9    Using the Storage Administrator

This chapter describes how to manage LSM objects, including disks, disk groups, volumes, plexes, and subdisks using the Storage Administrator GUI. The tasks described in this chapter can also be accomplished by using:

See Appendix A for more information on how to track Storage Administrator activities, how to use the Storage Administrator, and how to customize the Storage Administrator GUI.

9.1    Overview

The Storage Administrator is the Java-based graphical user interface (GUI) for LSM. The Storage Administrator displays a hierarchical view of LSM objects and their relationships. You use the Storage Administrator to view and manage LSM objects on a local or remote (client) system.

The Storage Administrator provides dialog boxes in which you enter information to create or manage LSM objects. Completing a dialog box may be the equivalent of entering several commands.

Storage Administrator consists of a server (daemon) and a client. The Storage Administrator server runs on a system on which the LSM software is initialized and running. The Storage Administrator client runs on any machine that supports the Java run-time environment.

Note the following considerations when using the Storage Administrator:

9.1.1    Installing and Starting the Storage Administrator

To install the Storage Administrator, choose the LSM GUI option during the LSM installation. Install the Storage Administrator on all systems on which you want to use the Storage Administrator to remotely manage the LSM software. See Chapter 2 for information on initializing the LSM software.

To use the Storage Administrator, you must either log in as root or have your user name is in the /etc/group file for the system in the group defined for the Storage Administrator administration (lsmsa_admin by default).

You can start the Storage Administrator in:

If the Storage Administrator does not start during the boot process, enter:

/sbin/init.d/lsmsa stop

Wait a few minutes and restart the Storage Administrator by entering the following command:

/sbin/init.d/lsmsa start

You can also access Storage Administrator from the SysMan Menu and from SysMan Station.

9.1.2    Main Window

The Storage Administrator main window consists of two panes. The left pane displays a hierarchical tree of objects. The right pane displays an object table that lists the properties of the object selected on the left. The window also has a menu bar and a command launcher that you can hide or display to initiate LSM options. Figure 9-1 shows the components of the main window.

Figure 9-1:  The Storage Administrator Main Window

The Object Tree displays the hierarchical relationship between LSM objects. Each object represents a group of components of the same type. When you select the icon to the left of an object, components of that type appear in the right pane. You can expand objects (by clicking on the plus sign) to display their hierarchy.

The Object Table displays component objects that belong to the currently selected object in the Object Tree. The Object Table is dynamic and constantly updates its contents to reflect changes to the system.

The Command Launcher displays a list of tasks that you can perform on objects. When you click on a task in the Command Launcher list, the task starts and the dialog box for the task appears.

The menu bar contains the following menus:

The toolbar consists of the following buttons that provide access to the following windows:

9.2    Running Tasks

You perform most tasks by selecting objects or tasks, then provide information in resulting dialog boxes. You run tasks using:

9.2.1    Selecting Objects

To select a single object, click on it. To deselect the object, click on it again.

To select or deselect multiple objects, hold down the Control key while selecting the objects. The objects that you select do not have to be adjacent.

To select a range of adjacent objects, select the first object and then hold down the Shift key while selecting the last object in the range. You can also select multiple adjacent objects by dragging the mouse over the desired objects while pressing the Shift key.

9.2.2    Using the Menu Bar

You can launch tasks from the Console and Selected menus in the menu bar. Choose New from the Console menu to create new LSM volumes, disk groups, and file systems. The context-sensitive Selected menu launches tasks on a selected object.

For example, to change a volume name, select Volumes in the Object Tree and the volume you want to rename in the Object Table. From the Selected menu, choose Rename, enter information in the Rename Volume dialog box, and click on the OK button.

9.2.3    Using the Right Mouse Button to Display a Pop-Up Menu

Click on the right mouse button to access a context-sensitive pop-up menu to display common task information that you can apply to the selected object. Additional tasks are available through the menu bar or the Command Launcher.

For example, to create a new volume in a disk group, select Disk Groups in the Object Tree, right click on the disk group in the Object Table, choose New Volume from the pop-up menu, enter information in the New Volume dialog box, and click on the OK button.

9.2.4    Using the Command Launcher

The Command Launcher window contains a list of objects and associated tasks. To display the Command Launcher, choose Command Launcher from the Window menu. To hide the Command Launcher, choose Command Launcher again from the Window menu.

To perform a task on a specific type of object, select the appropriate object-command combination from the Command Launcher list. For example, to create a volume, choose Volume-New from the Command Launcher, enter information in the New Volume dialog box, and click on the OK button.

Caution

The Command Launcher does not restrict context. If you choose inappropriate commands and ignore the warning dialogs, you can perform operations that may result in permanent loss of data.

9.2.5    Exiting the Storage Administrator

To exit the Storage Administrator client, from the Console menu, choose Exit. If you choose Close and there is no other Storage Administrator Main window open, the system exits.

9.3    Disk Management

The following sections describe disk management tasks that you can complete using the Storage Administrator.

9.3.1    Adding a Disk

When you add a disk for use with the LSM software, the disk is either initialized or encapsulated. If the disk is not set up, initialize it. If you want to use a disk with partitions that are in use with the LSM software, encapsulate it. Encapsulation preserves any existing data on the disk in the form of volumes. Initialization destroys any existing data on the disk. Initialized disks are placed in the free disk pool and are available to add to disk group.

Follow these steps to add a disk for use with the LSM software:

  1. In the Object Tree, select Disk and in the Object Table, choose a disk to be placed under LSM control.

    If the disk is new and does not show up in the Object Table when you highlight Disks, click on the Storage Administrator (at the top of the Object Tree) and choose Scan Disks from the Selected menu.

    In the Scan Disks dialog box, click on the Ok button to begin the search. From the Object Tree displayed, select a disk to add.

  2. From the Selected menu, choose Add.

    The Add Disk dialog box is displayed:

  3. In the Add Disk(s) dialog box:

    1. If the correct disk device name is not displayed, type the disk device name or click on Browse to select the disk. You can type more than one name separated by spaces.

    2. Specify where to add disks:

      To add disks to an existing disk group, select Existing Disk Group. Type the disk group name in the Disk Group Name field or click on Browse to select a disk group.

      To add disks to a new disk group, select New Disk Group. Type the name of the new disk group in the Disk Group Name field. The new disk group is created.

    3. To add disks to an existing disk group (that is, one that is not deported):

      Specify the LSM disk names for the disks, type a disk name in the Disk Name(s) field. This name must be unique within the disk group. If no LSM disk name is specified, the Storage Administrator assigns a default name to the disk.

      Enter a comment if desired.

    4. To place disks in the free disk pool, select Free Disk Pool. Disks in the free disk pool are under LSM control (initialized) but do not belong to a disk group and cannot be used to create volumes.

    5. Click on the Ok button.

You can add one or more unused disks to a disk group and designate them as hot-spares. If an I/O failure occurs, hot-relocation automatically relocates any redundant (mirrored or RAID5) subdisks to the spare disk(s) and restores the affected LSM objects and data. You are notified of the failure and relocation details by electronic mail. See Chapter 7 for more information on the hot-spare feature.

If you designate a hot-relocation spare, provide at least one per disk group. In the event of disk failure, any disk in the disk group can write to it. This will not work across disk groups.

To add a disk as a hot-relocation spare:

  1. In the Object Tree, select Disk Group and in the Object Table, select the LSM disk to be designated as a hot-relocation spare.

  2. Choose Properties from the Selected menu.

  3. In the Disk Properties window:

    1. Select the General tab.

    2. Select Spare.

    3. Click on the Ok button.

      Note

      The Properties dialog box is associated with the disk you have highlighted when you choose Properties. If you select a different disk from the Object Table, you must open a new Properties dialog box.

9.3.2    Evacuating a Disk

You can evacuate (or move) the contents of the volumes to other disks in the same disk group if there is sufficient free space. If no target disk is specified, LSM uses available disks with sufficient free space. Evacuating a disk is useful in the event of disk failure.

If the disk being evacuated contains part of a mirrored, striped, or RAID5 volume, do not move the contents of the disk to another disk containing a copy of the mirrored volume or part of the striped/RAID5 volume.

Follow these steps to evacuate a disk from LSM control:

  1. In the Object Tree, select Disk and in the Object Table, select the disk that contains the objects and data to be moved to another disk.

  2. From the Selected menu, choose Evacuate.

    The Evacuate dialog box is displayed:

  3. In the Evacuate Disk dialog box:

    1. If the correct disk name is not displayed in the Disk Name field, type the disk name or click on Browse then click on the Object Tree to select the disk.

    2. Type the name of the target disk to which you want to move the contents of the evacuated disk or click on Browse then click on the Object Tree to select one or more target disks.

      If you choose Browse, the total evacuated space (in kilobytes) is displayed so you can choose your target disk accordingly.

    3. Click on the Ok button.

9.3.3    Mirroring a Disk

Using LSM mirrored volumes (RAID1) is a common and effective way to improve data availability. If one disk fails on a mirrored volume, the data can still be accessed from the other copy, or plex. By mirroring data using disks connected to different controllers or buses, you can improve data availability even further because the data will still be accessible if a controller, cable, or storage cabinet fails. Therefore, it is helpful to understand a system's I/O hardware topology; that is, knowing which disk reside on which I/O bus.

Besides improving data availability, you can also use mirroring to significantly improve read performance because multiple reads to the same volume can be done simultaneously using the multiple copies of data. For example, read performance can potentially improve by a factor of two on a mirrored volume with two plexes because twice as many reads can be performed done at the same time.

Writes to the volume result in multiple, simultaneous write requests to each plex, so the time it takes to write to a volume may be slightly longer because of slight performance deviations between individual disks. For example, an individual write might take an additional 5 percent on average to complete because the volume write must wait for both writes to be completed on both plexes (disks).

You can improve overall I/O performance with mirroring because the larger performance gains for read often more than offset the slight degradation for writes. Comparing the number of read operations to the number of write operations on a volume using the volstat command can help give you better insight into whether mirroring can also help improve overall performance as well as provide higher data availability.

Because a volume can be changed to either add or remove a mirror on line with the LSM software, the overall performance implications can be measured on the actual I/O workload without stopping or disrupting service to a volume.

Mirrored volumes created with the volassist command will have dirty region logging (DRL) enabled by default. DRL is used with the LSM software mirrored volumes to keep track of which regions within the volume are dirty. While DRL may add a modest overhead to writes to the mirrored volume, DRL significantly reduces the time it takes to resynchronize a mirrored volume when rebooting a system after a failure because only the dirty regions within the volume need to be resynchronized instead of the entire volume.

While using a DRL with a mirrored volume is not required and has no affect on data integrity, DRL's dramatic reduction in resynchronization time often more than offset its overhead, so it is usually preferable to configure mirrored volumes with DRL enabled.

Note

In a TruCluster environment, the resynchronization overhead and time is significantly higher. Therefore, mirrored volumes should always be configured with DRL enabled when in a TruCluster environment.

Follow these steps to mirror all concatenated volumes on a disk:

  1. In the Object Tree, select Disk and in the Object Table, select the disk that contains the volumes to be mirrored onto another disk.

  2. From the Selected menu, choose Mirror.

    The Mirror Disk dialog box is displayed:

  3. In the Mirror Disk dialog box:

    1. If the correct disk name is not displayed in the disk name field, type the disk name or click on Browse to select the disk.

      If you choose Browse, total space being mirrored is displayed so you can choose your target disk accordingly.

    2. To specify the disks to contain the new mirrors, type the target disk name or click on Browse and complete the Target Disk dialog box.

    3. Click on Ok.

9.3.4    Placing a Disk On line

Placing a disk in an on line state restores access to a disk that is in an off line state. The disk is placed in the free disk pool and is accessible to LSM again. After bringing a disk back online, the disk must be added to a disk group before it can be used for volumes.

Only disks that are in an off line state can be placed in an on line state.

Follow these steps to place a disk on line:

  1. In the Object Tree, select Disk and in the Object Table and select the disk to be brought on line.

  2. From the Selected menu, choose Online.

    The Online Disk dialog box is displayed.

  3. In the Online Disk dialog box:

    1. If the correct disk name is not displayed, type the disk name or click on Browse to select the disk.

    2. Click on the Ok button.

9.3.5    Recovering Volumes on a Disk

A recovery operation depends on the types of volumes on the disk and includes starting disabled volumes, resynchronizing mirrors in mirrored volumes, and resynchronizing parity in RAID5 volumes. After successful recovery, the volumes should be available for use.

Alert icons and the Alert Monitor window may provide information when a volume recovery is needed.

If recovery of a volume is not possible, restore the volume from backup.

Follow these steps to recover all volumes on a disk:

  1. In the Object Tree, select Disk and in the Object Table and select the disk that contains the volumes to be recovered.

  2. From the Selected menu, choose Recover.

    The Recover Disk dialog box is displayed.

  3. In the Recover Disks dialog box:

    1. If the correct disk name is not displayed, type the disk name or click on Browse to select the disk.

    2. Click on the Ok button.

9.3.6    Removing a Disk from an LSM Disk Group

An LSM disk no longer in use can be removed from an disk group. Do not remove LSM disks that are currently in use (for example, contains subdisks for a volume), doing so can result in loss of data or of data redundancy.

After an LSM disk is removed from a disk group, it is still initialized for use with the LSM software. Therefore, after removing the disk from a disk group, it can be either immediately added to another disk group, removed from LSM, or left initialized for later use.

Follow these steps to remove a disk from an LSM disk group:

Follow these steps to remove a disk from an LSM disk group:

  1. In the Object Tree, select Disk and in the Object Table and select the disk to be removed.

  2. From the Selected menu, choose Remove.

    The Remove Disk dialog box is displayed:

  3. In the Remove Disk dialog box:

    1. If the correct disk name is not displayed, enter the disk name or click on Browse to select the disk.

    2. To move the contents of the disk to another disk(s) before the disk is removed, select Evacuate. Click Target Disks to specify the disk(s) to which you want the contents of the disk moved.

    3. Specify how to handle the disk after removal:

      To remove the disk from its disk group and place it in the free disk pool, select Return to Free Disk Pool. The disk remains under LSM control.

      To remove the disk from LSM control, select Return to Uninitialized State.

    4. Click on the Ok button.

9.3.7    Renaming a Disk

Because disk access names are defined by the operating system and media names are defined by you, you can only rename disk media names for disks in a disk group.

Follow these steps to rename the disk media name for an LSM disk:

  1. In the Object Tree, select Disk and in the Object Table, select the disk to be renamed.

  2. From the Selected menu, choose Rename.

    The Rename Disk dialog box is displayed.

  3. In the Rename Disk dialog box:

    1. If the correct disk name is not displayed, enter the disk name or click on Browse to select the disk.

    2. Enter the new LSM disk name.

    3. Click on the Ok button.

9.3.8    Replacing a Disk

You can replace an existing disk with a new physical disk, move volumes to the new disk, and attempt to recover any redundant (mirrored or RAID5) volumes on the disk. You cannot recover nonredundant volumes. You should restore nonredundant volumes from backup. If the disk being replaced is a boot disk, you can set up the new disk as a boot disk. You may need to replace a disk if the disk fails and needs to be removed and repaired.

If you replace a good disk, you need to remove the disk from its disk group and place it in the free disk pool before you replace the disk. If you replace a disk that has failed and is disconnected, you do not need to remove the disk from the disk group.

Follow these steps to replace a disk:

  1. In the Object Tree, select Disk and in the Object Table, select the disk to be replaced.

  2. From the Selected menu, choose Replace.

    The Replace Disk dialog box is displayed:

  3. In the Replace Disk dialog box:

    1. If the correct disk name is not displayed, enter the LSM disk name for the disk to be replaced or click on Browse to select the disk.

    2. Enter the physical disk name for the new (replacement) disk or click on Browse to select a disk.

    3. Click on the Ok button.

9.3.9    Scanning for New Disks

You can search your configuration for disks that are not under LSM control. Disks that are found are added to the free disk pool.

Follow these steps to scan for a new disk:

  1. Click on the scan button if it appears in the dialog box or click on Disk Scan in the Command Launcher.

    The Scan Disk dialog box is displayed.

  2. In the Scan Disks dialog box, click on the Ok button.

  3. To view disks that are found, click on the Free Disk Pool in the Object Tree.

9.3.10    Taking a Disk Off Line

You can place a disk in an off line state to prevent LSM from accessing it. You must remove a disk from its disk group before you take it off line. An off line disk remains unavailable until you restore access to the disk by placing it on line.

You place a disk in an off line state to protect it from unintentional use, for example, if attempts to access it may have a negative effect on the system. You cannot take a disk that is in use off line.

Follow these steps to take a disk offline:

  1. In the Object Tree, select Disk and in the Object Table, select the disk to be taken offline.

  2. From the Selected menu, choose Offline.

    The Offline Disk dialog box is displayed.

  3. In the Offline Disk dialog box:

    1. If the correct disk name is not displayed, enter the disk name or click on the Browse button to select the disk.

    2. Click on the Ok button.

9.4    Disk Group Management

The following sections describe disk group management tasks that you can complete using the Storage Administrator.

9.4.1    Adding a Disk to a Disk Group

To add a disk to a disk group, follow the instructions for adding a disk and in the Add Disk dialog box and specify an existing disk group as described in Section 9.3.1.

The LSM disk name must be unique within the disk group. If multiple disks are specified in the Disk Device(s) field and only one disk name is specified in Disk Name(s) field, LSM appends numbers to the disk name so that each disk name is unique within its disk group.

You must place disks that belong to a disk group in the free disk pool before you can add them to another disk group. You must add disks in the free disk pool to a disk group before you can use them to create volumes.

Disks must be on line before they can be added to a disk group or the free disk pool. Disks cannot be added to deported disk groups.

You must place the root disk in the root disk group (rootdg). If the root disk is placed in any other disk group, you cannot use the root disk to boot the system.

9.4.2    Creating a Disk Group

You must place disks in to a disk group before LSM can use them. The default disk group (rootdg) is usually created during LSM installation and always exists on a system running LSM. You can create additional disk groups to organize your disks into logical sets.

Each new disk group must contain at least one disk and its name must be unique. You can only use disks that are online and do not belong to a disk group to create a disk group.

Follow these steps to create a disk group:

  1. From the Console menu, choose New then Disk Group.

    The New Disk Group dialog box is displayed:

  2. In the New Disk Group dialog box:

    1. Enter the name of the disk group to be created. Click View to view the names of existing disk groups.

    2. To set up any new disks on the system, click Scan Disks. This runs the disk setup commands appropriate for the operating system.

    3. Select the disk devices to be placed in the new disk group or click on Browse to select the devices.

    4. There are two options:

      To specify the LSM disk name for the disk, enter a disk name in the Disk Name(s) field. If no LSM disk name is specified, Storage Administrator assigns a default name to the disk.

      Enter a comment if desired.

    5. Click on the Ok button.

9.4.3    Deporting a Disk Group

After a disk group is created, the LSM software automatically imports it for use whenever the system is booted.

To disable access to a disk group, you deport the disk group. All the volumes within the disk group should be stopped before deporting the disk group.

Follow these steps to deport a disk group:

  1. Select the Storage Administrator (at the top of the Object Tree).

  2. From the Selected menu, choose Deport Disk Group.

    The Deport Disk Group dialog box is displayed:

  3. In the Deport Disk Group dialog box:

    1. If the correct disk group name is not displayed, enter the disk group name or click on Browse to select the disk group.

    2. Use the following Expert Options with caution:

      To change the name of the disk group at deport, enter a new disk group name in the New Name field.

      To set up a host machine to import the deported disk group at reboot, enter the host ID in the New Host field

    3. Click on the Ok button.

9.4.4    Importing a Disk Group

You can import a disk group to make a deported (inaccessible) disk group and its volumes accessible again. To import a deported disk group, you must know the disk group's former name and this disk group name must have remained unused. In addition, at least one disk formerly assigned to the deported disk group must remain unused. If all disks associated with a deported disk group were reused because the disk group was deported, that disk group cannot be imported.

Import may fail for a number of reasons. It may fail if the host cannot find one or more disks in the disk group. If the import fails because a disk has failed, you can import the disk group by selecting the Force Import expert option. If the import fails for another reason, a forced import can cause serious problems.

When you import a disk group, the system stamps its host ID on all disks in the disk group. A disk group import fails if one of the disks is stamped with a host ID that does not match the others. This ensures that dual-ported disks cannot be managed (and possibly corrupted) by two systems at the same time. If you are sure that the disk group is not in use by another host, you can clear the host IDs and import the disk group by selecting the Clear Host ID expert option.

Follow these steps to import a disk group:

  1. Select the Storage Administrator (at the top of the Object Tree).

  2. From the Selected menu, choose Import Disk Group.

    The Import Disk Group dialog box is displayed:

  3. In the Import Disk Group dialog box:

    1. If the correct disk group name is not displayed, enter the disk group name or click on Browse to select the disk group.

    2. Use the following Expert Options with caution:

      To start all volumes in the disk group at import, select Start All Volumes.

      To clear the existing host ID stamp on all disks in the disk group at import, select Clear Host ID. Do not use this option if another host is using any disk(s) in the disk group.

      To force the disk group import when the host cannot access all disks in the disk group, select Force Import. Use this option with caution.

    3. Enter the name of the disk group to be imported or click on Browse to select the disk group.

    4. To change the name of the disk group at import, enter a new disk group name in the New Name field. To indicate that the name change is temporary, select Use New Name as Temporary. If you indicate a temporary name change, the original name is returned when the system is rebooted.

    5. Click on the Ok button.

9.4.5    Destroying a Disk Group

You can destroy a disk group permanently to remove the group from LSM control. It reinitializes all of the disks in the disk group as empty disks and places them in the free disk pool for reuse. You cannot destroy a disk group if any volumes in that disk group are in use. When a disk group is destroyed, the volumes in the disk group are removed.

Destroy a disk group only if you are sure that you no longer need the volumes and data in the disk group. Because the last disk in an existing disk group cannot be removed, destroying a disk group is a way to free the last disk in a disk group for reuse.

You cannot destroy the rootdg disk group.

Follow these steps to destroy a disk group:

  1. Select the Storage Administrator (at the top of the Object Tree).

  2. From the Selected menu, choose Destroy Disk Group.

    The Destroy Disk Group dialog box is displayed.

  3. In the Destroy Disk Group dialog box:

    1. Enter the name of the disk group to be destroyed or click on Browse to select the disk group.

    2. Click on the Ok button.

9.4.6    Moving a Disk Group

You can move a disk group (and LSM objects in that disk group) from one system to another. LSM and the Storage Administrator (server) must be running on both systems.

Follow these steps to move a disk group from one system to another:

  1. Unmount and stop all volumes in the disk group to be moved.

  2. Follow the instructions in Section 9.4.3 to deport the disk group to be moved to the other system.

  3. Attach all of the physical disks in the disk group to the new system.

  4. On the new system, follow the instructions in Section 9.4.4 to import the disk group.

  5. Select the Storage Administrator (at the top of the Object Tree) and from the Selected menu, choose Scan Disks to set up the newly attached disks on the system. This runs the disk setup commands appropriate for the operating system.

  6. Follow the instructions in Section 9.4.7 to restart and recover all volumes in the disk group on the new system.

9.4.7    Recovering Volumes in a Disk Group

You can recover volumes in a given disk group. The recovery operations depend on the types of volumes in the disk group and include starting disabled volumes, resynchronizing mirrors in mirrored volumes, and resynchronizing parity in RAID5 volumes. After successful recovery, the volumes are available for use.

Alert icons and the Alert Monitor window may provide you information to know when volume recovery is needed.

In some cases, recovery may not be possible. If the volume recovery fails, you can attempt to restore the volume from backup.

Follow these steps to recover all volumes in a disk group:

  1. In the Object Tree, select Disk Group and in the Object Table, select the disk group containing the volumes to be recovered.

  2. From the Selected menu, choose Recover

    The Recover Disk Groups dialog box is displayed.

  3. In the Recover Disk Groups dialog box:

    1. Enter the name of the disk group to be recovered or click on Browse to select the disk group.

    2. Click on the Ok button.

9.4.8    Renaming a Disk Group

You can rename a disk group. If volumes in the disk group are in use (mounted), the disk group is not renamed. Renaming a disk group updates the /etc/fstab file.

The new disk group name must be unique.

Follow these steps to rename a disk group:

  1. In the Object Tree, select Disk Group and in the Object Table, select the disk group to be renamed.

  2. From the Selected menu, choose Rename.

    The Rename Disk Group dialog box is displayed:

  3. In the Rename Disk Group dialog box:

    1. If the correct disk group name is not displayed, enter the disk group name or click on Browse to select the disk group.

    2. Enter the new name for the disk group

    3. Click on the Ok button.

9.5    Subdisk Management

Subdisks are created as the result of creating a volume. You cannot use the Storage Administrator to create subdisks.

The following sections describe the subdisk management tasks that you can complete by using the Storage Administrator:

9.5.1    Joining Subdisks

You can join two or more subdisks together to form a single, larger subdisk. Subdisks can only be joined together if they belong to the same volume and occupy adjacent regions of the same disk and mirror. The joined subdisk can retain the name of one of the subdisks being joined.

For a striped volume, the subdisks must be in the same column.

Follow these steps to join subdisks:

  1. In the Object Tree, select Volume and in the Object Table, select the volume with the subdisks to be joined.

  2. From the Selected menu, choose Show Layout.

  3. In the Volume Layout Details window, hold down the Shift key and click to select the subdisks to be combined. Subdisks must be contiguous.

    1. From the Selected menu, choose Join.

      The Join Subdisk dialog box is displayed:

    2. In the Join Subdisks dialog box:

      Enter the name of the disk group that contains the subdisks to be joined.

      If the correct subdisk names are not displayed, enter the subdisk names or click on Browse to select the subdisks. Specify at least two subdisk names separated by a space.

      Enter the name of the new, combined subdisk.

      Click on the Ok button.

  4. Close the Volume Layout Details window. The new volume layout can be viewed by reopening the window.

9.5.2    Moving a Subdisk

You can move portions of a volume to a different disk to improve performance. The disk space occupied by the original subdisk is returned to the free space pool.

Do not move a subdisk in a mirrored, striped, or RAID5 volume to a disk that already contains a copy or part of that volume.

If this task fails and leaves some unused subdisks (that is, subdisks that are not associated with a volume) on the system, you can use the Remove Subdisk task to free the space occupied by the unused subdisks.

Follow these steps to move a subdisk:

  1. In the Object Tree, select Volume and in the Object Table, select the volume with the subdisk to be moved.

  2. From the Selected menu, choose Show Layout.

  3. In the Volume Layout Details window, select the subdisk to be moved to another disk.

    1. From the Selected menu, choose Move.

      The Move Subdisks dialog box is displayed:

    2. In the Move Subdisks dialog box:

      Enter the name of the disk group that contains the subdisk to be moved.

      If the correct source subdisk name is not displayed, enter the subdisk's name or click on Browse to select the subdisk.

      Enter the name of the target disk to which the subdisk should be moved or click on Browse to select a disk.

      Choose the Move Policy to specify whether the subdisk can be split into smaller subdisks that fit in available space(s) on the target disk. The One to One options do not split the subdisk(s). The Split as Required option allows the subdisk(s) to be split if needed.

      Specify the minimum disk offset for the subdisk. Enter the offset in the Starting Offset for Gap Search field.

    3. Click on the Ok button.

  4. Close the Volume Layout Details window. You can view the new volume layout by reopening the window.

9.5.3    Removing a Subdisk

You can remove a subdisk that is not associated with a volume. This returns the disk space occupied by unused subdisks to the free space pool.

Follow these steps to remove a subdisk:

  1. In the Object Tree, select Disk and in the Object Table, select the disk with the subdisk to be removed.

  2. From the Selected menu, choose Show Layout.

  3. In the Volume Layout Details window, select the subdisk(s) to remove:

  4. From the Selected menu, choose Remove.

    The Remove Subdisks dialog box is displayed:

    In the Remove Subdisks dialog box:

    Enter the name of the disk group that contains the subdisks to be removed.

    If the correct subdisk names are not displayed, enter the subdisk names or click on Browse to select the subdisks.

    Click on the Ok button.

  5. Close the Volume Layout Details window. You can view the new volume layout by reopening the window.

9.5.4    Splitting a Subdisk

You can divide a subdisk into two or more subdisks. Once split, the smaller subdisks can be moved elsewhere or rejoined later. This is useful for reorganizing volumes or for improving performance. The original subdisk must contain a sufficient number of sectors for the specified split to work.

The name of the first subdisk remains the same as the selected subdisk. Other subdisks are automatically named by the Storage Administrator. The new, smaller subdisks occupy the same regions of the disk that the original subdisk occupied.

A log subdisk cannot be split.

Follow these steps to split a subdisk into multiple subdisks:

  1. In the Object Tree, select Volume and in the Object Table, select the volume with the subdisk to be split.

  2. From the Selected menu, choose Show Layout.

  3. In the Volume Layout Details window, select the subdisk to be split into multiple subdisks.

    1. From the Selected menu, choose Split.

      The Split Subdisk dialog box is displayed:

    2. In the Split Subdisk dialog box:

      If the correct subdisk name is not displayed, enter the subdisk's name or click on Browse to select the subdisk.

      Enter the number of subdisks into which the subdisk should be split. A subdisk can be split into two or more subdisks.

      Click on the Ok button.

  4. Close the Volume Layout Details window. You can view the new volume layout by reopening the window.

9.6    Volume Management

The following sections describe the volume management tasks that you can complete by using the Storage Administrator. Most tasks described in this section are appropriate only for UFS.

AdvFS file domains (file systems) operate differently from UFS file systems. Because the volume is part of the AdvFS file domain, once you assign a volume to AdvFS, it is out of the control of the Storage Administrator. Therefore, you cannot stop, remove, rename, mount, or unmount an AdvFS volume. That is, you cannot use the Storage Administrator to perform tasks that compromises the integrity of the AdvFS file domain.

In a cluster, AdvFS file systems are supported in all modes; UFS file systems are supported in read-only mode.

9.6.1    Adding a File System to a Volume

You can place a new file system on an existing volume and mount the file system. If Mount at Boot is selected, the /etc/fstab file is automatically updated.

Follow these steps to add a file system to an existing volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to contain the file system.

  2. From the Selected menu, choose File System then New.

    The New File System dialog box is displayed:

  3. In the New File System dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on Browse to select the volume.

    2. Type the mount point for the file system. The mount point must be an absolute pathname (that is, it must begin with root (/)). If the path specified for the mount point does not exist, it is created.

    3. Select Mount at Boot if desired.

    4. Select the file system type:

      If you select AdvFS, enter the Domain Name and the Fileset name.

      To add a volume to an existing AdvFS domain, you must have an AdvFS Advanced Utilities license. If you supply an existing domain name and do not have a license, an error message is displayed. You can create a new domain for the volume without the Advanced Utilities license.

      If you select UFS, you can also select Extra Options in the Mkfs Details dialog box. Click on Help in the Mkfs Details dialog box for more information. This option is not available in a cluster.

    5. Click on Mount Details to make the file system read only or to add extra options. Click on Help in the Mount Details dialog box for more information.

    6. Click on the Ok button.

9.6.2    Adding a Log to a Volume

You can add a log to a mirrored or RAID5 volume.

When you add a log to a mirrored volume, dirty region logging (DRL) is activated for that volume. DRL uses the log to track the regions of the volume that change due to I/O writes. If a system failure occurs, DRL uses the information in the log to recover only the portions of the volume that need recovery. This speeds up recovery time for mirrored volumes.

For DRL to be in effect, a mirrored volume must have at least one DRL log. You can create additional DRL logs (on different disks) to mirror the DRL information.

A RAID5 volume log speeds up the resynchronization time for RAID5 volumes after a system failure. A RAID5 log maintains a copy of the data and parity being written to the volume at any given time. If a system failure occurs, LSM can replay the RAID5 log to resynchronize the volume. This copies the data and parity that was being written at the time of failure from the log to the appropriate areas of the RAID5 volume.

You can create multiple RAID5 logs (on different disks) to mirror the log information. Ideally, each RAID5 volume should have at least two logs.

Follow these steps to add a log to a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to contain the log.

  2. From the Selected menu, choose Log then Add.

    The Add Log dialog box is displayed:

  3. In the Add Log dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on the Browse button to select the volume.

    2. To place the log on a specific disk, enter the name of the disk in the Disk Name field or click on Browse to select a disk.

    3. Click on the Ok button.

9.6.3    Adding a Mirror to a Volume

You can create a mirror (copy) of a volume on a disk that is not being used. Once mirrored, the data in the volume is redundant. If a disk fails, the data remains available on the surviving mirror. A volume can have multiple mirrors, but each must reside on a separate disk. Sufficient disk space must be available. You cannot mirror a RAID5 volume.

You can only use disks in the same disk group to create a new mirror. If no disks are assigned, LSM uses available disk space to create the mirror. Adding a mirror requires resynchronization, so this task may take some time.

A volume can contain up to 32 mirrors.

Follow these steps to add one or more mirrors to an existing volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to be mirrored.

  2. From the Selected menu, choose Mirror then Add.

    The Add Mirror dialog box is displayed:

  3. In the Add Mirror dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on Browse to select the volume.

    2. Click on the Layout button to:

      Specify the layout for the mirror: concatenated or striped and, if striped, unit size.

      Add more than one mirror and supply comments.

    3. Click on the Assign Disks button to place the mirror on a specific disk.

    4. Click on the Ok button.

9.6.4    Disabling a Mirror in a Volume

You can disable a mirror to temporarily detach the mirror from its volume. However, this can result in a loss of data redundancy because the mirroring process is not occurring. A detached mirror is inaccessible for reads and writes, but is still associated with its volume.

Once disabled, the mirror remains detached from its volume until you either reattach the mirror or restart the volume. If a volume only has two mirrors and one mirror is disabled, the volume is not redundant while the mirror is disabled.

The last mirror in a volume cannot be disabled.

When a volume is restarted, any disabled (detached) mirrors are reattached to the volume automatically.

Follow these steps to disable a mirror in a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume that contains the mirror to be disabled.

  2. From the Selected menu, choose Mirror then Disable.

    The Disable Mirror dialog box is displayed:

  3. In the Disable Mirror dialog box:

    1. Select the mirror to be disabled.

    2. Click on the Ok button.

9.6.5    Repairing a Mirror in a Volume

You can repair a disabled mirror and reattach it to its volume. Repairing a mirror involves copying data from an active mirror on the volume to the mirror being attached. Once attached, the mirror is accessible for reads and writes. This task recovers the mirror so that it has the same contents as other mirrors in the volume.

Alert icons and the Alert Monitor window may provide you with information when a mirror needs to be repaired.

Depending on the amount of data in the volume, this task may take some time.

Follow these steps to repair a mirror:

  1. In the Object Tree, select Volume and in the Object Table, select the volume that contains the mirror to be repaired.

  2. From the Selected menu, choose Mirror then Repair.

    The Repair Mirror dialog box is displayed:

  3. In the Repair Mirror dialog box

    1. Select the mirror to be repaired.

    2. Click on the Ok button.

9.6.6    Checking a File System on a Volume

Checking a file system only applies to UFS file systems because it uses the fsck utility, which is not compatible with AdvFS file systems. You can check the file system with or without repairing it. Checking a file system may take some time.

If you are running a cluster, UFS file systems are mounted read-only, so the Storage Administrator cannot check the file system.

Follow these steps to check a UFS file system on a volume:

  1. From the Object Tree, select File System and in the Object Table, select a UFS file system.

  2. From the Selected menu, choose Check.

    The Check File System dialog box is displayed:

  3. In the Check File System dialog box:

    1. If the correct file system name is not displayed, enter the file system name or click on Browse to select the file system.

    2. Choose the fsck option:

      Check with no repair

      Check and repair

    3. Choose the mount option:

      Unmount and check then mount

      Unmount and check

    4. Click on the Ok button.

9.6.7    Creating a Copy of a Volume Data

You can create a shapshot (temporary mirror) of a volume that tracks volume activity. In a cluster, you cannot take a snapshot of the clusterwide root.

Follow these steps to create or stop a volume snapshot:

  1. From the Object Tree, select Volume and in the Object Table, select a volume on which to collect data.

  2. From the Selected menu, choose Snapshot.

    The Volume Snapshot dialog box is displayed:

  3. In the Volume Snapshot dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on Browse to select the volume.

    2. If the correct snapshot name (volume) is not displayed, enter the volume name or click on Assign Disks to select the volume.

  4. You can create the mirror, start, or stop it:

9.6.8    Creating a Volume

You can create a volume that is less than or equal to the available free space on the disk(s). If no disks are assigned, the Storage Administrator uses available space on disks in the selected disk group.

The data in a striped or concatenated volume is not protected against disk failure unless the volume is mirrored.

Follow these steps to create a volume:

  1. From the Console menu, choose New then Volume.

    The New Volume dialog box is displayed:

  2. In the New Volume dialog box:

    1. If the correct disk group name is not displayed, enter the disk group name or click on Browse to select the disk group.

    2. Accept the default new volume name or enter a new volume name.

    3. Enter a comment if desired.

    4. Enter the volume size.

      To specify a size unit, attach an s (sectors), k (kilobytes), m (megabytes), or g (gigabytes) to the size. The default size unit is sectors.

      To determine the largest possible size for the volume, click Maxsize. Units are displayed in kilobytes.

    5. Choose the volume layout:

      Concatenated

      Striped - Enter the number of columns and stripe unit size.

      RAID5 - Enter the number of columns and stripe unit size. This option is not available in a cluster.

    6. Check the Enable Logging box to create logging for the new volume.

    7. If you have chosen a concatenated or striped volume, you can choose to mirror it.

      To mirror the volume, select Mirrored. In the Total Number of Mirrors field, type the total number of mirrors for the volume. Note that each plex is a mirror, so if you create a volume and one mirror of that volume, the total number of mirrors is 2.

    8. To place the volume on a specific disk, click Assign Disks. Select the disk you want to use from the Space Allocation - New Volume dialog box and click on the Ok button.

    9. To place a file system on the volume, click Add File System.

    10. In the Add File System dialog box:

      1. Type the mount point for the file system. The mount point must be an absolute pathname (that is, it must begin with root (/)). If the path specified for the mount point does not exist, it will be created.

      2. Select Mount at Boot if you want the /etc/fstab file automatically updated and the file system mounted at reboot.

      3. Select the file system type. If you select AdvFS, enter the Domain Name and the Fileset name. If you select UFS, you can also select Extra Options in the Mkfs Details dialog box. This option is not available in a TruCluster environment.

      Click on Mount Details to make the file system read only or to add extra options. Click on Help in the Mount Details dialog box for more information.

      Click on the Ok button to close the Add File System dialog box.

    11. Click on the Ok button.

9.6.9    Mounting a UFS File System on a Volume

You can mount a UFS file system that already exists on a volume. If Mount at Boot is selected, the /etc/fstab file is automatically updated.

Follow these steps to mount a file system on an existing volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume that contains the UFS file system to be mounted.

  2. From the Selected menu, choose File System then Mount.

    The Mount File System dialog box is displayed:

  3. In the Mount File System dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on the Browse button to select the volume.

    2. Enter the mount point for the file system. The mount point must be an absolute pathname (that is, it must begin with root (/)). If the path specified for the mount point does not exist, it is created.

    3. Select Mount at Boot if desired.

    4. Click on Mount Details to make the file system read only or to add extra options. Click on Help in the Mount Details dialog box for more information.

    5. Click on the Ok button.

Note

AdvFS filesets do not appear as mounted on volumes because AdvFS does not associate filesets with a specific volume.

9.6.10    Preparing to Restore a Volume from Backup

To reload a volume from backup, you can stop the volume, set the volume to an uninitialized state, and restart the volume (without resynchronizing the volume's mirrors). This procedure will not work for an AdvFS file domain.

If the volume contains a mounted UFS file system, you must unmount the file system before you proceed. This task does not remount the file system.

This procedure is useful for disaster recovery. If a volume's data is corrupted and you need to restore the volume from backup, this procedure prepares the volume for restoration.

Follow these steps to prepare a volume to restore it from backup:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to be restored from backup.

  2. From the Selected menu, choose Prepare For Restore.

    The Prepare Volume For Restore dialog box is displayed.

  3. In the Prepare Volume For Restore dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on Browse to select the volume.

    2. Click on the Ok button.

9.6.11    Recovering a Volume

You can recover a volume. The recovery operations depend on the type of volume and include starting disabled volumes, resynchronizing mirrors in mirrored volumes, and resynchronizing parity in RAID5 volumes. After successful recovery, the volume should be available for use.

Alert icons and the Alert Monitor window may provide information when a volume recovery is needed.

In some cases, recovery may not be possible. If the volume recovery fails, you can attempt to restore the volume from backup.

Follow these steps to recover a failed volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to be recovered.

  2. From the Selected menu, choose Recover.

    The Recover Volume dialog box is displayed.

  3. In the Recover Volume dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on the Browse button to select the volume.

    2. Click on the Ok button.

9.6.12    Removing a Volume

Removing a volume destroys all of the data in that volume. Only remove a volume if you are sure that you do not need the data in the volume (or the data is backed up elsewhere). When a volume is removed, the space it occupied is returned to the free space pool.

Removing a volume that has a file system on it will only work if the file system is UFS.

Follow these steps to remove a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume remove.

  2. From the Selected menu, choose Remove.

    The Remove Volume dialog box is displayed.

  3. In the Remove Volume dialog box, click on the Yes button to remove the volume.

9.6.13    Removing a Log from a Volume

You can remove a DRL log or a RAID5 log from a volume.

If you remove a volume's only log, logging (either DRL or RAID5 logging) is no longer in effect for that volume. If logging is disabled, recovery time increases.

Follow these steps to remove a log from a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume that contains the RAID5 or DRL log to be removed.

  2. From the Selected menu, choose Log then Remove.

    The Remove Log dialog box is displayed:

  3. In the Remove Log dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on the log table to select the volume.

    2. Click on the Ok button.

9.6.14    Removing a Mirror from a Volume

Removing a mirror for a volume breaks the link between the mirror and its volume and returns the mirror's disk space to the free space pool for reuse. However, this may leave the volume unmirrored and unprotected against disk failure.

If a volume only has two mirrors and one mirror is removed, the volume is no longer redundant. The last mirror cannot be removed from a volume, for that is equivalent to removing the volume.

Follow these steps to remove a mirror from a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume that contains the mirror to be removed.

  2. From the Selected menu, choose Mirror then Remove. The Remove Mirror dialog box is displayed.

  3. In the Remove Mirror dialog box:

    1. If the correct mirror name is not displayed, enter the mirror name.

    2. Click on the Ok button.

9.6.15    Renaming a Volume

When you rename a volume, the new name must be unique within the disk group. If the volume has a file system, renaming the volume automatically updates the /etc/fstab file and allows you to specify a new mount point for the file system. You cannot rename volumes that are part of an AdvFS domain.

Follow these steps to rename a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to be renamed.

  2. From the Selected menu, choose Rename.

    The Rename Volume dialog box is displayed:

  3. In the Rename Volume dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on Browse to select the volume.

    2. Enter the new name for the volume.

    3. Click on the Ok button.

9.6.16    Resizing a Volume

You can increase or decrease the size of a volume. If the volume contains a UFS file system, this procedure also resizes the file system, which destroys the data in it. A volume containing an unmounted file system cannot be shrunk.

You cannot resize an AdvFS file domain with the Storage Administrator. If you want to resize a domain, use the AdvFS command line commands addvol and rmvol. See AdvFS Administration for more information.

You can specify either the desired size or the amount of space to add to or subtract from the volume size. When a volume is shrunk, the resulting extra space is returned to the free disk pool. When the volume size is increased, sufficient disk space must be available. When increasing the size of a volume, LSM assigns the necessary new space from available disks.

Follow these steps to resize a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to be resized.

  2. From the Selected menu, choose Resize.

    The Resize Volume dialog box is displayed:

  3. In the Resize Volume dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on Browse to select the volume.

    2. To use a specific disk for the additional space, click Assign Disks and select the disk you want to use from the Space Allocation - Resize dialog box

    3. Specify one of the following:

      • To increase the volume size by a specific amount of space, use the Add By field to specify how much space should be added to the volume.

      • To decrease the volume size by a specific amount of space, use the Subtract By field to specify how much space should be removed from the volume.

      • To specify the new volume size, type the size in the Desired Size field.

      • To specify a size unit, attach an s (sectors), k (kilobytes), m (megabytes), or g (gigabytes) to the size. The default unit is sectors.

  4. Click on the Ok button.

9.6.17    Starting a Volume

You can start a volume. RAID5 is not supported in a cluster. If you are not running a cluster, starting a RAID5 volume enables the volume and resynchronizes parity, if necessary. Starting a mirrored volume enables the volume and resynchronizes the mirrors to ensure that they are consistent. When a volume is successfully restarted, the volume is again available for use.

Under normal circumstances, volumes are automatically started when the system reboots. You can restart a volume that you stopped manually or to attempt to restart a volume that was stopped in some other manner. If you cannot start a volume, the volume remains unusable. If the volume contains an AdvFS file domain, it cannot be started using the procedure described below.

Follow these steps to start a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the (stopped) volume to be started.

  2. From the Selected menu, choose Start.

    The Start Volume dialog box is displayed.

  3. In the Start Volume dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on the Browse button to select the volume.

    2. Click on the Ok button.

      The volume's state in the Object Table changes to Started.

9.6.18    Stopping a Volume

You can stop a volume. When you stop a volume, it is not available for use until you restart it . You cannot stop a volume if it is in use or it has a mounted file system. If the volume contains an AdvFS file domain, it cannot be stopped using the procedure described below.

Follow these steps to stop a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume to be stopped.

  2. From the Selected menu, choose Stop.

    The Stop Volume dialog box is displayed.

  3. In the Stop Volume dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on the Browse button to select the volume.

    2. Click on the Ok button.

      The volume's state in the Object Table changes to Stopped.

9.6.19    Unmounting a File System on a Volume

You can unmount a mounted UFS file system.

Follow these steps to unmount a file system on a volume:

  1. In the Object Tree, select Volume and in the Object Table, select the volume containing the file system to be unmounted.

  2. From the Selected menu, choose File System then Unmount.

    The Unmount File System dialog box is displayed.

  3. In the Unmount File System dialog box:

    1. If the correct volume name is not displayed, enter the volume name or click on Browse to select the volume.

    2. Click on the Ok button.