Using volumes e-Learning course

When you create a volume, you can designate it as thin-provisioned. A thin-provisioned volume has a virtual capacity and a real capacity. Virtual capacity is the volume storage capacity that is available to a host. Real capacity is the storage capacity that is allocated to a volume copy from a storage pool. In a fully allocated volume, the virtual capacity and real capacity are the same. In a thin-provisioned volume, however, the virtual capacity is typically much larger than the real capacity. The real storage is only consumed when a new block is written to the volume.

Thin-provisioned volumes can also help simplify server administration. When additional real storage is required, administrators can manually or automatically expand the real storage that is allocated to the volume. Thin-provisioned volumes help to ensure efficient use of storage resources.

You can also enable the auto-expand option for trusted users, which allows their storage to grow automatically as needed. Auto-expansion allows users to expand physical storage automatically, without asking the system administrator. When auto-expand is disabled, the storage space must be managed carefully. If a volume needs more than the allotted amount of storage, that volume will be automatically taken offline. To help prevent this, you can set up a warning level on each volume. When that warning level is reached, you receive an email notification to manually allocate additional real storage to that volume.

Thin-provisioned volumes use only the space that is needed. This simplifies storage management for servers and applications. For example, you could assign four 1-terabyte volumes to a web mail server with four users. The users each believe they have one terabyte of space. However, using thin-provisioned volumes, you can create those volumes in a 2-terabyte storage pool. With thin-provisioned volumes, you can manage the storage so that the space used can expand as needed for each user. So if two users are using half a terabyte each, they believe they still have half a terabyte free. If the other users are using one-fourth of a terabyte, they believe they still have three-fourths of a terabyte free. In reality, only half a terabyte of space is left on the 2-terabyte storage pool. This half a terabyte is available to any of the users, if needed. If more than two terabytes is needed by the storage pool later, you will receive a warning and can add more physical storage to the pool. As you can see, with thin-provisioned volumes, users can have the space they want while you give them only the space they actually need.In our example, with auto-expansion enabled, each user can expand physical storage automatically. It should be noted however, that a single user might consume the whole 2-terabyte storage pool very quickly.

It is important to understand the data growth characteristics of the applications that use the volumes. The system must maintain extra metadata that describes the contents of thin-provisioned volumes. As a result, thin-provisioned volumes that grow quickly might return I/O rates that are slower than those with fully allocated volumes on the same managed disks.With thin-provisioned volumes, instead of assigning a volume with some capacity to an application and increasing that capacity as the needs of the application change, you can configure a volume with a large virtual capacity for the application, and then increase or shrink the real capacity as the application needs change, without disrupting the application or server.

A compressed volume is a special type of thin-provisioned volume. Instead of simply storing a new data block when it is written, the data is compressed as it is written to a volume, so that less real storage capacity is required. Compressed volumes are best used with data that has not already been compressed. For example, .MP3 files are already compressed, so storing those files on compressed volumes would not be efficient, but storing documents and database applications on compressed volumes would minimize storage demands. Note that compressed volumes are not available on all Lenovo Storage products.

To view general system capacity information, go to the Volume by Pools panel. Right click on a selected volume and choose Properties from the dropdown options. A table containing volume details is displayed containing volume status, capacity, preferred nodes, and other volume information.

To view data compression using the management GUI, go to the Pools Panel >MDisks by Pools. Select a storage pool, right click and choose Properties from the dropdown option. Click the View More Details arrow to expand the window. The volume allocation and compression savings for the storage pool is displayed.

Mirrored volumes can be created to enable a volume to remain online, even when some of the associated managed disks cannot be accessed. The host is aware of only the logical volume. However, with mirrored volumes, there are actually two volume copies in the background. Although not required, it is advisable to map the copied volume to a different storage pool, on a different external storage system. This ensures that the volume is fully protected against a storage system failure.

When a server writes to a mirrored volume, the system writes the data to both copies. When a server reads a mirrored volume, the system selects one of the copies to read from. If one of the mirrored volume copies becomes temporarily unavailable, for example, because the storage system that provides the storage pool is unavailable, the volume will still be available to the servers. The system monitors and catalogues which areas of the volume are written, and resynchronizes these areas when both copies are available.

Mirroring is most useful for reliability and migration. Also, remember that when you create a copy, you consume twice as much storage. As a result, using mirrored volumes can cost twice as much when provisioning storage.

With mirrored volumes, the volume copy is thin-provisioned. Although the two copies of a mirrored volume can have different characteristics (for example, one copy is compressed, and the other is not), each copy has the same virtual capacity as the volume. A mirrored volume has all of the capabilities of the primary volume copy and also the same restrictions.

Mirrored volumes can also offer an alternative method of data migration. One method is to start with a non-mirrored volume in the source storage pool. Then, add a copy of that volume in the target storage pool. Once the volume is synchronized, the original copy can be deleted from the source storage pool. You might also use this method to change a thin-provisioned volume into a non-thin-provisioned volume, or the other way around.

Mirrored volumes are also used for external storage system maintenance. Let us say you are running a critical application on a volume in a storage pool that is backed up by a RAID storage system. However, that system needs to be taken offline for maintenance, and you cannot afford to lose access to the application. So, you mirror the volumes in a different storage pool that resides in a different storage system. After your data has been copied and is safe, you can take the system offline to complete the necessary work. Host I/O is able to continue, because a copy of the volume that the application is using is still available on the new storage pool and system. After the maintenance is complete, you can bring the storage system back online, and begin to resynchronize your volume copies so that the changes made to the mirrored volume are also on the original volume. After this transfer is complete, the mirrored volume is no longer needed and can be deleted, or it can remain in place to provide extra redundancy for a possible storage system failure in the future.

To create volumes, go to the Volumes panel. Select Create Volumes. The Create Volume pane is displayed. Select the preset type. Next, select a pool and enter the number of pools, capacity and volume name.You can edit the name and size of the individual volumes after they have been added. Right click on the volume you wish to rename. Choose Rename from the dropdown option. Enter the new name and click the Rename button. The renamed volume will now be displayed.

You typically migrate data in order to move workloads from external storage systems that are to be removed or replaced. To migrate a volume using the management GUI, go to the Volumes panel. Alternately, you can go to the Volumes by Pool or Volumes by Host panels. Select the volume or volumes to be migrated. Select Actions, and choose Migrate to Another Pool from the dropdown options. The Migrate Volume Copy panel is displayed. Select a pool and click Migrate to complete this task.

You can move a volume to a new I/O group to balance the workload across the canisters in the system without stopping host activity to the volumes. Note that moving volumes among I/O groups can only be performed on volumes accessed from Fibre Channel hosts.

To move a volume using the management GUI, go to the Volumes panel, Volumes by Pool, or Volumes by Host panel. Select the volume or volumes to be moved. Select Actions > Move to Another I/O Group. The Move Volumes to New I/O Group wizard displays. Select the new target I/O group and preferred node. The wizard automatically detects new paths and removes the volume from the old I/O group. Review the summary information that displays, then click finish to complete the wizard.

In this e-Learning module, you learned about thin-provisioned volumes, compressed volumes for select Lenovo Storage family products, and mirrored volumes. You also learned how to create, migrate, and move volumes.

To learn more, see the other e-Learning modules and help topics in IBM Knowledge Center.