is complete.
848 style volume label on a volume. It is a simple alternative to an appropriate
851 This is needed because some
853 commands still read the disk to find the label instead of using the correct
857 maintains a volume label separately from the volume data, so this command is not
860 This command is deprecated.
865 .Op volume | plex | subdisk
871 .Op volume | plex | subdisk
906 is used to show information about the specified object. If the argument is
907 omitted, information is shown about all objects known to
911 command is a synonym for
916 option relates to volumes and plexes: if specified, it recursively lists
917 information for the subdisks and (for a volume) plexes subordinate to the
918 objects. The commands
924 commands list only volumes, plexes, subdisks and drives respectively. This is
925 particularly useful when used without parameters.
931 to output device statistics, the
933 (verbose) option causes some additional information to be output, and the
935 causes considerable additional information to be output.
940 command removes the directory /dev/vinum and recreates it with device nodes
941 which reflect the current configuration. This command is not intended for
942 general use, and is provided for emergency use only.
953 command provides a simplified alternative to the
955 command for creating mirrored volumes. Without any options, it creates a RAID-1
956 (mirrored) volume with two concatenated plexes. The largest contiguous space
957 available on each drive is used to create the subdisks for the plexes. The
958 first plex is built from the odd-numbered drives in the list, and the second
959 plex is built from the even-numbered drives. If the drives are of different
960 sizes, the plexes will be of different sizes.
966 builds striped plexes with a stripe size of 256 kB. The size of the subdisks in
967 each plex is the size of the smallest contiguous storage available on any of the
968 drives which form the plex. Again, the plexes may differ in size.
972 command creates an arbitrary name for the volume and its components. The name
973 is composed of the text
975 and a small integer, for example
977 You can override this with the
979 option, which assigns the name specified to the volume. The plexes and subdisks
980 are named after the volume in the default manner.
982 There is no choice of name for the drives. If the drives have already been
985 drives, the name remains. Otherwise the drives are given names starting with
988 and a small integer, for example
994 option can be used to specify that a previous name should be overwritten. The
996 is used to specify verbose output.
998 See the section SIMPLIFIED CONFIGURATION below for some examples of this
1000 .It Nm mv Fl f Ar drive Ar object ...
1001 .It Nm move Fl f Ar drive Ar object ...
1003 Move all the subdisks from the specified objects onto the new drive. The
1004 objects may be subdisks, drives or plexes. When drives or plexes are specified,
1005 all subdisks associated with the object are moved.
1009 option is required for this function, since it currently does not preserve the
1010 data in the subdisk. This functionality will be added at a later date. In this
1011 form, however, it is suited to recovering a failed disk drive.
1012 .It Nm printconfig Op Pa file
1013 Write a copy of the current configuration to
1015 in a format that can be used to recreate the
1017 configuration. Unlike the configuration saved on disk, it includes definitions
1018 of the drives. If you omit
1026 program when running in interactive mode. Normally this would be done by
1035 command scans the specified disks for
1037 partitions containing previously created configuration information. It reads
1038 the configuration in order from the most recently updated to least recently
1039 updated configuration.
1041 maintains an up-to-date copy of all configuration information on each disk
1042 partition. You must specify all of the slices in a configuration as the
1043 parameter to this command.
1047 command is intended to selectively load a
1049 configuration on a system which has other
1051 partitions. If you want to start all partitions on the system, it is easier to
1058 encounters any errors during this command, it will turn off automatic
1059 configuration update to avoid corrupting the copies on disk. This will also
1060 happen if the configuration on disk indicates a configuration error (for
1061 example, subdisks which do not have a valid space specification). You can turn
1062 the updates on again with the
1066 commands. Reset bit 2 (numerical value 4) of the daemon options mask to
1067 re-enable configuration saves.
1068 .It Nm rebuildparity
1074 Rebuild the parity blocks on the specified RAID-4 or RAID-5 plex. This
1075 operation maintains a pointer in the plex, so it can be stopped and later
1076 restarted from the same position if desired. In addition, this pointer is used
1079 command, so rebuilding the parity blocks need only start at the location where
1080 the first parity problem has been detected.
1086 starts rebuilding at the beginning of the plex. If the
1090 first checks the existing parity blocks prints information about those found to
1091 be incorrect before rebuilding. If the
1095 prints a running progress report.
1098 .Ar [ drive | subdisk | plex | volume ]
1101 Change the name of the specified object. If the
1103 option is specified, subordinate objects will be named by the default rules:
1104 plex names will be formed by appending .p\f(BInumber\fP to the volume name, and
1105 subdisk names will be formed by appending .s\f(BInumber\fP to the plex name.
1111 Move all the subdisks from the specified drive onto the new drive. This will
1112 attempt to recover those subdisks that can be recovered, and create the others
1113 from scratch. If the new drive lacks the space for this operation, as many
1114 subdisks as possible will be fitted onto the drive, and the rest will be left on
1121 command completely obliterates the
1123 configuration on a system. Use this command only when you want to completely
1124 delete the configuration.
1126 will ask for confirmation: you must type in the words NO FUTURE exactly
1128 .Bd -unfilled -offset indent
1129 # \f(CBvinum resetconfig\f(CW
1131 WARNING! This command will completely wipe out your vinum
1132 configuration. All data will be lost. If you really want
1133 to do this, enter the text
1136 Enter text -> \f(BINO FUTURE\fP
1137 Vinum configuration obliterated
1141 As the message suggests, this is a last-ditch command. Don't use it unless you
1142 have an existing configuration which you never want to see again.
1145 .Op volume | plex | subdisk
1148 maintains a number of statistical counters for each object. See the header file
1150 for more information.
1151 .\" XXX put it in here when it's finalized
1154 command to reset these counters. In conjunction with the
1158 also resets the counters of subordinate objects.
1162 .Ar volume | plex | subdisk
1165 removes an object from the
1167 configuration. Once an object has been removed, there is no way to recover it.
1170 performs a large amount of consistency checking before removing an object. The
1174 to omit this checking and remove the object anyway. Use this option with great
1175 care: it can result in total loss of data on a volume.
1179 refuses to remove a volume or plex if it has subordinate plexes or subdisks
1180 respectively. You can tell
1182 to remove the object anyway by using the
1184 option, or you can cause
1186 to remove the subordinate objects as well by using the
1188 (recursive) option. If you remove a volume with the
1190 option, it will remove both the plexes and the subdisks which belong to the
1194 Save the current configuration to disk. This is primarily a maintenance
1195 function. For example, if an error occurs on startup, updates will be
1196 disabled. When you reenable them, the configuration is not automatically saved
1197 to disk. Use this command to save the configuration.
1202 .Ar volume | plex | subdisk | disk
1204 sets the state of the specified object to one of the valid states (see OBJECT
1205 STATES below). Normally
1207 performs a large amount of consistency checking before making the change. The
1211 to omit this checking and perform the change anyway. Use this option with great
1212 care: it can result in total loss of data on a volume.
1214 .Nm This command has not yet been implemented.
1220 sets a variable bitmask for the
1222 dæmon. This command is temporary and will be replaced. Currently, the bit mask
1223 may contain the bits 1 (log every action to syslog) and 4 (don't update
1224 configuration). Option bit 4 can be useful for error recovery.
1227 .Op Ar volume | plex | subdisk | drive
1230 sets the state of the specified objects to the specified state. This bypasses
1231 the usual consistency mechanism of
1233 and should be used only for recovery purposes. It is possible to crash the
1234 system by incorrect use of this command.
1236 .Op Fl i Ar interval
1242 starts (brings into to the
1248 If no object names are specified,
1250 scans the disks known to the system for
1252 drives and then reads in the configuration as described under the
1256 drive contains a header with all information about the data stored on the drive,
1257 including the names of the other drives which are required in order to represent
1262 encounters any errors during this command, it will turn off automatic
1263 configuration update to avoid corrupting the copies on disk. This will also
1264 happen if the configuration on disk indicates a configuration error (for
1265 example, subdisks which do not have a valid space specification). You can turn
1266 the updates on again with the
1270 command. Reset bit 4 of the daemon options mask to re-enable configuration
1273 If object names are specified,
1275 starts them. Normally this operation is only of use with subdisks. The action
1276 depends on the current state of the object:
1279 If the object is already in the
1285 If the object is a subdisk in the
1295 If the object is a subdisk in the
1297 state, the change depends on the subdisk. If it is part of a plex which is part
1298 of a volume which contains other plexes,
1300 places the subdisk in the
1302 state and attempts to copy the data from the volume. When the operation
1303 completes, the subdisk is set into the
1305 state. If it is part of a plex which is part of a volume which contains no
1306 other plexes, or if it is not part of a plex,
1312 If the object is a subdisk in the
1318 operation offline. When the operation completes, the subdisk is set into the
1323 When a subdisk comes into the
1327 automatically checks the state of any plex and volume to which it may belong and
1328 changes their state where appropriate.
1330 If the object is a plex,
1332 checks the state of the subordinate subdisks (and plexes in the case of a
1333 volume) and starts any subdisks which can be started.
1335 To start a plex in a multi-plex volume, the data must be copied from another
1336 plex in the volume. Since this frequently takes a long time, it is normally
1337 done in the background. If you want to wait for this operation to complete (for
1338 example, if you are performing this operation in a script), use the
1342 Copying data doesn't just take a long time, it can also place a significant load
1343 on the system. You can specify the transfer size in bytes or sectors with the
1345 option, and an interval (in milliseconds) to wait between copying each block with
1348 option. Both of these options lessen the load on the system.
1351 .Op volume | plex | subdisk
1353 If no parameters are specified,
1359 This can only be done if no objects are active. In particular, the
1361 option does not override this requirement. Normally, the
1363 command writes the current configuration back to the drives before terminating.
1364 This will not be possible if configuration updates are disabled, so
1366 will not stop if configuration updates are disabled. You can override this by
1373 command can only work if
1375 has been loaded as a kld, since it is not possible to unload a statically
1381 is statically configured.
1383 If object names are specified,
1385 disables access to the objects. If the objects have subordinate objects, they
1386 subordinate objects must either already be inactive (stopped or in error), or
1391 options must be specified. This command does not remove the objects from the
1392 configuration. They can be accessed again after a
1398 does not stop active objects. For example, you cannot stop a plex which is
1399 attached to an active volume, and you cannot stop a volume which is open. The
1403 to omit this checking and remove the object anyway. Use this option with great
1404 care and understanding: used incorrectly, it can result in serious data
1414 command provides a simplified alternative to the
1416 command for creating volumes with a single striped plex. The size of the
1417 subdisks is the size of the largest contiguous space available on all the
1418 specified drives. The stripe size is fixed at 256 kB.
1422 command creates an arbitrary name for the volume and its components. The name
1423 is composed of the text
1425 and a small integer, for example
1427 You can override this with the
1429 option, which assigns the name specified to the volume. The plexes and subdisks
1430 are named after the volume in the default manner.
1432 There is no choice of name for the drives. If the drives have already been
1435 drives, the name remains. Otherwise the drives are given names starting with
1438 and a small integer, for example
1444 option can be used to specify that a previous name should be overwritten. The
1446 is used to specify verbose output.
1448 See the section SIMPLIFIED CONFIGURATION below for some examples of this
1451 .Sh SIMPLIFIED CONFIGURATION
1452 This section describes a simplified interface to
1454 configuration using the
1459 commands. These commands create convenient configurations for some more normal
1460 situations, but they are not as flexible as the
1464 See above for the description of the commands. Here are some examples, all
1465 performed with the same collection of disks. Note that the first drive,
1467 is smaller than the others. This has an effect on the sizes chosen for each
1470 The following examples all use the
1472 option to show the commands passed to the system, and also to list the structure
1473 of the volume. Without the
1475 option, these commands produce no output.
1476 .Ss Volume with a single concatenated plex
1477 Use a volume with a single concatenated plex for the largest possible storage
1478 without resilience to drive failures:
1480 vinum -> concat -v /dev/da1h /dev/da2h /dev/da3h /dev/da4h
1482 plex name vinum0.p0 org concat
1483 drive vinumdrive0 device /dev/da1h
1484 sd name vinum0.p0.s0 drive vinumdrive0 size 0
1485 drive vinumdrive1 device /dev/da2h
1486 sd name vinum0.p0.s1 drive vinumdrive1 size 0
1487 drive vinumdrive2 device /dev/da3h
1488 sd name vinum0.p0.s2 drive vinumdrive2 size 0
1489 drive vinumdrive3 device /dev/da4h
1490 sd name vinum0.p0.s3 drive vinumdrive3 size 0
1491 V vinum0 State: up Plexes: 1 Size: 2134 MB
1492 P vinum0.p0 C State: up Subdisks: 4 Size: 2134 MB
1493 S vinum0.p0.s0 State: up PO: 0 B Size: 414 MB
1494 S vinum0.p0.s1 State: up PO: 414 MB Size: 573 MB
1495 S vinum0.p0.s2 State: up PO: 988 MB Size: 573 MB
1496 S vinum0.p0.s3 State: up PO: 1561 MB Size: 573 MB
1499 In this case, the complete space on all four disks was used, giving a volume
1501 .Ss Volume with a single striped plex
1502 A volume with a single striped plex may give better performance than a
1503 concatenated plex, but restrictions on striped plexes can mean that the volume
1504 is smaller. It will also not be resilient to a drive failure:
1506 vinum -> stripe -v /dev/da1h /dev/da2h /dev/da3h /dev/da4h
1507 drive vinumdrive0 device /dev/da1h
1508 drive vinumdrive1 device /dev/da2h
1509 drive vinumdrive2 device /dev/da3h
1510 drive vinumdrive3 device /dev/da4h
1512 plex name vinum0.p0 org striped 256k
1513 sd name vinum0.p0.s0 drive vinumdrive0 size 849825b
1514 sd name vinum0.p0.s1 drive vinumdrive1 size 849825b
1515 sd name vinum0.p0.s2 drive vinumdrive2 size 849825b
1516 sd name vinum0.p0.s3 drive vinumdrive3 size 849825b
1517 V vinum0 State: up Plexes: 1 Size: 1659 MB
1518 P vinum0.p0 S State: up Subdisks: 4 Size: 1659 MB
1519 S vinum0.p0.s0 State: up PO: 0 B Size: 414 MB
1520 S vinum0.p0.s1 State: up PO: 256 kB Size: 414 MB
1521 S vinum0.p0.s2 State: up PO: 512 kB Size: 414 MB
1522 S vinum0.p0.s3 State: up PO: 768 kB Size: 414 MB
1525 In this case, the size of the subdisks has been limited to the smallest
1526 available disk, so the resulting volume is only 1659 MB in size.
1527 .Ss Mirrored volume with two concatenated plexes
1528 For more reliability, use a mirrored, concatenated volume:
1530 vinum -> mirror -v -n mirror /dev/da1h /dev/da2h /dev/da3h /dev/da4h
1531 drive vinumdrive0 device /dev/da1h
1532 drive vinumdrive1 device /dev/da2h
1533 drive vinumdrive2 device /dev/da3h
1534 drive vinumdrive3 device /dev/da4h
1535 volume mirror setupstate
1536 plex name mirror.p0 org concat
1537 sd name mirror.p0.s0 drive vinumdrive0 size 0b
1538 sd name mirror.p0.s1 drive vinumdrive2 size 0b
1539 plex name mirror.p1 org concat
1540 sd name mirror.p1.s0 drive vinumdrive1 size 0b
1541 sd name mirror.p1.s1 drive vinumdrive3 size 0b
1542 V mirror State: up Plexes: 2 Size: 1146 MB
1543 P mirror.p0 C State: up Subdisks: 2 Size: 988 MB
1544 P mirror.p1 C State: up Subdisks: 2 Size: 1146 MB
1545 S mirror.p0.s0 State: up PO: 0 B Size: 414 MB
1546 S mirror.p0.s1 State: up PO: 414 MB Size: 573 MB
1547 S mirror.p1.s0 State: up PO: 0 B Size: 573 MB
1548 S mirror.p1.s1 State: up PO: 573 MB Size: 573 MB
1551 This example specifies the name of the volume:
1553 Since one drive is smaller than the others, the two plexes are of different
1554 size, and the last 158 MB of the volume is non-resilient. To ensure complete
1555 reliability in such a situation, use the
1557 command to create a volume with 988 MB.
1558 .Ss Mirrored volume with two striped plexes
1559 Alternatively, use the
1561 option to create a mirrored volume with two striped plexes:
1563 vinum -> mirror -v -n raid10 -s /dev/da1h /dev/da2h /dev/da3h /dev/da4h
1564 drive vinumdrive0 device /dev/da1h
1565 drive vinumdrive1 device /dev/da2h
1566 drive vinumdrive2 device /dev/da3h
1567 drive vinumdrive3 device /dev/da4h
1568 volume raid10 setupstate
1569 plex name raid10.p0 org striped 256k
1570 sd name raid10.p0.s0 drive vinumdrive0 size 849825b
1571 sd name raid10.p0.s1 drive vinumdrive2 size 849825b
1572 plex name raid10.p1 org striped 256k
1573 sd name raid10.p1.s0 drive vinumdrive1 size 1173665b
1574 sd name raid10.p1.s1 drive vinumdrive3 size 1173665b
1575 V raid10 State: up Plexes: 2 Size: 1146 MB
1576 P raid10.p0 S State: up Subdisks: 2 Size: 829 MB
1577 P raid10.p1 S State: up Subdisks: 2 Size: 1146 MB
1578 S raid10.p0.s0 State: up PO: 0 B Size: 414 MB
1579 S raid10.p0.s1 State: up PO: 256 kB Size: 414 MB
1580 S raid10.p1.s0 State: up PO: 0 B Size: 573 MB
1581 S raid10.p1.s1 State: up PO: 256 kB Size: 573 MB
1584 In this case, the usable part of the volume is even smaller, since the first
1585 plex has shrunken to match the smallest drive.
1586 .Ss CONFIGURATION FILE
1588 requires that all parameters to the
1590 commands must be in a configuration file. Entries in the configuration file
1591 define volumes, plexes and subdisks, and may be in free format, except that each
1592 entry must be on a single line.
1595 Some configuration file parameters specify a size (lengths, stripe sizes).
1596 These values can be specified as bytes, or one of the following scale factors
1600 specifies that the value is a number of sectors of 512 bytes.
1602 specifies that the value is a number of kilobytes (1024 bytes).
1604 specifies that the value is a number of megabytes (1048576 bytes).
1606 specifies that the value is a number of gigabytes (1073741824 bytes).
1608 is used for compatibility with VERITAS. It stands for blocks of 512 bytes.
1609 This abbreviation is confusing, since the word ``block'' is used in different
1610 meanings, and its use is deprecated.
1613 For example, the value 16777216 bytes can also be written as
1619 The configuration file can contain the following entries:
1622 .It Nm drive Ar name devicename
1625 Define a drive. The options are:
1627 .Bl -hang -width 18n
1628 .It Nm device Ar devicename
1629 Specify the device on which the drive resides.
1631 must be the name of a disk partition, for example
1635 and it must be of type
1639 partition, which is reserved for the complete disk.
1641 Define the drive to be a
1645 drive, which is maintained to automatically replace a failed drive.
1647 does not allow this drive to be used for any other purpose. In particular, it
1648 is not possible to create subdisks on it. This functionality has not been
1649 completely implemented.
1655 Define a volume with name
1660 .Bl -hang -width 18n
1661 .It Nm plex Ar plexname
1662 Add the specified plex to the volume. If
1667 will look for the definition of the plex as the next possible entry in the
1668 configuration file after the definition of the volume.
1669 .It Nm readpol Ar policy
1677 .Nm prefer Ar plexname .
1679 satisfies a read request from only one of the plexes. A
1681 read policy specifies that each read should be performed from a different plex
1682 in \fIround-robin\fR\| fashion. A
1684 read policy reads from the specified plex every time.
1687 When creating a multi-plex volume, assume that the contents of all the plexes
1688 are consistent. This is normally not the case, so by default
1690 sets all plexes except the first one to the
1694 command to first bring them to a consistent state. In the case of striped and
1695 concatenated plexes, however, it does not normally cause problems to leave them
1696 inconsistent: when using a volume for a file system or a swap partition, the
1697 previous contents of the disks are not of interest, so they may be ignored.
1698 If you want to take this risk, use the
1700 keyword. It will only apply to the plexes defined immediately after the volume
1701 in the configuration file. If you add plexes to a volume at a later time, you
1702 must integrate them manually with the
1706 Note that you \fImust\fP\| use the
1708 command with RAID-5 plexes: otherwise extreme data corruption will result if one
1712 .It Nm plex Op options
1714 Define a plex. Unlike a volume, a plex does not need a name. The options may
1717 .Bl -hang -width 18n
1718 .It Nm name Ar plexname
1719 Specify the name of the plex. Note that you must use the keyword
1721 when naming a plex or subdisk.
1723 .It Nm org Ar organization Op stripesize
1725 Specify the organization of the plex.
1736 plexes, the parameter
1738 must be specified, while for
1740 it must be omitted. For type
1742 it specifies the width of each stripe. For type
1744 it specifies the size of a group. A group is a portion of a plex which
1745 stores the parity bits all in the same subdisk. It must be a factor of the plex size (in
1746 other words, the result of dividing the plex size by the stripe size must be an
1747 integer), and it must be a multiple of a disk sector (512 bytes).
1749 For optimum performance, stripes should be at least 128 kB in size: anything
1750 smaller will result in a significant increase in I/O activity due to mapping of
1751 individual requests over multiple disks. The performance improvement due to the
1752 increased number of concurrent transfers caused by this mapping will not make up
1753 for the performance drop due to the increase in latency. A good guideline for
1754 stripe size is between 256 kB and 512 kB. Avoid powers of 2, however: they tend
1755 to cause all superblocks to be placed on the first subdisk.
1757 A striped plex must have at least two subdisks (otherwise it is a concatenated
1758 plex), and each must be the same size. A RAID-5 plex must have at least three
1759 subdisks, and each must be the same size. In practice, a RAID-5 plex should
1760 have at least 5 subdisks.
1762 .It Nm volume Ar volname
1763 Add the plex to the specified volume. If no
1765 keyword is specified, the plex will be added to the last volume mentioned in the
1768 .It Nm sd Ar sdname Ar offset
1769 Add the specified subdisk to the plex at offset
1774 .It Nm subdisk Op options
1776 Define a subdisk. Options may be:
1778 .Bl -hang -width 18n
1782 Specify the name of a subdisk. It is not necessary to specify a name for a
1783 subdisk\(emsee OBJECT NAMING above. Note that you must specify the keyword
1785 if you wish to name a subdisk.
1787 .It Nm plexoffset Ar offset
1788 Specify the starting offset of the subdisk in the plex. If not specified,
1790 allocates the space immediately after the previous subdisk, if any, or otherwise
1791 at the beginning of the plex.
1793 .It Nm driveoffset Ar offset
1794 Specify the starting offset of the subdisk in the drive. If not specified,
1796 allocates the first contiguous
1798 bytes of free space on the drive.
1800 .It Nm length Ar length
1801 Specify the length of the subdisk. This keyword must be specified. There is no
1802 default, but the value 0 may be specified to mean
1803 .if t ``use the largest available contiguous free area on the drive''.
1804 .if n "use the largest available contiguous free area on the drive".
1805 If the drive is empty, this means that the entire drive will be used for the
1812 Specify the plex to which the subdisk belongs. By default, the subdisk belongs
1813 to the last plex specified.
1815 .It Nm drive Ar drive
1816 Specify the drive on which the subdisk resides. By default, the subdisk resides
1817 on the last drive specified.
1822 .Sh EXAMPLE CONFIGURATION FILE
1824 # Sample vinum configuration file
1827 drive drive1 device /dev/da1h
1828 drive drive2 device /dev/da2h
1829 drive drive3 device /dev/da3h
1830 drive drive4 device /dev/da4h
1831 drive drive5 device /dev/da5h
1832 drive drive6 device /dev/da6h
1833 # A volume with one striped plex
1835 plex org striped 512b
1836 sd length 64m drive drive2
1837 sd length 64m drive drive4
1839 plex org striped 512b
1840 sd length 512m drive drive2
1841 sd length 512m drive drive4
1845 sd length 100m drive drive2
1846 sd length 50m drive drive4
1848 sd length 150m drive drive4
1849 # A volume with one striped plex and one concatenated plex
1851 plex org striped 512b
1852 sd length 100m drive drive2
1853 sd length 100m drive drive4
1855 sd length 150m drive drive2
1856 sd length 50m drive drive4
1857 # a volume with a RAID-5 and a striped plex
1858 # note that the RAID-5 volume is longer by
1859 # the length of one subdisk
1861 plex org striped 64k
1862 sd length 1000m drive drive2
1863 sd length 1000m drive drive4
1865 sd length 500m drive drive1
1866 sd length 500m drive drive2
1867 sd length 500m drive drive3
1868 sd length 500m drive drive4
1869 sd length 500m drive drive5
1871 .Ss DRIVE LAYOUT CONSIDERATIONS
1873 drives are currently BSD disk partitions. They must be of type
1875 in order to avoid overwriting data used for other purposes. Use
1878 to edit a partition type definition. The following display shows a typical
1879 partition layout as shown by
1883 # size offset fstype [fsize bsize bps/cpg]
1884 a: 81920 344064 4.2BSD 0 0 0 # (Cyl. 240*- 297*)
1885 b: 262144 81920 swap # (Cyl. 57*- 240*)
1886 c: 4226725 0 unused 0 0 # (Cyl. 0 - 2955*)
1887 e: 81920 0 4.2BSD 0 0 0 # (Cyl. 0 - 57*)
1888 f: 1900000 425984 4.2BSD 0 0 0 # (Cyl. 297*- 1626*)
1889 g: 1900741 2325984 vinum 0 0 0 # (Cyl. 1626*- 2955*)
1892 In this example, partition
1896 partition. Partitions
1905 partitions. Partition
1907 is a swap partition, and partition
1909 represents the whole disk and should not be used for any other purpose.
1912 uses the first 265 sectors on each partition for configuration information, so
1913 the maximum size of a subdisk is 265 sectors smaller than the drive.
1916 maintains a log file, by default
1917 .Pa /var/log/vinum_history ,
1918 in which it keeps track of the commands issued to
1920 You can override the name of this file by setting the environment variable
1922 to the name of the file.
1924 Each message in the log file is preceded by a date. The default format is
1925 .Li %e %b %Y %H:%M:%S
1928 for further details of the format string. It can be overridden by the
1929 environment variable
1930 .Ev VINUM_DATEFORMAT .
1931 .Sh HOW TO SET UP VINUM
1932 This section gives practical advice about how to implement a
1935 .Ss Where to put the data
1936 The first choice you need to make is where to put the data. You need dedicated
1939 They should be partitions, not devices, and they should not be partition
1941 For example, good names are
1949 both of which represent a device, not a partition, and
1951 which represents a complete disk and should be of type
1953 See the example under DRIVE LAYOUT CONSIDERATIONS above.
1954 .Ss Designing volumes
1957 volumes depends on your intentions. There are a number of possibilities:
1960 You may want to join up a number of small disks to make a reasonable sized file
1961 system. For example, if you had five small drives and wanted to use all the
1962 space for a single volume, you might write a configuration file like:
1963 .Bd -literal -offset 4n
1964 drive d1 device /dev/da2e
1965 drive d2 device /dev/da3e
1966 drive d3 device /dev/da4e
1967 drive d4 device /dev/da5e
1968 drive d5 device /dev/da6e
1971 sd length 0 drive d1
1972 sd length 0 drive d2
1973 sd length 0 drive d3
1974 sd length 0 drive d4
1975 sd length 0 drive d5
1978 In this case, you specify the length of the subdisks as 0, which means
1979 .if t ``use the largest area of free space that you can find on the drive''.
1980 .if n "use the largest area of free space that you can find on the drive".
1981 If the subdisk is the only subdisk on the drive, it will use all available
1986 to obtain additional resilience against disk failures. You have the choice of
1988 .if t ``mirroring'', or RAID-5, also called ``parity''.
1989 .if n "mirroring", or RAID-5, also called "parity".
1991 To set up mirroring, create multiple plexes in a volume. For example, to create
1992 a mirrored volume of 2 GB, you might create the following configuration file:
1993 .Bd -literal -offset 4n
1994 drive d1 device /dev/da2e
1995 drive d2 device /dev/da3e
1998 sd length 2g drive d1
2000 sd length 2g drive d2
2003 When creating mirrored drives, it is important to ensure that the data from each
2004 plex is on a different physical disk so that
2006 can access the complete address space of the volume even if a drive fails.
2007 Note that each plex requires as much data as the complete volume: in this
2008 example, the volume has a size of 2 GB, but each plex (and each subdisk)
2009 requires 2 GB, so the total disk storage requirement is 4 GB.
2011 To set up RAID-5, create a single plex of type
2013 For example, to create an equivalent resilient volume of 2 GB, you might use the
2014 following configuration file:
2015 .Bd -literal -offset 4n
2016 drive d1 device /dev/da2e
2017 drive d2 device /dev/da3e
2018 drive d3 device /dev/da4e
2019 drive d4 device /dev/da5e
2020 drive d5 device /dev/da6e
2023 sd length 512m drive d1
2024 sd length 512m drive d2
2025 sd length 512m drive d3
2026 sd length 512m drive d4
2027 sd length 512m drive d5
2030 RAID-5 plexes require at least three subdisks, one of which is used for storing
2031 parity information and is lost for data storage. The more disks you use, the
2032 greater the proportion of the disk storage can be used for data storage. In
2033 this example, the total storage usage is 2.5 GB, compared to 4 GB for a mirrored
2034 configuration. If you were to use the minimum of only three disks, you would
2035 require 3 GB to store the information, for example:
2036 .Bd -literal -offset 4n
2037 drive d1 device /dev/da2e
2038 drive d2 device /dev/da3e
2039 drive d3 device /dev/da4e
2042 sd length 1g drive d1
2043 sd length 1g drive d2
2044 sd length 1g drive d3
2047 As with creating mirrored drives, it is important to ensure that the data from
2048 each subdisk is on a different physical disk so that
2050 can access the complete address space of the volume even if a drive fails.
2054 to allow more concurrent access to a file system. In many cases, access to a
2055 file system is limited by the speed of the disk. By spreading the volume across
2056 multiple disks, you can increase the throughput in multi-access environments.
2057 This technique shows little or no performance improvement in single-access
2060 uses a technique called
2063 or sometimes RAID-0, to increase this concurrency of access. The name RAID-0 is
2064 misleading: striping does not provide any redundancy or additional reliability.
2065 In fact, it decreases the reliability, since the failure of a single disk will
2066 render the volume useless, and the more disks you have, the more likely it is
2067 that one of them will fail.
2069 To implement striping, use a
2072 .Bd -literal -offset 4n
2073 drive d1 device /dev/da2e
2074 drive d2 device /dev/da3e
2075 drive d3 device /dev/da4e
2076 drive d4 device /dev/da5e
2078 plex org striped 512k
2079 sd length 512m drive d1
2080 sd length 512m drive d2
2081 sd length 512m drive d3
2082 sd length 512m drive d4
2085 A striped plex must have at least two subdisks, but the increase in performance
2086 is greater if you have a larger number of disks.
2088 You may want to have the best of both worlds and have both resilience and
2089 performance. This is sometimes called RAID-10 (a combination of RAID-1 and
2090 RAID-0), though again this name is misleading. With
2092 you can do this with the following configuration file:
2093 .Bd -literal -offset 4n
2094 drive d1 device /dev/da2e
2095 drive d2 device /dev/da3e
2096 drive d3 device /dev/da4e
2097 drive d4 device /dev/da5e
2098 volume raid setupstate
2099 plex org striped 512k
2100 sd length 512m drive d1
2101 sd length 512m drive d2
2102 sd length 512m drive d3
2103 sd length 512m drive d4
2104 plex org striped 512k
2105 sd length 512m drive d4
2106 sd length 512m drive d3
2107 sd length 512m drive d2
2108 sd length 512m drive d1
2111 Here the plexes are striped, increasing performance, and there are two of them,
2112 increasing reliablity. Note that this example shows the subdisks of the second
2113 plex in reverse order from the first plex. This is for performance reasons and
2114 will be discussed below. In addition, the volume specification includes the
2117 which ensures that all plexes are
2121 .Ss Creating the volumes
2122 Once you have created your configuration files, start
2124 and create the volumes. In this example, the configuration is in the file
2127 # vinum create -v configfile
2128 1: drive d1 device /dev/da2e
2129 2: drive d2 device /dev/da3e
2132 5: sd length 2g drive d1
2134 7: sd length 2g drive d2
2135 Configuration summary
2137 Drives: 2 (4 configured)
2138 Volumes: 1 (4 configured)
2139 Plexes: 2 (8 configured)
2140 Subdisks: 2 (16 configured)
2142 Drive d1: Device /dev/da2e
2143 Created on vinum.lemis.com at Tue Mar 23 12:30:31 1999
2144 Config last updated Tue Mar 23 14:30:32 1999
2145 Size: 60105216000 bytes (57320 MB)
2146 Used: 2147619328 bytes (2048 MB)
2147 Available: 57957596672 bytes (55272 MB)
2150 Drive d2: Device /dev/da3e
2151 Created on vinum.lemis.com at Tue Mar 23 12:30:32 1999
2152 Config last updated Tue Mar 23 14:30:33 1999
2153 Size: 60105216000 bytes (57320 MB)
2154 Used: 2147619328 bytes (2048 MB)
2155 Available: 57957596672 bytes (55272 MB)
2159 Volume mirror: Size: 2147483648 bytes (2048 MB)
2163 Read policy: round robin
2165 Plex mirror.p0: Size: 2147483648 bytes (2048 MB)
2168 Organization: concat
2169 Part of volume mirror
2170 Plex mirror.p1: Size: 2147483648 bytes (2048 MB)
2173 Organization: concat
2174 Part of volume mirror
2176 Subdisk mirror.p0.s0:
2177 Size: 2147483648 bytes (2048 MB)
2179 Plex mirror.p0 at offset 0
2181 Subdisk mirror.p1.s0:
2182 Size: 2147483648 bytes (2048 MB)
2184 Plex mirror.p1 at offset 0
2191 to list the file as it configures. Subsequently it lists the current
2192 configuration in the same format as the
2195 .Ss Creating more volumes
2196 Once you have created the
2200 keeps track of them in its internal configuration files. You do not need to
2201 create them again. In particular, if you run the
2203 command again, you will create additional objects:
2206 # vinum create sampleconfig
2207 Configuration summary
2209 Drives: 2 (4 configured)
2210 Volumes: 1 (4 configured)
2211 Plexes: 4 (8 configured)
2212 Subdisks: 4 (16 configured)
2214 D d1 State: up Device /dev/da2e Avail: 53224/57320 MB (92%)
2215 D d2 State: up Device /dev/da3e Avail: 53224/57320 MB (92%)
2217 V mirror State: up Plexes: 4 Size: 2048 MB
2219 P mirror.p0 C State: up Subdisks: 1 Size: 2048 MB
2220 P mirror.p1 C State: up Subdisks: 1 Size: 2048 MB
2221 P mirror.p2 C State: up Subdisks: 1 Size: 2048 MB
2222 P mirror.p3 C State: up Subdisks: 1 Size: 2048 MB
2224 S mirror.p0.s0 State: up PO: 0 B Size: 2048 MB
2225 S mirror.p1.s0 State: up PO: 0 B Size: 2048 MB
2226 S mirror.p2.s0 State: up PO: 0 B Size: 2048 MB
2227 S mirror.p3.s0 State: up PO: 0 B Size: 2048 MB
2231 As this example (this time with the
2233 option) shows, re-running the
2235 has created four new plexes, each with a new subdisk. If you want to add other
2236 volumes, create new configuration files for them. They do not need to reference
2239 already knows about. For example, to create a volume
2247 you only need to mention the other two:
2249 drive d3 device /dev/da1e
2250 drive d4 device /dev/da4e
2259 With this configuration file, we get:
2261 # vinum create newconfig
2262 Configuration summary
2264 Drives: 4 (4 configured)
2265 Volumes: 2 (4 configured)
2266 Plexes: 5 (8 configured)
2267 Subdisks: 8 (16 configured)
2269 D d1 State: up Device /dev/da2e Avail: 51176/57320 MB (89%)
2270 D d2 State: up Device /dev/da3e Avail: 53220/57320 MB (89%)
2271 D d3 State: up Device /dev/da1e Avail: 53224/57320 MB (92%)
2272 D d4 State: up Device /dev/da4e Avail: 53224/57320 MB (92%)
2274 V mirror State: down Plexes: 4 Size: 2048 MB
2275 V raid State: down Plexes: 1 Size: 6144 MB
2277 P mirror.p0 C State: init Subdisks: 1 Size: 2048 MB
2278 P mirror.p1 C State: init Subdisks: 1 Size: 2048 MB
2279 P mirror.p2 C State: init Subdisks: 1 Size: 2048 MB
2280 P mirror.p3 C State: init Subdisks: 1 Size: 2048 MB
2281 P raid.p0 R5 State: init Subdisks: 4 Size: 6144 MB
2283 S mirror.p0.s0 State: up PO: 0 B Size: 2048 MB
2284 S mirror.p1.s0 State: up PO: 0 B Size: 2048 MB
2285 S mirror.p2.s0 State: up PO: 0 B Size: 2048 MB
2286 S mirror.p3.s0 State: up PO: 0 B Size: 2048 MB
2287 S raid.p0.s0 State: empty PO: 0 B Size: 2048 MB
2288 S raid.p0.s1 State: empty PO: 512 kB Size: 2048 MB
2289 S raid.p0.s2 State: empty PO: 1024 kB Size: 2048 MB
2290 S raid.p0.s3 State: empty PO: 1536 kB Size: 2048 MB
2293 Note the size of the RAID-5 plex: it is only 6 GB, although together its
2294 components use 8 GB of disk space. This is because the equivalent of one
2295 subdisk is used for storing parity data.
2296 .Ss Restarting Vinum
2297 On rebooting the system, start
2306 This will start all the
2308 drives in the system. If for some reason you wish to start only some of them,
2312 .Ss Performance considerations
2313 A number of misconceptions exist about how to set up a RAID array for best
2314 performance. In particular, most systems use far too small a stripe size. The
2315 following discussion applies to all RAID systems, not just to
2318 The FreeBSD block I/O system issues requests of between .5kB and 128 kB; a
2319 typical mix is somewhere round 8 kB. You can't stop any striping system from
2320 breaking a request into two physical requests, and if you make the stripe small
2321 enough, it can be broken into several. This will result in a significant drop
2322 in performance: the decrease in transfer time per disk is offset by the order of
2323 magnitude greater increase in latency.
2325 With modern disk sizes and the FreeBSD I/O system, you can expect to have a
2326 reasonably small number of fragmented requests with a stripe size between 256 kB
2327 and 512 kB; with correct RAID implementations there is no obvious reason not to
2328 increase the size to 2 or 4 MB on a large disk.
2330 When choosing a stripe size, consider that most current ufs file systems have
2331 cylinder groups 32 MB in size. If you have a stripe size and number of disks
2332 both of which are a power of two, it is probable that all superblocks and inodes
2333 will be placed on the same subdisk, which will impact performance significantly.
2334 Choose an odd number instead, for example 479 kB.
2336 The easiest way to consider the impact of any transfer in a multi-access system
2337 is to look at it from the point of view of the potential bottleneck, the disk
2338 subsystem: how much total disk time does the transfer use? Since just about
2339 everything is cached, the time relationship between the request and its
2340 completion is not so important: the important parameter is the total time that
2341 the request keeps the disks active, the time when the disks are not available to
2342 perform other transfers. As a result, it doesn't really matter if the transfers
2343 are happening at the same time or different times. In practical terms, the time
2344 we're looking at is the sum of the total latency (positioning time and
2345 rotational latency, or the time it takes for the data to arrive under the disk
2346 heads) and the total transfer time. For a given transfer to disks of the same
2347 speed, the transfer time depends only on the total size of the transfer.
2349 Consider a typical news article or web page of 24 kB, which will probably be
2350 read in a single I/O. Take disks with a transfer rate of 6 MB/s and an average
2351 positioning time of 8 ms, and a file system with 4 kB blocks. Since it's 24 kB,
2352 we don't have to worry about fragments, so the file will start on a 4 kB
2353 boundary. The number of transfers required depends on where the block starts:
2354 it's (S + F - 1) / S, where S is the stripe size in file system blocks, and F is
2355 the file size in file system blocks.
2359 Stripe size of 4 kB. You'll have 6 transfers. Total subsystem load: 48 ms
2360 latency, 2 ms transfer, 50 ms total.
2362 Stripe size of 8 kB. On average, you'll have 3.5 transfers. Total subsystem
2363 load: 28 ms latency, 2 ms transfer, 30 ms total.
2365 Stripe size of 16 kB. On average, you'll have 2.25 transfers. Total subsystem
2366 load: 18 ms latency, 2 ms transfer, 20 ms total.
2368 Stripe size of 256 kB. On average, you'll have 1.08 transfers. Total subsystem
2369 load: 8.6 ms latency, 2 ms transfer, 10.6 ms total.
2371 Stripe size of 4 MB. On average, you'll have 1.0009 transfers. Total subsystem
2372 load: 8.01 ms latency, 2 ms transfer, 10.01 ms total.
2375 It appears that some hardware RAID systems have problems with large stripes:
2376 they appear to always transfer a complete stripe to or from disk, so that a
2377 large stripe size will have an adverse effect on performance.
2379 does not suffer from this problem: it optimizes all disk transfers and does not
2380 transfer unneeded data.
2382 Note that no well-known benchmark program tests true multi-access conditions
2383 (more than 100 concurrent users), so it is difficult to demonstrate the validity
2384 of these statements.
2386 Given these considerations, the following factors affect the performance of a
2391 Striping improves performance for multiple access only, since it increases the
2392 chance of individual requests being on different drives.
2394 Concatenating UFS file systems across multiple drives can also improve
2395 performance for multiple file access, since UFS divides a file system into
2396 cylinder groups and attempts to keep files in a single cylinder group. In
2397 general, it is not as effective as striping.
2399 Mirroring can improve multi-access performance for reads, since by default
2401 issues consecutive reads to consecutive plexes.
2403 Mirroring decreases performance for all writes, whether multi-access or single
2404 access, since the data must be written to both plexes. This explains the
2405 subdisk layout in the example of a mirroring configuration above: if the
2406 corresponding subdisk in each plex is on a different physical disk, the write
2407 commands can be issued in parallel, whereas if they are on the same physical
2408 disk, they will be performed sequentially.
2410 RAID-5 reads have essentially the same considerations as striped reads, unless
2411 the striped plex is part of a mirrored volume, in which case the performance of
2412 the mirrored volume will be better.
2414 RAID-5 writes are approximately 25% of the speed of striped writes: to perform
2417 must first read the data block and the corresponding parity block, perform some
2418 calculations and write back the parity block and the data block, four times as
2419 many transfers as for writing a striped plex. On the other hand, this is offset
2420 by the cost of mirroring, so writes to a volume with a single RAID-5 plex are
2421 approximately half the speed of writes to a correctly configured volume with two
2426 configuration changes (for example, adding or removing objects, or the change of
2427 state of one of the objects),
2429 writes up to 128 kB of updated configuration to each drive. The larger the
2430 number of drives, the longer this takes.
2432 .Ss Creating file systems on Vinum volumes
2433 You do not need to run
2435 before creating a file system on a
2441 option to state that the device is not divided into partitions. For example, to
2442 create a file system on volume
2444 enter the following command:
2445 .Bd -literal -offset 4n
2446 # newfs -v /dev/vinum/mirror
2449 A number of other considerations apply to
2454 There is no advantage in creating multiple drives on a single disk. Each drive
2455 uses 131.5 kB of data for label and configuration information, and performance
2456 will suffer when the configuration changes. Use appropriately sized subdisks instead.
2458 It is possible to increase the size of a concatenated
2460 plex, but currently the size of striped and RAID-5 plexes cannot be increased.
2461 Currently the size of an existing UFS file system also cannot be increased, but
2462 it is planned to make both plexes and file systems extensible.
2464 .Sh STATE MANAGEMENT
2465 Vinum objects have the concept of
2469 for more details. They are only completely accessible if their state is
2471 To change an object state to
2475 command. To change an object state to
2479 command. Normally other states are created automatically by the relationship
2480 between objects. For example, if you add a plex to a volume, the subdisks of
2481 the plex will be set in the
2483 state, indicating that, though the hardware is accessible, the data on the
2484 subdisk is invalid. As a result of this state, the plex will be set in the
2487 .Ss The 'reviving' state
2488 In many cases, when you start a subdisk the system must copy data to the
2489 subdisk. Depending on the size of the subdisk, this can take a long time.
2490 During this time, the subdisk is set in the
2492 state. On successful completion of the copy operation, it is automatically set
2495 state. It is possible for the process performing the revive to be stopped and
2496 restarted. The system keeps track of how far the subdisk has been revived, and
2499 command is reissued, the copying continues from this point.
2501 In order to maintain the consistency of a volume while one or more of its plexes
2504 writes to subdisks which have been revived up to the point of the write. It may
2505 also read from the plex if the area being read has already been revived.
2507 The following points are not bugs, and they have good reasons for existing, but
2508 they have shown to cause confusion. Each is discussed in the appropriate
2513 drives are UNIX disk partitions and must have the partition type
2515 This is different from
2517 which expects partitions of type
2519 This behaviour of ccd is an invitation to shoot yourself in the foot: with
2521 you can easily overwrite a file system.
2523 will not permit this.
2525 For similar reasons, the
2528 command will not accept a drive on partition
2532 is used by the system to represent the whole disk, and must be of type
2534 Clearly there is a conflict here, which
2536 resolves by not using the
2540 When you create a volume with multiple plexes,
2542 does not automatically initialize the plexes. This means that the contents are
2543 not known, but they are certainly not consistent. As a result, by default
2545 sets the state of all newly-created plexes except the first to
2547 In order to synchronize them with the first plex, you must
2551 to copy the data from a plex which is in the
2553 state. Depending on the size of the subdisks involved, this can take a long
2556 In practice, people aren't too interested in what was in the plex when it was
2557 created, and other volume managers cheat by setting them
2561 provides two ways to ensure that newly created plexes are
2565 Create the plexes and then synchronize them with
2569 Create the volume (not the plex) with the keyword
2573 to ignore any possible inconsistency and set the plexes to be
2577 Some of the commands currently supported by
2579 are not really needed. For reasons which I don't understand, however, I find
2580 that users frequently try the
2584 commands, though especially
2586 outputs all sort of dire warnings. Don't use these commands unless you have a
2587 good reason to do so.
2589 Some state transitions are not very intuitive. In fact, it's not clear whether
2590 this is a bug or a feature. If you find that you can't start an object in some
2591 strange state, such as a
2593 subdisk, try first to get it into
2599 commands. If that works, you should then be able to start it. If you find
2600 that this is the only way to get out of a position where easier methods fail,
2601 please report the situation.
2603 If you build the kernel module with the
2605 option, you must also build
2609 option, since the size of some data objects used by both components depends on
2610 this option. If you don't do so, commands will fail with the message
2611 .Ar Invalid argument ,
2612 and a console message will be logged such as
2615 vinumioctl: invalid ioctl from process 247 (vinum): c0e44642
2618 This error may also occur if you use old versions of kld or userland program.
2623 command has a particularly emetic syntax. Once it was the only way to start
2625 but now the preferred method is with
2630 should be used for maintenance purposes only. Note that its syntax has changed,
2631 and the arguments must be disk slices, such as
2633 not partitions such as
2639 - directory with device nodes for
2643 .Ar /dev/vinum/control
2644 - control device for
2648 - directory containing device nodes for
2653 - directory containing device nodes for
2659 The name of the log file, by default /var/log/vinum_history.
2660 .It VINUM_DATEFORMAT
2661 The format of dates in the log file, by default %e %b %Y %H:%M:%S.
2663 The name of the editor to use for editing configuration files, by default
2671 .Pa http://www.lemis.com/vinum.html ,
2672 .Pa http://www.lemis.com/vinum-debugging.html .
2674 .An Greg Lehey Aq grog@lemis.com .
2678 command first appeared in
2680 The RAID-5 component of
2682 was developed for Cybernet Inc.
2683 .Pa www.cybernet.com
2684 for its NetMAX product.