Maintenance Page

The Maintenance page appears when you choose Management > Maintenance from the menu bar.

The Maintenance page lets you perform maintenance tasks on existing units on the current controller (shown in the drop-down list on the menu bar), and lets you create new units by configuring available drives.

Rescan Controller

Use the Rescan Controller button to have 3DM scan the ports. Rescanning updates the list of available drives shown. It also updates the status of all ports, so if error conditions have been fixed, the status is updated to reflect that. This is useful in variety of maintenance tasks. For example, if you physically plug in a drive and want the controller to recognize the newly plugged in drive, use Rescan.

Note: If you unplug a drive without first removing it through 3DM, Rescan will not recognize it as gone unless the drive was in use or until it is required by the system. Always use the Remove link to remove a drive before unplugging it. This is required for controllers earlier than the 9550SX models, and recommended for 9550SX and 9590SE models.

Warning: Physically removing or adding drives which are not in hotswap carriers can result in a system hang or may even damage the system and the drive.

In 9000-series controllers, rescan checks all ports. It checks empty ports for newly plugged in drives. If those drives were previously part of a 3ware RAID configuration and they still have valid DCB (Disk Configuration Block) information on them, the controller tries to piece them back together into a working unit. If a working unit can be formed, it will appear in the Unit Maintenance list when the scan is complete, and the operating system will be notified of the unit. In Linux or FreeBSD, a device node will be associated with each unit created. In Windows, the device manager will reflect the changes under the disk drives icon. This process is known as importing drives.

If new drives do not have any data indicating they were previously part of a 3ware RAID configuration, they will appear in the Available Drives list.

In addition, if there is a unit with the status Inoperable before a rescan (for example, a RAID 5 unit missing 2 or more drives), and a rescan finds drives that complete the unit, the inoperable unit will become a valid unit.

Note: In 8000-series controllers, only JBOD units can be imported. Clicking Rescan Controller after removing units other than a JBOD unit (e.g. RAID 5) will cause the drives that make up the unit to be seen as individual available drives; they will not be pieced together to form a unit. This is unlike the 9000 series controllers, which will piece together the unit for you and display the drives as a unit. If you still want to use the drives together as a unit, restart the computer.. After you restart, the 8000 will again see the drives as a valid unit. Creating a new unit with those drives will destroy any existing data on them.

Unit Maintenance

The Unit Maintenance section of the page lists all existing units on the current controller, and displays summary information about them.

The top row shows information about the unit, while subsequent rows show summary information about each drive in the unit.

Unit Information

Unit Number. The unit number assigned to the unit by the firmware. Unit numbers for 9000 series are in sequential order. Unit numbers for the 7/8000 series will begin with the lowest port number of the unit. Use the checkbox next to the unit to select a unit before clicking one of the task buttons.

Type of Unit. Type of unit: RAID 0, RAID 1, RAID 5, RAID 10, RAID 50, Single Disk, Spare, or JBOD. If the unit has been given a unique name, it shows beneath the RAID type.

Name of Unit. User-assigned unique name of the unit. The default setting is blank.

Status. Operational status of the unit: Ok, Rebuilding, Initializing, Verifying, Migrating, Degraded, or Inoperable (missing drives). When Rebuilding, Initializing, Migrating, or Verifying, the percentage (%) complete is also shown. The percentage complete can be active or paused. To see whether this task is currently active or paused, click on the unit number to display the Unit Information page, which has that information.

Drive Information

Status. The status of the drive: OK, Not Supported, Not Present, and so forth. If you need help regarding a status displayed here, please contact Technical Support.

Remove Drive. The Remove Drive link removes a drive from the controller so that you can safely unplug it. In the Unit Maintenance section, this link is only provided for drives that can be safely removed without creating an inoperable unit. (For example, a RAID 5 missing 2 or more drives or a RAID 0 missing 1 or more drives would become inoperable.) If you remove a drive from a redundant unit, the unit will become degraded. Once a unit has become degraded, additional drives cannot be removed without making it inoperable, so no Remove Drive link will display.

Warning: Physically removing drives which are not in hotswap carriers can result in a system hang or may even damage the system and the drive.

Maintenance Task Buttons

Below the list of units, a row of task buttons lets you preform maintenance and configuration tasks related to the unit. Before clicking one of these buttons, select the appropriate unit.

Verify Unit. Puts the selected unit in verifying mode. If verify scheduling is enabled on the Scheduling page, the unit will not start actively verifying until the scheduled time, and the status will indicate "Verify-Paused." (The Unit Details page will indicate whether a unit is actively verifying.) If verify scheduling is not enabled, clicking Verify Unit begins the verification process.

If the unit you selected to verify is a redundant unit, the redundancy of the unit will be verified. For example it will check parity for a RAID 5 or check data consistency for a RAID 1. If the unit you checked is not a redundant unit, verify will do a surface scan of the media. During verification, I/O continues normally. For RAID 0, single disks, JBODs, and spares, there is only a slight performance loss. For redundant units, you can set the background task rate on the Controller Settings page to specify whether more processing time should be given to verifying or to I/O.

While a unit is verifying, the status changes to Verifying and a Stop Verify link appears in the right-most column of the Unit Maintenance table.

Note: If the unit has not previously been initialized and you click Verify Unit, the initialization process starts. Initialization cannot be halted, so no Stop Verify link appears. (Initialization can be paused, however, through Scheduling. Initialization follows the Rebuild schedule, so turning on scheduling for Rebuild will pause initialization, as well.) For more information about initialization, see About Initialization.

Rebuild Unit. Replaces a degraded drive in a degraded unit with an available drive and begins rebuilding the RAID. When you select a degraded unit and click Rebuild Unit, a dialog box listing available drives appears, so that you can select the drive you want to use. If the degraded unit has more than one degraded drive (for example, a RAID 10 where both mirrored pairs each have a degraded drive), you will repeat this process a second time.

If rebuild scheduling is enabled on the Scheduling page, the unit will not start actively rebuilding until the scheduled time, and the status will change to say "Rebuild-Paused." (The Unit Details page indicates whether a unit is actively rebuilding.) If rebuild scheduling is not enabled, the rebuild process will begin right away.

Migrate Unit. Reconfigures a unit while it is on-line. Migration can be used to change only the RAID level, to expand the capacity by adding additional drives, or to change the stripe size.

Caution: Once migration of a unit is started, it cannot be cancelled.

When you select a unit and click Migrate Unit, a dialog box appears which lists the drives in the unit and any additional available drives. In the dialog box are two drop-down menus, one for choosing the RAID level and one for choosing stripe size.

You can only migrate a unit to a RAID level that will be larger than the original unit. For example, you can migrate from a RAID 5 array with 4 drives to a RAID 0 with four drives but you cannot migrate from a RAID 5 with four drives to a RAID 10 with four drives.

After you have specified changes to the unit, the Unit Maintenance screen reflects your changes and shows the percentage of migration completed.

While the unit is migrating, you can still access the unit as normal but the performance will be lower. You can adjust the I/O rate with the radio buttons on the Controller Settings page.

Remove Unit. Removes a selected unit and allows you to unplug the drives and move the unit to another controller. The data on the unit remain intact.

Caution: Before you click Remove Unit, make sure the unit you are removing is unmounted and no I/Os are being issued. (For example, make sure you are not copying files to the unit, and make sure that there are no applications with open files on that unit.)

If a unit is not unmounted and you remove it, it is the equivalent of physically yanking a hard drive out from under the operating system. Resulting behavior depends on which operating system you are using and what kind of I/O is being done, however you will typically see really bad results, bad errors, and most likely a reset of the controller or a system hang.

To unmount a unit under windows, use Administrative Tools > Computer Management > Disk Management. In the Computer Management window, right-click on the partition and Remove the logical drive letter associated with the unit.

To unmount a unit under Linux, unmount the mount point to where the RAID unit is mounted. For example, if you want to remove unit 0 and you know that 0 corresponds to /dev/sdb, you should unmount all partitions for sdbx (where x is the number of the partition).

umount /dev/sdbx

For FreeBSD, the command would be

When you click Remove Unit, you will be asked to confirm that you want to proceed. When you confirm the removal, the unit number and information will be removed from 3DM. (Units created in the future can reclaim this unit number.)

The operating system is notified that the unit was removed. In Linux the device node associated with this unit is removed. In Windows the Device Manager will reflect the changes under the disk drives icon.

Information about the unit remains intact on the drives. This allows the drive or drives to be reassembled into a unit again on this controller, or if moved to another controller.

Warning: Physically removing drives which are not in hotswap carriers can result in a system hang or may even damage the system and the drive.

Delete Unit. Deletes the selected unit and allows you to use the drives to create another unit. The drives appear in the list of Available Drives.

Caution: Before you click Delete Unit, make sure the unit you are removing is unmounted and no I/Os are being issued. If a unit is not unmounted and you delete it, it is the equivalent of physically yanking a hard drive out from under the operating system. Resulting behavior depends on which operating system you are using and what kind of I/O is being done, however you will typically see really bad results, bad errors, and most likely a reset of the controller or a system hang.

Devices can be unmounted through the operating system. For details, see the discussion under Remove Unit, above.

Warning: When a unit is deleted, the data will be permanently deleted: the drives cannot be reassembled into the same unit. If you want to reassemble the drives on another controller and access the existing data, use Remove Unit instead of Delete Unit.

After deletion, the operating system is notified that the unit was deleted. In Linux the device node associated with this unit is removed. In Windows the Device Manager will reflect the changes under the disk drives icon.

Available Drives (to Create Units)

This section lists the drives on the controller which are not currently configured as part of a unit. The Port number, model, capacity, and status are all displayed, as they are for drives in existing units.

Remove Drive. The Remove Drive link removes a drive from the controller so that you can safely unplug it. Any drive in the Available Drives list can be removed.

Warning: Physically removing drives which are not in hotswap carriers can result in a system hang or may even damage the system and the drive.

Create Unit

Use the Create Unit button to create a unit for use on the current controller. Begin by selecting the drives you want to use in the list of Available Drives, and then click Create Unit. You will be prompted to select the unit Type, Name, Stripe size (if applicable), and unit policy settings.

A window shows the drives you selected, and lets you specify configuration settings.

Type. The drop-down list lists the possible RAID configurations for the drives selected in the list of Available Drives. Available configurations may include RAID 0, RAID 1, RAID 5, RAID 10, RAID 50, Single Disk, and Spare Disk. For information about these configurations, see 3ware Serial ATA RAID Controller User Guide.

When you are configuring a RAID 50 with twelve drives, an additional field appears, in which you select the number of drives per subunit-3, 4, or 6.

Stripe. The drop-down list of stripe sizes lists the possible stripe sizes for the configuration you selected in the RAID level drop-down.

The default stripe size of 64KB will give the best performance with applications that have many sequential reads and writes. A larger stripe size will give better performance with applications that have a lot of random reads and writes. In general, the smaller the stripe size, the better the sequential I/O and the worse the random I/O. The larger the stripe size, the worse the sequential I/O and the better the random I/O.

Write Cache, Auto-Verify, and Continue on Source Error during Rebuild. These check boxes let you set the policies for the unit. These policies can also be set and changed on the Controller Settings page. For details about these policies, see 3ware Serial ATA RAID Controller User Guide.

Note: If the configuration window disappears while you are selecting drives, 3DM 2 may have refreshed. Click Create Unit again. If desired, you can reduce the frequency with which information refreshes in 3DM 2, or disable refresh temporarily, on the 3DM 2 Settings page.

StorSave. This drop-down lets you specify the StorSave Profile to be used for the unit. Three profiles are available: Protection, Balanced, and Performance. For additional information, see Setting the StorSave Profile for a Unit.

Maintenance Page

Rescan Controller

Unit Maintenance

Unit Information

Drive Information

Maintenance Task Buttons

Available Drives (to Create Units)

Create Unit

Related Topics