7    Administering the UNIX File System

This chapter introduces file systems, disk partitions, and swap space, and explains how to perform the following system administration tasks related to the UNIX File System (UFS):

• Adding swap space

• Managing file system directories and files

• Configuring file system types and locks

• Creating file systems

• Mounting and unmounting file systems

• Tuning and checking file systems

• Managing disk space and troubleshooting disks

• Repartitioning disks

• Cloning disks

• Checking for overlapping partitions on disks

7.1    File Systems and Logical Storage

The Tru64 UNIX operating system supports many file systems and logical storage schemes, including:

• UNIX File System (UFS)

• POLYCENTER Advanced File System (AdvFS)

  For more information about AdvFS, refer to the advfs(4) reference page and the AdvFS Administration guide.

• ISO 9660 Compact Disk File System (CDFS)

• dvdfs - The Digital Versatile Disk File System

• Network File System (NFS)

• Virtual File System (VFS) interface and framework

• Logical Storage Manager (LSM)

For a survey of these capabilities, read the Tru64 UNIX Technical Overview manual. For information about administering these file systems, see AdvFS Administration for AdvFS and Chapter 8 for LSM.

Administration of the NFS is documented in the Network Administration manual.

The Technical Overview points you to sources of information about these file systems:

• Memory File System (MFS)

• /proc File System (PROCFS)

• File-on-File Mounting File System (FFM)

• File Descriptor File System (FDFS)

7.1.1    Disk Partitions

A disk consists of storage units called sectors. Each sector is usually 512 bytes. A sector is addressed by the logical block number (LBN). The LBN is the basic unit of the disk's user-accessible data area that you can address. The first LBN is numbered 0, and the highest LBN is numbered one less than the number of LBNs in the user-accessible area of the disk.

Sectors are grouped together to form up to eight disk partitions. However, disks differ in the number and size of partitions. The /etc/disktab file contains a list of supported disks and the default partition sizes for the system. Refer to the disktab(4) reference page for more information.

Disk partitions are logical divisions of a disk that allow you to organize files by putting them into separate areas of varying sizes. Partitions hold data in structures called file systems and can also be used for system operations such as paging and swapping. File systems have a hierarchical structure of directories and files, which is described in Section 7.1.3. By selecting the file systems to be placed in a partition, you can monitor the growth and activity of the disk.

Disk partitions have default sizes that depend on the type of disk and that can be altered by using the disklabel command. Partitions are named a to h. While the allocated space for a partition can overlap another partition, a properly partitioned disk should not have file systems on overlapping partitions.

Figure 7-1 shows the default partitions and starting (offset) sectors for an RZ73 disk:

Figure 7-1:  RZ73 Default Disk Partitions

The disk label is located in block 0 (zero) in one of the first sectors of the disk. The disk label provides detailed information about the geometry of the disk and the partitions into which the disk is divided. The system disk driver and the boot program use the disk label information to recognize the drive, the disk partitions, and the file systems. Other information is used by the operating system to use the disk most efficiently and to locate important file system information.

The disk label description of each partition contains an identifier for the partition type (for example, standard file system, swap space, and so on). There are two copies of a disk label, one located on the disk and one located in system memory. Because it is faster to access system memory than to perform I/O, when a system recognizes a disk, it copies the disk label into memory. The file system updates the in-memory copy of the label if it contains incomplete information about the file system. If a disk is not labeled and does not contain an ULTRIX-style partition table, the partitions are obtained from the /etc/disktab file. You can change the label with the disklabel command. Refer to Section 7.8 and to the disklabel(8) reference page for more information.

7.1.2    Adding Swap Space

The Tru64 UNIX operating system uses a combination of physical memory and disk space to create virtual memory, which can be much larger than the physical memory. Virtual memory can support more processes than the physical memory alone. This section and the sections that follow describe important virtual memory concepts that you should consider when configuring swap space.

The basic unit of virtual memory and physical memory is the page. Virtual memory attempts to keep a process' most recently referenced virtual pages in physical memory. When a process references virtual pages, they are brought into physical memory from their storage locations on disk. Modified virtual pages can be moved to a temporary location on the disk called swap space if the physical pages (the pages in physical memory) that contain the virtual pages are needed by either a newly referenced virtual page or by a page with a higher priority. Therefore, a process' virtual address space can consist of pages that are located in physical memory, stored temporarily in swap space, and stored permanently on disk in executable or data files. The operating system uses two operations to move virtual pages between physical memory and disk: paging and swapping.

Paging involves moving a single virtual page or a small cluster of pages between disk and physical memory. If a process references a virtual page that is not in physical memory, the operating system reads a copy of the virtual page from its permanent location on disk or from swap space into physical memory. This operation is called a pagein. Pageins typically occur when a process executes a new image and references locations in the executable image that have not been referenced before.

If a physical page is needed to hold a newly referenced virtual page or a page with a higher priority, the operating system writes a modified virtual page (or a small cluster of pages) that has not been recently referenced to the swap space. This operation is called modified page writing or a pageout. Note that only modified virtual pages are written to swap space because there is always a copy of the unmodified pages in their permanent locations on disk.

Swapping involves moving a large number of virtual pages between physical memory and disk. The operating system requires a certain amount of physical memory for efficient operation. If the number of free physical pages drops below the system-defined limit, and if the system is unable to reclaim enough physical memory by paging out individual virtual pages or clusters of pages, the operating system selects a low priority process and reclaims all the physical pages that it is using. It does this by writing all of its modified virtual pages to swap space. This operation is called a swapout. Swapouts typically occur on systems that are memory constrained.

7.1.2.1    How Swap Space Is Allocated

Swap space is initially allocated during system installation. You can add swap space after the installation by including swap space entries in the /etc/fstab file and then rebooting. Additionally, you can use the swapon command to add more swap space--overriding the /etc/fstab definitions--until the next time the system is rebooted. Refer to Section 7.4 and the swapon(8) reference page for more information.

See Section 7.10 for information about how this command interacts with overlapping partitions. The amount of swap space that your system requires depends on the swap space allocation strategy that you use and your system workload. Strategies are described in the following section.

7.1.2.2    Estimating Swap Space Requirements

There are two strategies for swap space allocation: immediate mode and deferred or over-commitment mode. The two strategies differ in the point in time at which swap space is allocated. In immediate mode, swap space is allocated when modifiable virtual address space is created. In deferred mode, swap space is not allocated until the system needs to write a modified virtual page to swap space.

Note

The operating system will terminate a process if it attempts to write a modified virtual page to swap space that is depleted.

Immediate mode is more conservative than deferred mode because each modifiable virtual page is assigned a page of swap space when it is created. If you use the immediate mode of swap space allocation, you must allocate a swap space that is at least as large as the total amount of modifiable virtual address space that will be created on your system. Immediate mode requires significantly more swap space than deferred mode because it guarantees that there will be enough swap space if every modifiable virtual page is modified.

If you use the deferred mode of swap space allocation, you must estimate the total amount of virtual address space that will be both created and modified, and compare that total amount with the size of your system's physical memory. If this total amount is greater than the size of physical memory, the swap space must be large enough to hold the modified virtual pages that do not fit into your physical memory. If your system's workload is complex and you are unable to estimate the appropriate amount of swap space by using this method, you should first use the default amount of swap space and adjust the swap space as needed.

You should always monitor your system's use of swap space. If the system issues messages that indicate that swap space is almost depleted, you can use the swapon command to allocate additional swap space. If you use the immediate mode, swap space depletion prevents you from creating additional modifiable virtual address space. If you use the deferred mode, swap space depletion may result in one or more processes being involuntarily terminated.

7.1.2.3    Selecting the Swap Space Allocation Method

To determine which swap space allocation method is being used, check for the existence of a soft link named /sbin/swapdefault, which points to the primary swap partition. If the /sbin/swapdefault file exists, the system uses the immediate method of swap space allocation. To enable the deferred method, rename or delete this soft link.

You may receive the following informational messages when you remove the /sbin/swapdefault file and when you boot a system that is using the deferred method:

vm_swap_init:  warning sbin/swapdefault swap device not found
vm_swap_init:  in swap over-commitment mode

If the /sbin/swapdefault file does not exist and you want to enable the immediate method of swap allocation, become the root user and create the file by using the following command syntax:

ln -s ../dev/rzxy /sbin/swapdefault

The x variable specifies the device number for the device that holds the primary swap partition, and the y variable specifies the swap partition. Usually, the swap device number is the same as the boot device number, and the primary swap partition is partition b.

You must reboot the system for the new method to take effect.

7.1.3    UNIX File System Structure

This section discusses the UNIX File System (UFS). For information on the Advanced File System (AdvFS) structure, refer to AdvFS Administration.

A UFS file system has four major parts:

• Boot block

  The first block of every file system (block 0) is reserved for a boot, or initialization, program.

• Superblock

  Block 1 of every file system is called the superblock and contains the following information:

  • Total size of the file system (in blocks)

  • Number of blocks reserved for inodes

  • Name of the file system

  • Device identification

  • Date of the last superblock update

  • Head of the free-block list, which contains all of the free blocks (the blocks available for allocation) in the file system

    When new blocks are allocated to a file, they are obtained from the free-block list. When a file is deleted, its blocks are returned to the free-block list.

  • List of free inodes, which is the partial listing of inodes available to be allocated to newly created files

• Inode blocks

  A group of blocks follows the superblock. Each of these blocks contains a number of inodes. Each inode has an associated inumber. An inode describes an individual file in the file system. There is one inode for each possible file in the file system. File systems have a maximum number of inodes; therefore there is a maximum number of files that a file system can contain. The maximum number of inodes depends on the size of the file system.

  The first inode (inode 1) on each file system is unnamed and unused. The second inode (inode 2) must correspond to the root directory for the file system. All other files in the file system are under the file system's root directory. After inode 2, you can assign any inode to any file. You can also assign any data block to any file. The inodes and blocks are not allocated in any particular order.

  If an inode is assigned to a file, the inode can contain the following information:

  • File type

    The possible types are regular, device, named pipes, socket, and symbolic link files.

  • File owner

    The inode contains the user and group identification numbers that are associated with the owner of the file.

  • Protection information

    Protection information specifies read, write, and execute access for the file owner, members of the group associated with the file, and others. The protection information also includes other mode information specified by the chmod command.

  • Link count

    A directory entry (link) consists of a name and the inumber (inode number) that represents the file. The link count indicates the number of directory entries that refer to the file. A file is deleted if the link count is zero; the file's inode is returned to the list of free inodes, and its associated data blocks are returned to the free-block list.

  • Size of the file in bytes

  • Last file access date

  • Last file modification date

  • Last inode modification date

  • Pointers to data blocks

    These pointers indicate the actual location of the data blocks on the physical disk.

• Data blocks

  Data blocks contain user data or system files.

7.1.4    File System and Directory Hierarchy

The standard Tru64 UNIX system directory hierarchy is set up for efficient organization. It separates files by function and intended use. Effective use of the file system includes placing command files in directories that are in the normal search path as specified by the users' .profile or .login file, as appropriate. Figure 7-2 shows the major directories in the file system. Not all of the directories in the Tru64 UNIX hierarchy are shown; you should use those shown in Figure 7-2 to ensure that your product will be portable to other systems. Some of the directories are actually symbolic links.

Figure 7-2:  Partial Tru64 UNIX Directory Hierarchy

Table 7-1 describes the contents and purposes of the directories shown in Figure 7-2.

Table 7-1:  Contents of the Tru64 UNIX Directories

Mounting a file system makes it available for use. Use the mount command to attach file systems to the file system hierarchy under the system root directory; use the umount command to detach them. When you mount a file system, you specify a location (the mount point under the system root directory) to which the file system will attach.

The root directory of a mounted file system is also its mount point. Only one system root directory can exist because the system uses the root directory as its source for system initialization files. Consequently, file systems are mounted under the system root directory.

7.1.5    Directories and File Types

The operating system views files as bit streams, allowing you to define and handle on-disk data, named pipes, UNIX domain sockets, and terminals as files. This object-type transparency provides a simple mechanism for defining and working with a wide variety of storage and communication facilities. The operating system handles the various levels of abstraction as it organizes and manages its internal activities.

While you notice only the external interface, you should understand the various file types recognized by the system. The system supports the following file types:

• Regular files contain data in the form of a program, a text file, or source code, for example.

• Directories are a type of regular file and contain the names of files or other directories.

• Character and block device files identify physical and pseudodevices on the system.

• UNIX domain socket files provide a connection between network processes. The socket system call creates socket files.

• Named pipes are device files that communicating processes operating on a host machine use.

• Linked files point to target files or directories. A linked file contains the name of the target file. A symbolically linked file and its target file can be located on the same file system or on different file systems. A file with a hard link and its target file must be located on the same file system.

7.1.6    Device Special Files

Device special files represent physical devices, pseudodevices, and named pipes. The /dev directory contains device special files. Device special files serve as the link between the system and the device drivers. Each device special file corresponds to a physical device (for example, a disk, tape, printer, or terminal) or a pseudodevice (for example, a network interface, a named pipe, or a UNIX domain socket). The driver handles all read and write operations and follows the required protocols for the device.

Note

Support for rz SCSI device names will be retired in a future release. Any code that derives knowledge about a device from the ASCII name or minor number may be impacted. All code that uses the current device names will be compatible when the change occurs because a mechanism that ensures binary compatability will be provided. Existing interfaces such as names and minor numbers will be fully supported.

There are three types of device files:

• Block device files

  Block device files are used for devices whose driver handles I/O in large blocks and where the kernel handles I/O buffering. Physical devices such as disks are defined as block device files. An example of the block device files in the /dev directory follows:

  brw-------  1 root  system     8,   1 Jan 19 11:20 /dev/rz0a
  brw-------  1 root  system     8,   1 Jan 19 10:09 /dev/rz0b
  

• Character device files

  Character device files are used for devices whose drivers handle their own I/O buffering. Disk, terminal, pseudoterminal, and tape drivers are typically defined as character device files. An example of the character device files in the /dev directory follows:

  crw-rw-rw-  1 root  system    7,    0 Jan 31 16:02 /dev/ptyp0
  crw-rw-rw-  1 root  system    7,    1 Jan 31 16:00 /dev/ptyp1
  crw-rw-rw-  1 root  system    9, 1026 Jan 11 14:20 /dev/rmt1h
  

• Socket device files

  The printer daemon ( lpd ) and error logging daemon ( syslogd ) use the socket device files. An example of the socket device files in the /dev directory follows:

  srw-rw-rw-  1 root  system         0 Jan 22 03:40 log
  srwxrwxrwx  1 root  system         0 Jan 22 03:41 printer
  

Because disk and tape drivers often handle more than one device, each device file has a major and a minor number. The major number specifies (to the kernel) the driver that handles the device. The minor number is passed to the appropriate driver and tells it the device on which to perform the operation.

For static drivers, use the MAKEDEV command or the mknod command to create device special files. The kmknod command creates device special files for third-party kernel layered products. Refer to the MAKEDEV(8), mknod(8), and kmknod(8) reference pages for more information.

For loadable drivers, the sysconfig command creates the device special files by using the information specified in the driver's stanza entry in the /etc/sysconfigtab database file.

7.1.7    Administering Log Files

System log files, such as those found in /var/adm/syslog.dated, are often created and updated automatically. On a well managed system where everything is running smoothly, log files will not grow excessively and will provide important data for administering the system. You will therefore want to keep a certain number of recent log files on hand and ensure that others are archived or recorded in system backups.

On systems with little available disk space, you may want to actively manage the log files. For example the /var/adm/syslog.dated directory contains preserved copies of log files that are used for debugging. Normally, these files do not contain many entries. However, under certain error conditions, a kernel subsystem might log an excessive amount of entries to a file and cause a problem. You should either physically check the logs on a regular basis or use the cron utility to set up a regular job to clear the log files.

The default root crontab file in the /var/spool/cron/crontabs directory contains the following sample line for clearing up the /var/adm/syslog.dated directory:

40 4 * * * find /var/adm/syslog.dated -depth -type d -ctime +5 -exec rm -rf {} ;

If enabled, this cron job will be activated every morning at 4:40 a.m. and will delete any log file in /var/adm/syslog.dated that has not been updated for the last five days. You can edit the crontab file to uncomment and modify this line or add a similar line by using the following command:

# crontab -e

For more information, see the crontab(8) reference page.

Other log files that you may want to administer in this way are:

• The default printer log file, /usr/adm/lperr. Note that you can specify alternate locations for these log files when configuring printers.

• The binary error log file, /usr/adm/binary.errlog.

7.2    Creating File Systems

The newfs command formats a disk partition and creates a usable UNIX file system. For information on creating an AdvFS, refer to AdvFS Administration. Using the information in the disk label or the default values specified in the /etc/disktab file, the newfs command builds a file system on the specified disk partition. You can also use newfs command options to specify the disk geometry.

Note

Changing the default disk geometry values may make it impossible for the fsck program to find the alternate superblocks if the standard superblock is lost.

The newfs command has the following syntax:

/sbin/newfs [-N] [fs_options] device [disk_type]

You must specify the unmounted, raw device (for example, /dev/rrz0a).

Refer to the newfs(8) reference page for information on the command options specific to file systems.

See Section 7.10 for information about how this command interacts with overlapping partitions.

7.3    Checking File Systems

The fsck program checks UNIX file systems and performs some corrections to help ensure a reliable environment for file storage on disks. The fsck program can correct file system inconsistencies such as unreferenced inodes, missing blocks in the free list, or incorrect counts in the superblock.

File systems can become corrupted in many ways, such as improper shutdown procedures, hardware failures, and power outages and power surges. A file system can also become corrupted if you physically write protect a mounted file system, take a mounted file system off line, or if you do not synchronize the system before you shut the system down.

At boot time, the system runs fsck noninteractively, making any corrections that can be done safely. If it encounters an unexpected inconsistency, the fsck program exits, leaves the system in single-user mode, and displays a recommendation that you run the program manually, which allows you to respond yes or no to the prompts that fsck displays.

The command to invoke the fsck program has the following syntax:

/usr/sbin/fsck [options ...] [file_system ...]

If you do not specify a file system, all the file systems in the /etc/fstab file are checked. If you specify a file system, you should always use the raw device.

Refer to the fsck(8) reference page for information about command options.

See Section 7.10 for information about how this command interacts with overlapping partitions.

Note

To check the root file system, you must be in single-user mode, and the file system must be mounted read only. To shut down the system to single-user mode, use the shutdown command.

AdvFS uses write-ahead logging instead of the fsck utility. As your system mounts, AdvFS checks all records in the recovery log for system inconsistencies and makes corrections as needed. Refer to AdvFS Administration for more information.

7.4    Accessing File Systems

You attach a file system to the file system tree by using the mount command, which makes the file system available for use. The mount command attaches the file system to an existing directory (mount point).

Note

The Tru64 UNIX operating system does not support 4-KB block-size file systems. The default block size for Tru64 UNIX file systems is 8 KB. To access the data on a disk that has 4-KB block-size file systems, you must back up the disk to either a tape or a disk that has 8-KB block-size file systems.

When you boot the system, file systems that are defined in the /etc/fstab file are mounted. The /etc/fstab file contains entries that specify the device and partition where the file system is located, the mount point, and additional information about the file system, such as file system type. If you are in single-user mode, the root file system is mounted read only.

Note

To change a file system's mount status, use the mount command with the -u option. This is useful if you try to reboot and the /etc/fstab file is unavailable.

If you try to reboot and the /etc/fstab file is corrupted, use a command similar to the following:

# mount -u /dev/rz0a /

The /dev/rz0a device is the root file system.

The operating system uses the UFS for the root file system. The operating system supports only one root file system from which it accesses the executable kernel ( /vmunix ) and other binaries and files that it needs to boot and initialize. The root file system is mounted at boot time and cannot be unmounted.

The /etc/fstab file contains descriptive information about file systems and swap space and is read by commands such as the mount command. When you boot the system, the /etc/fstab file is read and the file systems described in the file are mounted in the order that they appear in the file. A file system or swap space is described on a single line; information on each line is separated by tabs or spaces. Refer to the swapon(8) reference page for more information about adding swap space.

The order of entries in the /etc/fstab file is important because the mount and umount commands read the file entries in the order that they appear.

You must be root user to edit the /etc/fstab file. To apply the additions that you make to the file, use the mount -a command. Any changes you make to the file become effective when you reboot.

The following is an example of an /etc/fstab file:

/dev/rz2a        /                  ufs     rw      1       1
/dev/rz0g        /usr               ufs     rw      1       2
/dev/rz2b        swap1              ufs     sw      0       2
/dev/rz0b        swap2              ufs     sw      0       2
/dev/rz2g        /var               ufs     rw      1       2
/usr/man@tuscon  /usr/man           nfs     rw,bg   0       0
proj_dmn#testing /projects/testing  advfs   rw      0       0
  [1]                  [2]             [3]       [4]     [5]      [6]

Each line contains an entry and the information is separated either by tabs or spaces. An /etc/fstab file entry has the following information:

1. Specifies the block special device or remote file system to be mounted. For UFS, the special file name is the block special file name, not the character special file name. [Return to example]

2. Specifies the mount point for the file system or remote directory (for example, /usr/man) or swapn for a swap partition. [Return to example]

3. Specifies the type of file system, as follows:

   cdfs	Specifies an ISO 9600 or HS formatted (CD-ROM) file system.
   nfs	Specifies NFS.
   procfs	Specifies a /proc file system, which is used for debugging.
   ufs	Specifies a UFS file system or a swap partition.
   advfs	Specifies an AdvFS file system.

   [Return to example]

4. Describes the mount options associated with the partition. You can specify a list of options separated by commas. Usually, you specify the mount type and any additional options appropriate to the file system type, as follows:

   ro	Specifies that the file system is mounted with read-only access.
   rw	Specifies that the file system is mounted with read-write access.
   sw	Specifies that the partition is used as swap space.
   rq	Specifies that the file system is mounted with read-write access and quotas imposed.
   userquota groupquota	Specifies that the file system is automatically processed by the quotacheck command and that disk quotas are enabled with the quotaon command. By default, user and group quotas for a file system are contained in the quota.user and quota.group files, which are located in the directory specified by the mount point. For example, the quotas for the file system on which /usr is mounted are located in the /usr directory. You also can specify another file name and location. For example: userquota=/var/quotas/tmp.user
   xx	Specifies that the file system entry should be ignored.

   [Return to example]

5. Used by the dump command to determine which UFS file systems should be backed up. If you specify the value 1, the file system is backed up. If you do not specify a value or if you specify 0 (zero), the file system is not backed up. [Return to example]

6. Used by the fsck command to determine the order in which the UNIX file system is checked at boot time. For the root file system, specify 1; for other file systems that you want to check, specify 2. If you do not specify a value or if you specify 0 (zero), the file system is not checked. File systems that use the same disk drive are checked sequentially. File systems on different drives are checked simultaneously to utilize the available parallelism. [Return to example]

7.4.1    Using the mount Command

You use the mount command to make a file system available for use. Unless you add the file system to the /etc/fstab file, the mount will be temporary and will not exist after you reboot the system.

The mount command supports the UFS, AdvFS, NFS, CDFS, and /proc file system types.

The following mount command syntax is for all file systems:

mount [- adflruv ] [-o option ] [-t type] [file_system ] [ mount_point ]

For AdvFS, the file system argument has the following form:

filedomain#fileset

Specify the file system and the mount point, which is the directory on which you want to mount the file system. The directory must already exist on your system. If you are mounting a remote file system, use one of the following syntaxes to specify the file system:

host : remote_directory

remote_directory @ host

The following command lists the currently mounted file systems and the file system options. The backslash contained in this example indicates line continuation and is not in the actual display.

# mount -l
/dev/rz2a on / type ufs (rw,exec,suid,dev,nosync,noquota)
/dev/rz0g on /usr type ufs (rw,exec,suid,dev,nosync,noquota)
/dev/rz2g on /var type ufs (rw,exec,suid,dev,nosync,noquota)
/dev/rz3c on /usr/users type ufs (rw,exec,suid,dev,nosync,noquota)
/usr/share/man@tuscon on /usr/share/man type nfs (rw,exec,suid,dev,
nosync,noquota,hard,intr,ac,cto,noconn,wsize=8192,rsize=8192,
timeo=10,retrans=10,acregmin=3,acregmax=60,acdirmin=30,acdirmax=60)
proj_dmn#testing on /alpha_src type advfs (rw,exec,suid,dev,nosync,\
noquota)

The following command mounts the /usr/homer file system located on host acton on the local /homer mount point with read-write access:

# mount -t nfs -o rw acton:/usr/homer /homer

Refer to the mount(8) reference page for more information on general options and options specific to a file system type.

See Section 7.10 for information about how this command interacts with overlapping partitions.

7.4.2    Using the umount Command

Use the umount command to unmount a file system. You must unmount a file system if you want to check it with the fsck command or if you want to change its partitions with the disklabel command. The umount command has the following syntax:

umount [- afv ] [- h host ] [- t type ] [ mount_point ]

If any user process (including a cd command) is in effect within the file system, you cannot unmount the file system. If the file system is in use when the command is invoked, the system returns the following error message and does not unmount the file system:

mount device busy

You cannot unmount the root file system with the umount command.

7.5    Tuning File Systems

To enhance the efficiency of UFS reads, use the tunefs command to change a file system's dynamic parameters, which affect layout policies.

The tunefs command has the following syntax:

tunefs [- a maxc ] [- d rotd ] [- e maxb ] [- m minf ] [- o op t] [ file_s ]

You can use the tunefs command on both mounted and unmounted file systems; however, changes are applied only if you use the command on unmounted file systems. If you specify the root file system, you must also reboot to apply the changes.

You can use command options to specify the dynamic parameters that affect the disk partition layout policies. Refer to the tunefs(8) reference page for more information on the command options.

7.6    Maintaining Disks

The radisk program and the scu program allow you to maintain your Digital Storage Architecture (DSA) and Small Computer System Interface (SCSI) disk devices, respectively.

With the radisk program, you can perform the following tasks on a DSA disk device:

• Clear a forced error indicator

• Set or clear the exclusive access attribute

• Replace a bad block

• Scan the disk for bad blocks

Refer to the radisk(8) reference page for more information.

The scu program allows you to perform the following tasks on SCSI disk devices, in addition to other tasks:

• Format media

• Reassign a defective block

• Reserve and release a device

• Display and set device and program parameters

• Enable and disable a device

Refer to Appendix B and to the scu(8) reference page for more information.

7.7    Monitoring Disk Use

To ensure an adequate amount of free disk space, you should regularly monitor the disk use of your configured file systems. You can do this in any of the following ways:

• Check available free space by using the df command

• Check disk use by using the du command or the quot command

• Verify disk quotas (if imposed) by using the quota command

You can use the quota command only if you are the root user.

7.7.1    Checking Available Free Space

To ensure sufficient space for your configured file systems, you should regularly use the df command to check the amount of free disk space in all of the mounted file systems. The df command displays statistics about the amount of free disk space on a specified file system or on a file system that contains a specified file.

The df command has the following syntax:

df [- eiknPt ] [-F fstype ] [file ] [ file_system ...]

With no arguments or options, the df command displays the amount of free disk space on all of the mounted file systems. For each file system, the df command reports the file system's configured size in 512-byte blocks, unless you specify the -k option, which reports the size in kilobyte blocks. The command displays the total amount of space, the amount presently used, the amount presently available (free), the percentage used, and the directory on which the file system is mounted.

For AdvFS file domains, the df command displays disk space usage information for each fileset.

If you specify a device that has no file systems mounted on it, df displays the information for the root file system.

You can specify a file pathname to display the amount of available disk space on the file system that contains the file.

Refer to the df(1) reference page for more information.

Note

You cannot use the df command with the block or character special device name to find free space on an unmounted file system. Instead, use the dumpfs command.

The following example displays disk space information about all the mounted file systems:

# /sbin/df 
Filesystem         512-blks   used  avail capacity Mounted on
/dev/rz2a            30686  21438   6178    77%   /
/dev/rz0g           549328 378778 115616    76%   /usr
/dev/rz2g           101372   5376  85858     5%   /var
/dev/rz3c           394796     12 355304     0%   /usr/users
/usr/share/man@tsts 557614 449234  52620    89%   /usr/share/man
domain#usr          838432 680320  158112   81%   /usr

Note

The newfs command reserves a percentage of the file system disk space for allocation and block layout. This can cause the df command to report that a file system is using more than 100 percent of its capacity. You can change this percentage by using the tunefs command with the -minfree flag.

7.7.2    Checking Disk Use

If you determine that a file system has insufficient space available, check how its space is being used. You can do this with the du command or the quot command.

The du command pinpoints disk space allocation by directory. With this information you can decide who is using the most space and who should free up disk space.

The du command has the following syntax:

/usr/bin/du [- aklrsx ] [ directory ... filename ...]

The du command displays the number of blocks contained in all directories (listed recursively) within each specified directory, file name, or (if none are specified) the current working directory. The block count includes the indirect blocks of each file in 1-kilobyte units, independent of the cluster size used by the system.

If you do not specify any options, an entry is generated only for each directory. Refer to the du(1) reference page for more information on command options.

The following example displays a summary of blocks that all main subdirectories in the /usr/users directory use:

# /usr/bin/du -s /usr/users/*  
440     /usr/users/barnam
43      /usr/users/broland
747     /usr/users/frome
6804    /usr/users/morse
11183   /usr/users/rubin
2274    /usr/users/somer

From this information, you can determine that user rubin is using the most disk space.

The following example displays the space that each file and subdirectory in the /usr/users/rubin/online directory uses:

# /usr/bin/du -a /usr/users/rubin/online 
1	/usr/users/rubin/online/inof/license
2	/usr/users/rubin/online/inof
7	/usr/users/rubin/online/TOC_ft1
16	/usr/users/rubin/online/build
  .
  .
  .
251	/usr/users/rubin/online

Note

As an alternative to the du command, you can use the ls -s command to obtain the size and usage of files. Do not use the ls -l command to obtain usage information; ls -l displays only file sizes.

You can use the quot command to list the number of blocks in the named file system currently owned by each user. You must be root user to use the quot command.

The quot command has the following syntax:

/usr/sbin/quot [-c] [-f] [-n] [file_system]

The following example displays the number of blocks used by each user and the number of files owned by each user in the /dev/rz0h file system:

# /usr/sbin/quot -f /dev/rrz0h 

Note

The character device special file must be used to return the information, because when the device is mounted the block special device file is busy.

Refer to the quot(8) reference page for more information.

7.7.3    Setting User and Group Quotas for UFS

This section provides information on setting user and group quotas for UFS. As a system administrator, you establish usage limits for user accounts and for groups by setting file system quotas, also known as disk quotas, for them. For information on setting AdvFS user and group quotas, refer to AdvFS Administration. For more information on user and group quotas on UFS, refer to Section 9.3.4.

You can apply quotas to file systems to establish a limit on the number of blocks and inodes (or files) that a user account or a group of users can allocate. You can set a separate quota for each user or group of users on each file system. You may want to set quotas on file systems that contain home directories, such as /usr/users, because the sizes of these file systems can increase more significantly than other file systems. You should avoid setting quotas on the /tmp file system.

7.7.3.1    Hard and Soft Quota Limits

File systems can have both soft and hard quota limits. When a hard limit is reached, no more disk space allocations or file creations that would exceed the limit are allowed. The soft limit may be reached for a period of time (called the grace period). If the soft limit is reached for an amount of time that exceeds the grace period, no more disk space allocations or file creations are allowed until enough disk space is freed or enough files are deleted to bring the disk space usage or number of files below the soft limit.

Caution

With both hard and soft limits, you can end up with a partially-written file if the quota limit is reached while you are writing to the file.

If you are in an editor and exceed a quota limit, do not abort the editor or write the file because data may be lost. Instead, use the editor exclamation point ( ! ) shell escape command to remove files. You can also write the file to another file system, such as /tmp, remove files from the file system whose quota you reached, and then move the file back to that file system.

It is important to note that a hard limit is one more unit (blocks, files, or inodes) than will be allowed when the quota limit is active. The quota is up to, but not including the limit. For example, if you set a hard limit of 10,000 disk blocks for each user account in a file system, an account reaches the hard limit when 9,999 disk blocks have been allocated. If you want a maximum of 10,000 complete blocks for the user account, you must set the hard limit to 10,001.

7.7.3.2    Activating File System Quotas

To activate quotas on a UNIX file system, perform the following steps.

1. Configure the system to include the disk quota subsystem by editing the /sys/conf/NAME system configuration file to include the following line:

   options       QUOTA
   

2. Edit the /etc/fstab file and change the fourth field of the file system's entry to read rw, userquota, and groupquota.

3. Use the quotacheck command to create a quota file where the quota subsystem stores current allocations and quota limits. Refer to the quotacheck(8) reference page for command information.

4. Use the edquota command to activate the quota editor and create a quota entry for each user.

   For each user or group you specify, edquota creates a temporary ASCII file that you edit with the vi editor. Edit the file to include entries for each file system with quotas enforced, the soft and hard limits for blocks and inodes (or files), and the grace period.

   If you specify more than one user name or group name in the edquota command line, the edits will affect each user or group. You can also use prototypes that allow you to quickly set up quotas for groups of users. Refer to the edquota(8) reference page for more information.

5. Use the quotaon command to activate the quota system. Refer to the quotaon(8) reference page for more information.

6. To check and enable disk quotas during system startup, use the following command to set the disk quota configuration variable in the /etc/rc.config file:

   # /usr/sbin/rcmgr set QUOTA_CONFIG yes
   

If you want to turn off quotas, use the quotaoff command. Also, the umount command turns off quotas before it unmounts a file system. Refer to the quotaoff(8) reference page for more information.

7.7.4    Verifying Disk Quotas

If you are enforcing user disk quotas, you should periodically verify your quota system. You can use the quotacheck, quota, and repquota commands to compare the established limits with actual use.

The quotacheck command verifies that the actual block use is consistent with established limits. You should run the quotacheck command twice: when quotas are first enabled on a file system and after each reboot. The command gives more accurate information when there is no activity on the system.

The quota command displays the actual block use for each user in a file system. Only the root user can execute the quota command.

The repquota command displays the actual disk use and quotas for the specified file system. For each user, the current number of files and the amount of space (in KB) is displayed along with any quotas.

If you find it necessary to change the established quotas, use the edquota command, which allows you to set or change the limits for each user.

Refer to the quotacheck(8), quota(8), and repquota(8) reference pages for more information on disk quotas.

7.8    Partitioning Disks

This section provides the information you need to change the partition scheme of your disks. In general, you allocate disk space during the initial installation or when adding disks to your configuration. Usually, you do not have to alter partitions; however, there are cases when it is necessary to change the partitions on your disks to accommodate changes and to improve system performance.

The disk label provides detailed information about the geometry of the disk and the partitions into which the disk is divided. You can change the label with the disklabel command. You must be the root user to use the disklabel command.

There are two copies of a disk label, one located on the disk and one located in system memory. Because it is faster to access system memory than to perform I/O, when the system boots, it copies the disk label into memory. Use the disklabel -r command to directly access the label on the disk instead of going through the in-memory label.

Note

Before you change disk partitions, back up all the file systems if there is any data on the disk. Changing a partition overwrites the data on the old file system, destroying the data. Refer to Section 7.1.1 for more information on disk partitions.

The following rules apply to changing partitions:

• You cannot change the offset, which is the beginning sector, or shrink any partition on a mounted file system or on a file system that has an open file descriptor.

• If you need only one partition on the entire disk, use partition c.

• Specify the raw device for partition a, which begins at the start of the disk sector (sector 0), when you change the label.

Before changing the size of a disk partition, review the current partition setup by viewing the disk label. The disklabel command allows you to view the partition sizes. The bottom, top, and size of the partitions are in 512-byte sectors.

To review the current disk partition setup, use the following disklabel command syntax:

disklabel -r device

Specify the device with its directory name (/dev) followed by the raw device name, drive number, and partition a or c. You can also specify the disk unit and number, such as rz1.

An example of using the disklabel command to view a disk label follows:

# disklabel -r /dev/rrz3a 
type: SCSI
disk: rz26
label:
flags:
bytes/sector: 512
sectors/track: 57
tracks/cylinder: 14
sectors/cylinder: 798
cylinders: 2570
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # milliseconds
track-to-track seek: 0  # milliseconds
drivedata: 0
 
8 partitions:
#       size offset   fstype [fsize bsize cpg]
 a:  131072       0   4.2BSD   1024  8192  16  # (Cyl.    0 - 164*)
 b:  262144  131072   unused   1024  8192      # (Cyl.  164*- 492*)
 c: 2050860       0   unused   1024  8192      # (Cyl.    0 - 2569)
 d:  552548  393216   unused   1024  8192      # (Cyl.  492*- 1185*)
 e:  552548  945764   unused   1024  8192      # (Cyl. 1185*- 1877*)
 f:  552548 1498312   unused   1024  8192      # (Cyl. 1877*- 2569*)
 g:  819200  393216   unused   1024  8192      # (Cyl.  492*- 1519*)
 h:  838444 1212416   4.2BSD   1024  8192 16   # (Cyl. 1519*- 2569*)

You must be careful when you change partitions because you can overwrite data on the file systems or make the system inefficient. If the partition label becomes corrupted while you are changing the partition sizes, you can return to the default partition label by using the disklabel command with the -w option, as follows:

# disklabel -r -w /dev/rrz1a rz26

The disklabel command allows you to change the partition label of an individual disk without rebuilding the kernel and rebooting the system. Use the following procedure:

1. Display disk space information about the file systems by using the df command.

2. View the /etc/fstab file to determine if any file systems are being used as swap space.

3. Examine the disk's label by using the disklabel command with the -r option. Refer to the rz(7) and ra(7) reference pages and to the /etc/disktab file for information on the default disk partitions.

4. Unmount the file systems on the disk whose label you want to change.

5. Calculate the new partition parameters. You can increase or decrease the size of a partition. You can also cause partitions to overlap.

6. Edit the disk label by using the disklabel command with the -e option to change the partition parameters, as follows:

   disklabel -e [-r] disk

   An editor, either the vi editor or that specified by the EDITOR environment variable, is invoked so you can edit the disk label, which is in the format displayed with the disklabel -r command.

   The -r option writes the label directly to the disk and updates the system's in-memory copy, if possible. The disk parameter specifies the unmounted disk (for example, rz0 or /dev/rrz0a).

   After you quit the editor and save the changes, the following prompt is displayed:

   write new label? [?]:
   

   Enter y to write the new label or n to discard the changes.

7. Use the disklabel command with the -r option to view the new disk label.

8. Edit the /etc/fstab file to include the new file systems.

7.9    Cloning Disks

You can use the dd command to clone a complete disk or a disk partition; that is, you can produce a physical copy of the data on the disk or disk partition.

Because the dd command was not meant for copying multiple files, you should clone a disk or a partition only on a disk that is used as a data disk or one that does not contain a file system. Use the dump and restore commands described in Chapter 11 to clone disks or partitions that contain a file system. For example the following command can be used to clone the/usr partition of a disk to a new disk partition mlunted on /newdisk

#dump -0f - /usr | (cd /newdisk ; restore -xf -)

Using this command, you can clone a complete disk. For example, if you wanted to move your system disk to a disk with larger capacity. To clone a complete disk, use the following procedure:

1. set up the target disk using diskconfig. The new disk should have target partitions for each partition on the source disk and requires a boot block if it is to be a boot disk. Not that you do not need identical partitions if you do not plan to clone all data, however, you will need to adjust your /etc/fstab file before using the new disk.

   Identify the device special files for the source and target disks (dev/disk/dskNx). Use dsfmgr or hwmgr to identify and check disk characteristics.

2. Before exiting diskconfig, note the physical location (bust, target, and lun) of the source and target disks.

3. The /etc/sysconfigtab file lists the swap partitions that you will need to recreate on the target disk. Use the swapon command to check the swap partitions.

4. Create a /newdisk mount point and mount each partition of the target in turn, beginning with the apartition. For example:

   # /mount /dev/disk/disk6a /newdisk
   

5. Dump each partition in turn using the following command syntax:

   
   # dump -0u -f - /dev/disk/disk0a | (cd /newdisk ; restore -r -f -)
   

6. Verify file ownerships and protection and ensure that all required filesystem branches were dumped. You can do this by using the diff command on the directory listings for the source and target disks.

Tru64 UNIX protects the first block of a disk with a valid disk label because this is where the disk label is stored. As a result, if you clone a partition to a partition on a target disk that contains a valid disk label, you must decide whether you want to keep the existing disk label on that target disk.

If you want to maintain the disk label on the target disk, use the dd command with the skip and seek options to move past the protected disk label area on the target disk. Note that the target disk must be the same size as or larger than the original disk.

To determine if the target disk has a label, use the following disklabel command syntax:

disklabel -r target_device

You must specify the target device directory name (/dev) followed by the raw device name, drive number, and partition c. If the disk does not contain a label, the following message is displayed:

Bad pack magic number (label is damaged, or pack is unlabeled)

The following example shows a disk that already contains a label:

# disklabel -r /dev/rrz1c

type: SCSI
disk: rz26
label:
flags:
bytes/sector: 512
sectors/track: 57
tracks/cylinder: 14
sectors/cylinder: 798
cylinders: 2570
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # milliseconds
track-to-track seek: 0  # milliseconds
drivedata: 0
 
8 partitions:
#      size  offset  fstype [fsize bsize  cpg]
 a:  131072       0  unused 1024 8192 # (Cyl.    0 - 164*)
 b:  262144  131072  unused 1024 8192 # (Cyl.  164*- 492*)
 c: 2050860       0  unused 1024 8192 # (Cyl.    0 - 2569)
 d:  552548  393216  unused 1024 8192 # (Cyl.  492*- 1185*)
 e:  552548  945764  unused 1024 8192 # (Cyl. 1185*- 1877*)
 f:  552548 1498312  unused 1024 8192 # (Cyl. 1877*- 2569*)
 g:  819200  393216  unused 1024 8192 # (Cyl.  492*- 1519*)
 h:  838444 1212416  unused 1024 8192 # (Cyl. 1519*- 2569*)

If the target disk already contains a label and you do not want to keep the label, you must zero (clear) the label by using the disklabel -z command. For example:

# disklabel -z /dev/rrz1c

To clone the original disk to the target disk and keep the target disk label, use the following dd command syntax:

dd if=original_disk of=target_disk skip=16 seek=16 bs=512

Specify the device directory name (/dev) followed by the raw device name, drive number, and the original and target disk partitions. For example:

# dd if=/dev/rrz0c of=/dev/rrz1c skip=16 seek=16 bs=512

7.10    Checking for Overlapping Partitions

Commands to mount or create file systems, add a new swap device, and add disks to the Logical Storage Manager first check whether the disk partition specified in the command already contains valid data, and whether it overlaps with a partition that is already marked for use. The fstype field of the disk label is used to determine when a partition or an overlapping partition is in use.

If the partition is not in use, the command continues to execute. In addition to mounting or creating file systems, commands like mount, newfs, fsck, voldisk, mkfdmn, rmfdmn, and swapon also modify the disk label, so that the fstype field specifies how the partition is being used. For example, when you add a disk partition to an AdvFS domain, the fstype field is set to AdvFS.

If the partition is not available, these commands return an error message and ask if you want to continue, as shown in the following example:

# newfs /dev/rrz8c 
WARNING: disklabel reports that rz8c currently
is being used as "4.2BSD" data. Do you want to
continue with the operation and possibly destroy
existing data? (y/n) [n]

Applications, as well as operating system commands, can modify the fstype of the disk label, to indicate that a partition is in use. See the check_usage(3) and set_usage(3) reference pages for more information.