JavaScript is required to for searching.
Skip Navigation Links
Exit Print View
man pages section 1M: System Administration Commands     Oracle Solaris 10 1/13 Information Library
search filter icon
search icon

Document Information



System Administration Commands - Part 1

System Administration Commands - Part 2

System Administration Commands - Part 3
































































































- incremental file system restore


/usr/sbin/ufsrestore i | r | R | t | x [abcdfhlmostvyLT] 
     [archive_file] [factor] [dumpfile] [n] [label] 
     [timeout] [filename]...


The ufsrestore utility restores files from backup media created with the ufsdump command. ufsrestores's actions are controlled by the key argument. The key is exactly one function letter (i, r, R , t, or x) and zero or more function modifiers (letters). The key string contains no SPACE characters. Function modifier arguments are listed on the command line in the same order as their corresponding function modifiers appear in the key string.

filename arguments which appear on the command line, or as arguments to an interactive command, are treated as shell glob patterns by the x and t functions; any files or directories matching the patterns are selected. The metacharacters *, ?, and [ ] must be protected from the shell if they appear on the command line. There is no way to quote these metacharacters to explicitly match them in a filename.

The temporary files rstdir* and rstmode* are placed in /tmp by default. If the environment variable TMPDIR is defined with a non-empty value, that location is used instead of /tmp.


Function Letters

You must specify one (and only one) of the function letters listed below. Note that i, x, and r are intended to restore files into an empty directory. The R function is intended for restoring into a populated directory.


Interactive. After reading in the directory information from the media, ufsrestore invokes a shell-like interface that allows you to browse through the dump file's directory hierarchy and select individual files to be extracted. Restoration has the same semantics as x (see below). See Interactive Commands, below, for a description of available commands.


Recursive. Starting with an empty directory and a level 0 dump, the r function recreates the filesystem relative to the current working directory, exactly as it appeared when the dump was made. Information used to restore incremental dumps on top of the full dump (for example, restoresymtable) is also included. Several ufsrestore runs are typical, one for each higher level of dump (0, 1, ..., 9). Files that were deleted between the level 0 and a subsequent incremental dump will not exist after the final restore. To completely restore a file system, use the r function restore the level 0 dump, and again for each incremental dump. Although this function letter is intended for a complete restore onto a new file system (one just created with newfs(1M)), if the file system contains files not on the backup media, they are preserved.


Resume restoring. If an r-mode ufsrestore was interrupted, this function prompts for the volume from which to resume restoring and continues the restoration from where it was left off. Otherwise identical to r.


Table of contents. List each filename that appears on the media. If no filename argument is given, the root directory is listed. This results in a list of all files on the media, unless the h function modifier is in effect. The table of contents is taken from the media or from the specified archive file, when the a function modifier is used. The a function modifier is mutually exclusive with the x and r function letters.


Extract the named files from the media. Files are restored to the same relative locations that they had in the original file system.

If the filename argument matches a directory whose contents were written onto the media, and the h modifier is not in effect, the directory is recursively extracted, relative to the current directory, which is expected to be empty. For each file, the owner, modification time, and mode are restored (if possible).

If you omit the filename argument or specify ., the root directory is extracted. This results in the entire tape being extracted, unless the h modifier is in effect. . With the x function, existing files are overwritten and ufsrestore displays the names of the overwritten files. Overwriting a currently-running executable can have unfortunate consequences.

Use the x option to restore partial file system dumps, as they are (by definition) not entire file systems.

Function Modifiers

a archive_file

Read the table of contents from archive_file instead of the media. This function modifier can be used in combination with the t, i, or x function letters, making it possible to check whether files are on the media without having to mount the media. When used with the x and interactive (i) function letters, it prompts for the volume containing the file(s) before extracting them.

b factor

Blocking factor. Specify the blocking factor for tape reads. For variable length SCSI tape devices, unless the data was written with the default blocking factor, a blocking factor at least as great as that used to write the tape must be used; otherwise, an error will be generated. Note that a tape block is 512 bytes. Refer to the man page for your specific tape driver for the maximum blocking factor.


Convert the contents of the media in 4.1BSD format to the new ufs file system format.


Debug. Turn on debugging output.

f dump_file

Use dump_file instead of /dev/rmt/0 as the file to restore from. Typically dump_file specifies a tape or diskette drive. If dump_file is specified as `-', ufsrestore reads from the standard input. This allows ufsdump(1M) and ufsrestore to be used in a pipeline to copy a file system:

example# ufsdump 0f - /dev/rdsk/c0t0d0s7 \
 | (cd /home;ufsrestore xf -)

If the name of the file is of the form machine:device, the restore is done from the specified machine over the network using rmt(1M). Since ufsrestore is normally run by root, the name of the local machine must appear in the /.rhosts file of the remote machine. If the file is specified as user@machine:device, ufsrestore will attempt to execute as the specified user on the remote machine. The specified user must have a .rhosts file on the remote machine that allows the user invoking the command from the local machine to access the remote machine.


Extract or list the actual directory, rather than the files that it references. This prevents hierarchical restoration of complete subtrees from the tape.


Autoload. When the end-of-tape is reached before the restore is complete, take the drive off-line and wait up to two minutes (the default, see the T function modifier) for the tape drive to be ready again. This gives autoloading (stackloader) tape drives a chance to load a new tape. If the drive is ready within two minutes, continue. If it is not, prompt for another tape and wait.

L label

The label that should appear in the header of the dump file. If the labels do not match, ufsrestore issues a diagnostic and exits. The tape label is specific to the ufsdump tape format, and bears no resemblance to IBM or ANSI-standard tape labels.


Extract by inode numbers rather than by filename to avoid regenerating complete pathnames. Regardless of where the files are located in the dump hierarchy, they are restored into the current directory and renamed with their inode number. This is useful if only a few files are being extracted.


Offline. Take the drive off-line when the restore is complete or the end-of-media is reached and rewind the tape, or eject the diskette. In the case of some autoloading 8mm drives, the tape is removed from the drive automatically.

s n

Skip to the nth file when there are multiple dump files on the same tape. For example, the command:

example# ufsrestore xfs /dev/rmt/0hn 5

would position you to the fifth file on the tape when reading volume 1 of the dump. If a dump extends over more than one volume, all volumes except the first are assumed to start at position 0, no matter what “s n” value is specified.

If “s n” is specified, the backup media must be at BOT (beginning of tape). Otherwise, the initial positioning to read the table of contents will fail, as it is performed by skipping the tape forward n-1 files rather than by using absolute positioning. This is because on some devices absolute positioning is very time consuming.

T timeout [hms]

Sets the amount of time to wait for an autoload command to complete. This function modifier is ignored unless the l function modifier has also been specified. The default timeout period is two minutes. The time units may be specified as a trailing h (hours), m (minutes), or s (seconds). The default unit is minutes.


Verbose. ufsrestore displays the name and inode number of each file it restores, preceded by its file type.


Do not ask whether to abort the restore in the event of tape errors. ufsrestore tries to skip over the bad tape block(s) and continue as best it can.

Interactive Commands

ufsrestore enters interactive mode when invoked with the i function letters. Interactive commands are reminiscent of the shell. For those commands that accept an argument, the default is the current directory. The interactive options are:

add [filename]

Add the named file or directory to the list of files to extract. If a directory is specified, add that directory and its files (recursively) to the extraction list (unless the h modifier is in effect).

cd directory

Change to directory (within the dump file).

delete [filename]

Delete the current directory, or the named file or directory from the list of files to extract. If a directory is specified, delete that directory and all its descendents from the extraction list (unless the h modifier is in effect). The most expedient way to extract a majority of files from a directory is to add that directory to the extraction list, and then delete specific files to omit.


Extract all files on the extraction list from the dump media. ufsrestore asks which volume the user wishes to mount. The fastest way to extract a small number of files is to start with the last volume and work toward the first. If “s n” is given on the command line, volume 1 will automatically be positioned to file n when it is read.


Display a summary of the available commands.

ls [directory]

List files in directory or the current directory, represented by a `.' (period). Directories are appended with a `/' (slash). Entries marked for extraction are prefixed with a `*' (asterisk). If the verbose option is in effect, inode numbers are also listed.

marked [directory]

Like ls, except only files marked for extraction are listed.


Toggle the pagination of the output from the ls and marked commands. The pager used is that defined by the PAGER environment variable, or more(1) if that envar is not defined. The PAGER envar may include white-space-separated arguments for the pagination program.


Print the full pathname of the current working directory.


ufsrestore exits immediately, even if the extraction list is not empty.


Prompts: set owner/mode for `.' (period). Type y for yes to set the mode (permissions, owner, times) of the current directory `.' (period) into which files are being restored equal to the mode of the root directory of the file system from which they were dumped. Normally, this is what you want when restoring a whole file system, or restoring individual files into the same locations from which they were dumped. Type n for no, to leave the mode of the current directory unchanged. Normally, this is what you want when restoring part of a dump to a directory other than the one from which the files were dumped.

setpager command

Sets the command to use for paginating output instead of the default or that inherited from the environment. The command string may include arguments in addition to the command itself.


Toggle the status of the v modifier. While v is in effect, the ls command lists the inode numbers of all entries, and ufsrestore displays information about each file as it is extracted.


Display the dump header on the media.


The following operands are supported.


Specifies the pathname of files (or directories) to be restored to disk. Unless the h function modifier is also used, a directory name refers to the files it contains, and (recursively) its subdirectories and the files they contain. filename is associated with either the x or t function letters, and must come last.


See largefile(5) for the description of the behavior of ufsrestore when encountering files greater than or equal to 2 Gbyte ( 231 bytes).

Exit Status

The following exit values are returned:


Successful completion.


An error occurred. Verbose messages are displayed.

Environment Variables


The command to use as a filter for paginating output. This can also be used to specify the options to be used. Default is more(1).


Selects the directory for temporary files. Defaults to /tmp if not defined in the environment.



the default tape drive


file containing directories on the tape


owner, mode, and timestamps for directories


information passed between incremental restores


See attributes(5) for descriptions of the following attributes:


See Also

more(1), mkfs(1M), mount(1M), rmt(1M), ufsdump(1M), ufsdump(4), attributes(5), largefile(5)


ufsrestore complains about bad option characters.

Read errors result in complaints. If y has been specified, or the user responds y, ufsrestore will attempt to continue.

If the dump extends over more than one tape, ufsrestore asks the user to change tapes. If the x or i function letter has been specified, ufsrestore also asks which volume the user wishes to mount. If the s modifier has been specified, and volume 1 is mounted, it is automatically positioned to the indicated file.

There are numerous consistency checks that can be listed by ufsrestore. Most checks are self-explanatory or can “never happen”. Common errors are given below.

Converting to new file system format

A dump tape created from the old file system has been loaded. It is automatically converted to the new file system format.

filename: not found on tape

The specified file name was listed in the tape directory, but was not found on the tape. This is caused by tape read errors while looking for the file, using a dump tape created on an active file system, or restoring a partial dump with the r function.

expected next file inumber, got inumber

A file that was not listed in the directory showed up. This can occur when using a dump tape created on an active file system.

Incremental tape too low

When doing an incremental restore, a tape that was written before the previous incremental tape, or that has too low an incremental level has been loaded.

Incremental tape too high

When doing incremental restore, a tape that does not begin its coverage where the previous incremental tape left off, or one that has too high an incremental level has been loaded.

media read error: invalid argument

Blocking factor specified for read is smaller than the blocking factor used to write data.

Tape read error while restoring
Tape read error while skipping over inode inumber
Tape read error while trying to resynchronize
A tape read error has occurred

If a file name is specified, then its contents are probably partially wrong. If an inode is being skipped or the tape is trying to resynchronize, then no extracted files have been corrupted, though files may not be found on the tape.

resync ufsrestore, skipped num

After a tape read error, ufsrestore may have to resynchronize itself. This message lists the number of blocks that were skipped over.

Incorrect tape label. Expected `foo', got `bar'.

The L option was specified, and its value did not match what was recorded in the header of the dump file.


ufsrestore can get confused when doing incremental restores from dump tapes that were made on active file systems.

A level 0 dump must be done after a full restore. Because ufsrestore runs in user mode, it has no control over inode allocation. This means that ufsrestore repositions the files, although it does not change their contents. Thus, a full dump must be done to get a new set of directories reflecting the new file positions, so that later incremental dumps will be correct.