What is a Controller?Go to Table of Contents for Storage Management Online Help
Dell OpenManage™ Storage Management User's Guide
Controller-supported RAID Levels
Controller-supported Stripe Sizes
RAID Controller Read, Write, and Cache Policy
Cluster-enabled RAID Controllers
Integrated Mirroring and the PERC 4/IM Controller
Controller Properties and Tasks
This section describes the controllers supported by Storage Management. Use this section to understand the different controller features.
Most operating systems (OS) do not read and write data directly from the disks, but instead send read and write instructions to a controller. The controller is the hardware in your system that interacts directly with the disks to write and retrieve data. A controller has channels which are attached to one or more hard disks or an enclosure containing hard disks. RAID controllers can span the boundaries of the disks so as to create an extended amount of storage space -- or a virtual disk -- using the capacity of more than one disk.
Controllers also perform other tasks, such as initiating rebuilds, initializing disks, and so on. To complete their tasks, controllers require special software known as firmware and drivers. In order to function properly, the controller must have the required version of the firmware and drivers installed.
Storage Management supports different types of controllers. If your system has a supported controller, the controller is displayed by expanding the Storage object in the tree view. You can select the controller to display tabs for executing controller tasks and viewing controller properties.
Different controllers have different characteristics in the way they read and write data and execute tasks. It is helpful to understand these features to most efficiently manage your storage. The following sections describe the supported controllers and their features.
This section describes the RAID controllers supported by Storage Management. RAID controllers provide RAID functions such as creating virtual disks.
The PERC 2/SC, 2/DC, 3/SC, 3/DCL, 3/DC, 3/QC, 4/SC, 4/DC, 4e/DC, 4/Di, 4e/Si, and 4e/Di controllers are high-performance intelligent PCI-to-SCSI host adapters with RAID control capabilities.
The CERC ATA100/4ch controller is a PCI-to-IDE host adapter with RAID control capabilities. Each channel on the CERC ATA100/4ch supports 100 MB data transfer. Each channel on this controller supports a single device.
Each channel on the PERC 2/SC and PERC 2/DC controllers supports SCSI data transfer rates of up to 80 MB per second per channel and up to 160 MB per second per channel for the PERC 3/SC, 3/DCL, 3/DC, 3/QC, 4/Di, 4e/Si, and 4e/Di controllers. The channels on these controllers also support up to 15 devices.
Each channel on the PERC 4/SC, 4/DC, 4e/DC, 4/Di, 4e/Si, and 4e/Di controllers supports SCSI data transfer rates of up to 320 MB.
The PERC 2 RAID controller contains four channels. PERC 2/Si and PERC 3/Si are single integrated channel controllers. The PERC 3/Di is a dual integrated channel controller.
The PERC 4/IM controller provides integrated mirroring. This controller supports SCSI data transfer rates of up to 320 MB per second.
For more information about specific controller features, refer to the hardware manual that came with the controller.
 |
Note: Supported features may vary from controller to controller. |
Different controllers have different features. If you have more than one controller attached to your system, you may notice that the tasks displayed on the controller's Information/Configuration subtab are different for each controller.
Controllers may also have differences in their read, write, and cache policies as well as how they handle hot spares. You should be aware of these differences when creating virtual disks and assigning hot spares.
The following describes controller features and provides links to a more detailed explanation:
The following table indicates the RAID controllers supported by Storage Management. These controllers provide RAID functions such as creating virtual disks.
The table indicates which RAID levels are supported by each controller. Refer to the Comments column for information on how a controller implements a particular RAID level. For a list of supported controllers, see Supported Controllers.
For more information on storage concepts used in this section, see What Is RAID? and Choosing RAID Levels and Concatenation.
For information on controller features and virtual disks, see the following:
The following table indicates the stripe size supported by the RAID controllers when creating a virtual disk.
For more information on storage concepts used in this section, see What Is RAID? and Choosing RAID Levels and Concatenation.
For information on controller features and virtual disks, see the following:
When creating a virtual disk, you specify the read, write, and cache policies for the virtual disk. The following describes how each of these policies functions.
|
Note: Read, write, and cache policies are not supported on the CERC SATA1.5/2s controller. |
The read policies indicate whether or not the controller should read sequential sectors of the logical drive when seeking data.
These controllers have the following read policies:
These controllers have the following read policies:
The write policies specify whether the controller sends a write-request completion signal as soon as the data is in the cache or after it has been written to disk.
These controllers have the following write policies.
|
Note: Storage Management does not allow you to select the Write-Back policy for controllers that do not have a battery. This restriction protects a controller without a battery from the data loss that may occur in the event of a power failure. On some controllers, the Write-Back policy may be available in the controller BIOS even though it is not available in Storage Management. |
|
Note: Write-through is the default write policy setting when cluster mode is enabled. In cluster mode, the PERC 3/DC, 4/DC, and 4e/DC controllers only allow write-through caching. |
These controllers have the following write policies:
The Direct I/O and Cache I/O cache policies apply to reads on a specific logical drive. These settings do not affect the read-ahead policy. The cache policies are as follows:
Storage Management supports PERC 3/DC, 4/DC, and 4e/DC controllers that have Microsoft Cluster Server (MSCS) enabled.
A cluster refers to two or more servers that are connected so that their resources can be shared and accessed as if the clustered servers were a single machine. Clusters provide increased availability because when one server in the cluster experiences downtime, another server can take over the processing and workload requests of the failed server.
|
Note: Storage Management does not set resource ownership in a Microsoft Windows cluster configuration. |
After a failover of cluster resources, it is necessary to perform a rescan operation in order for Storage Management to display the most up-to-date information about shared resources.
When downloading firmware to a clustered controller, it is recommended that you shut down the other systems in the cluster first. When restarted, the other systems in the cluster should be able to see the firmware updates that you have applied.
If, however, you have downloaded firmware to a controller without first shutting down the other systems in the cluster, you may find that the other systems cannot see the firmware update until you restart the disk management service on those systems. For example, if you download firmware onto system A, and system B cannot see the firmware update, then restart the disk management service on system B.
If you are using a PERC 3/DC, 4/DC, or 4e/DC controller in a cluster configuration, you must shut down and power off the other systems in the cluster before creating or deleting the virtual disk. The following procedure describes the sequence of actions required to create or delete a virtual disk from a cluster-enabled controller. For the purposes of this procedure, the system on which you are creating or deleting the virtual disk is identified as "system A" and the other system in the cluster is identified as "system B."
The PERC 4/IM controller enables you to mirror a disk drive that resides internally in the server. This feature can be used to mirror a system's boot drive, ensuring that the system remains running in the event of a boot drive failure. The PERC 4/IM controller firmware maintains the mirrored data on both drives so that the system's CPU is not burdened with the extra processing required to maintain the mirrored data.
When implementing mirroring on a PERC 4/IM controller, you use the controller BIOS to create a virtual disk from hard disks. A hard disk is a physical disk that is directly visible to the operating system. Unlike creating a virtual disk on other controllers, the PERC 4/IM controller is able to implement a mirror for a hard disk that already contains data. The data is then copied to the mirror. Any data previously residing on the mirror is overwritten.
Once you have created the integrated mirror using the controller BIOS, the operating system sees the mirror as a virtual disk and a virtual disk object for the mirror is displayed in the Storage Management tree view. Expanding the virtual disks object displays the disks included in the mirror. These disks are now labeled "array disks" since they are part of a virtual disk and no longer individually visible to the operating system.
On PERC 3/SC, 3/DCL, 3/DC, 3/QC, 4/SC, 4/DC, 4e/DC, 4/Di, 4e/Si, 4e/Di, and CERC ATA100/4ch controllers, background initialization of a redundant virtual disk begins automatically within 0 to 5 minutes after the virtual disk is created. The background initialization of a redundant virtual disk prepares the virtual disk to maintain redundant data and improves write performance. For example, after the background initialization of a RAID 5 virtual disk completes, the parity information has been initialized. After the background initialization of a RAID 1 virtual disk completes, the array disks are mirrored.
The background initialization process helps the controller identify and correct problems that may occur with the redundant data at a later time. In this regard, the background initialization process is similar to a check consistency.
The background initialization should be allowed to run to completion. If cancelled, the background initialization will automatically restart within 0 to 5 minutes. Other processes such as read and write operations are possible while the background initialization is running. Other processes, such as creating a virtual disk, cannot be run concurrently with a background initialization. These processes cause the background initialization to cancel.
The SCSI controllers are non-RAID controllers that support SCSI devices. Because these controllers are non-RAID, they do not support virtual disks. You can manage these non-RAID controllers and their attached SCSI devices with Storage Management.
The Ultra SCSI controllers supported by Storage Management are a high voltage differential (HVD) SCSI drop-in controller that fits in a PCI slot on the server or is embedded in the server. This controller can be used to attach HVD devices such as an HVD tape drive.
The Ultra2 SCSI and Ultra160 SCSI controllers are low voltage differential (LVD) drop-in controllers that fit in a PCI slot on the server or are embedded in the controller. These controllers can be used to attach external enclosures and other LVD SCSI devices.
The following information applies to the SCSI controllers:
|
Note: Supported features may vary from controller to controller. |
This screen displays the status of the controller and the components attached to the controller.
Component status is indicated by the severity. A component with a Warning or Critical/Failure status requires immediate attention to avoid data loss if possible. It may be useful to review the Alert Log for events indicating why a component has a Warning or Critical status. For additional troubleshooting information, see Alert Messages and Troubleshooting.
The component status displayed on this screen reflects the status at a given point in time. If you believe the status has changed and wish to update this screen, select the controller's Information/Configuration tab and execute the Rescan task.
For information on the controller, see the following topics:
For information on attached components, see the following topics:
Use this window to view information about the controller and execute controller tasks.
The controller properties can vary depending on the model of the controller. Controller properties may include:
|
Property |
Definition | ||
|---|---|---|---|
| These icons represent the severity or health of the storage component. See Storage Component Severity for more information. | ||
Name | This property displays the name of the controller. | ||
State | This property displays the current status of the controller. Possible values are:
| ||
Firmware Version | This property displays the version of the controller's firmware. | ||
Driver Version | This property displays the version of the controller's driver.
| ||
Minimum Required Firmware Version | This property displays the minimum firmware version that is required by Storage Management. | ||
Number of Channels | This property displays the number of channels the controller has. Each channel can be attached to array disks or an enclosure. | ||
Rebuild Rate | The rebuild rate is the percentage of the system's resources dedicated to rebuilding a failed disk when a rebuild is necessary. See Set Rebuild Rate for more information. | ||
Alarm State | This property displays whether the controller's alarm is enabled or disabled. | ||
SCSI Initiator ID | This property displays the SCSI ID of the controller. The default value is 6. You can change the default value in the BIOS. In cluster mode, the value is 6 or 7. |
Do the following to execute a controller task:
|
Note: Different controllers support different features. For this reason, the tasks displayed on the Tasks drop-down menu can vary depending on which controller is selected in the tree view. |
A controller rescan updates configuration changes (such as new or removed devices) for all components attached to the controller. For information on when you may want to do a rescan, see Rescan to Update Storage Configuration Changes.
On non-RAID SCSI controllers, the operating system must be able to see any hardware changes before initiating the controller rescan. Otherwise, the hardware changes may not be reflected in the Storage Management graphical user interface (GUI). Refer to the operating system documentation for information on updating hardware changes. You can also reboot the system to update hardware changes with the operating system.
To rescan a controller:
Use the Create Virtual Disk task to launch the Create Virtual Disk Express Wizard. See Create Virtual Disk Express Wizard (Step 1 of 2) for more information.
Use the Enable Alarm task to enable the controller's alarm. When enabled, the alarm sounds in the event of a device failure.
Use the Disable Alarm task to disable the controller's alarm. When disabled, the alarm does not sound in the event of a device failure.
Use the Quiet Alarm task to quiet the controller's alarm once it is sounding. Once quieted, the alarm is still enabled in the event of a future device failure.
Use the Test Alarm task to test whether the controller alarm is functional. The alarm will sound for about two seconds.
|
Note: The Test Alarm task is only available on the CERC SATA1.5/6ch controller. |
Use the Set Rebuild Rate task to change the rebuild rate. See Set Rebuild Rate for more information.
Use the Reset Configuration task to erase all information on the controller, so that you can perform a fresh configuration. This operation destroys all virtual disks on the controller.
|
Caution: Resetting a configuration permanently destroys all data on all virtual disks attached to the controller. If the system or boot partition resides on these virtual disks, it will be destroyed. |
Use this task to export the controller log to a text file. See Export Log for more information.
For information on attached components, see the following topics:
The Set Rebuild Rate task changes the controller's rebuild rate.
During a rebuild, the complete contents of an array disk are reconstructed. The rebuild rate, configurable between 0% and 100%, represents the percentage of the system resources dedicated to rebuilding failed array disks. At 0%, the rebuild will have the lowest priority for the controller, will take the most time to complete, and will be the setting with the least impact to system performance. A rebuild rate of 0% does not mean that the rebuild is stopped or paused.
At 100%, the rebuild will be at the highest priority for the controller, will minimize the rebuild time, and will be the setting with the most impact to system performance.
|
Note: On the PERC 2/SC, 2/DC, 3/SC, 3/DCL, 3/DC, 3/QC, 4/SC, 4/DC, 4e/DC, 4/Di, 4e/Si, 4e/Di, and CERC ATA100/4ch controllers, the controller firmware also uses the rebuild rate setting to control the system resource allocation for the Check Consistency and Background Initialization tasks. The rebuild rate setting applies to these tasks in the same manner that it applies to the Rebuild task. |
Use the Reset Configuration task to erase all information on the controller so that you can perform a fresh configuration. This operation destroys all data and virtual disks on the controller and unassigns any hot spares.
You will need to completely reconfigure your storage after performing this operation.
|
|
Caution: Resetting a configuration permanently destroys all data on all virtual disks attached to the controller. If the system or boot partition resides on these virtual disks, it will be destroyed. |
The Export Log task exports the controller log to a text file. The text file is located in the Storage Management installation directory at the location displayed on the Export Log screen.
The log gives detailed information on the controller activities and can be useful for troubleshooting. By default, the log file is exported to the <install-directory>/sm directory where <install-directory> is the directory where you have installed Server Administrator.
Depending on the controller, the log file name will be either afa_<mmdd>.log or lsi_<mmdd>.log where <mmdd> is the month and date. For example, a log file exported on September 21 will be named either afa_0921.log or lsi_0921.log.
Exporting the controller log file clears the log file information from the controller log.
|
Note: The Export Log File command is not supported on the PERC 2/SC, 2/DC, 4/IM, CERC ATA100/4ch, and CERC SATA1.5/2s controllers. |
Click Export Log File when ready. To exit without exporting the controller log file, click Go Back To Controller Page.
Go to Table of Contents for Storage Management Online Help
