Storage architecture refers to how Data ONTAP utilizes storage systems to make data storage resources available to host or client systems and applications. Data ONTAP 7.0 and later versions distinguish between the physical layer of data storage resources and the logical layer that includes the file systems and the data that reside on the physical resources.
The physical layer includes disks, Redundant Array of Independent Disks (RAID) groups they are assigned to, plexes, and aggregates. The logical layer includes volumes, qtrees, Logical Unit Numbers (LUNs), and the files and directories that are stored in them. Data ONTAP also provides SnapshotTM technology to take point-in-time images of volumes and aggregates.
Storage systems use disks from a variety of manufacturers. All new storage systems use block checksum disks (BCDs) for RAID parity checksums. These disks provide better performance for random reads than zoned checksum disks (ZCDs), which were used in older storage systems. For more information about disks, see Understanding disks.
Data ONTAP organizes disks into RAID groups, which are collections of data and parity disks to provide parity protection. Data ONTAP 7.1 supports the following RAID types for NetApp storage systems (including the R100 series, the R200 series, the F87, the F800 series, the FAS200 series, the FAS900 series, and the FAS3000 series storage systems).
NetApp V-Series systems support systems that use RAID1, RAID5, and RAID10 levels, although the V-Series systems do not themselves use RAID1, RAID5, or RAID10. For information about V-Series systems and how they support RAID types, see the V-Series Systems Planning Guide.
Choosing the right size and the protection level for a RAID group depends on the kind of data you intend to store on the disks in that RAID group. For more information about RAID groups, see Understanding RAID groups.
A plex is a collection of one or more RAID groups that together provide the storage for one or more WAFL® (Write Anywhere File Layout) file system volumes. Data ONTAP uses plexes as the unit of RAID-level mirroring when the SyncMirror® feature is enabled. All RAID groups in one plex are of the same level, but may have a different number of disks.
An aggregate is a collection of one or two plexes, depending on whether you want to take advantage of RAID-level mirroring. If the aggregate is unmirrored, it contains a single plex. If the SyncMirror feature is licensed and enabled, Data ONTAP adds a second plex to the aggregate, which serves as a RAID-level mirror for the first plex in the aggregate.
When you create an aggregate, Data ONTAP assigns data disks and parity disks to RAID groups, depending on the options you choose, such as the size of the RAID group (based on the number of disks to be assigned to it) or the level of RAID protection.
You use aggregates to manage plexes and RAID groups because these entities only exist as part of an aggregate. You can increase the usable space in an aggregate by adding disks to existing RAID groups or by adding new RAID groups. Once you've added disks to an aggregate, you cannot remove them to reduce storage space without first destroying the aggregate.
If the SyncMirror feature is licensed and enabled, you can convert an unmirrored aggregate to a mirrored aggregate and vice versa without any downtime.
Consists of one plex, automatically named by Data ONTAP as plex0. This is the default configuration. In the following diagram, the unmirrored aggregate, arbitrarily named aggrA by the user, consists of one plex, which is made up of four double-parity RAID groups, automatically named rg0, rg1, rg2, and rg3 by Data ONTAP.
Notice that RAID-DP requires that both a parity disk and a double parity disk be in each RAID group. In addition to the disks that have been assigned to RAID groups, there are sixteen hot spare disks in one pool of disks waiting to be assigned.
Consists of two plexes, which provide an even higher level of data redundancy via RAID-level mirroring. For an aggregate to be enabled for mirroring, the storage system's disk configuration must support RAID-level mirroring, and the storage system must have the necessary licenses installed and enabled, as follows.
When you enable SyncMirror, Data ONTAP divides all the hot spare disks into two disk pools to ensure a single failure does not affect disks in both pools. This allows the creation of mirrored aggregates. Mirrored aggregates have two plexes. Data ONTAP uses disks from one pool to create the first plex, always named plex0, and another pool to create a second plex, typically named plex1. This provides fault isolation of plexes. A failure that affects one plex will not affect the other plex.
The plexes are physically separated (each plex has its own RAID groups and its own disk pool), and the plexes are updated simultaneously during normal operation. This provides added protection against data loss if there is a double-disk failure or a loss of disk connectivity, because the unaffected plex continues to serve data while you fix the cause of the failure. Once the plex that had a problem is fixed, you can resynchronize the two plexes and reestablish the mirror relationship.
In the following diagram, SyncMirror is enabled, so plex0 has been copied and automatically named plex1 by Data ONTAP. Notice that plex0 and plex1 contain copies of one or more file systems and that the hot spare disks have been separated into two pools, Pool0 and Pool1.
For more information about aggregates, see Understanding aggregates.
A volume is a logical file system whose structure is made visible to users when you export the volume to a UNIX host through an NFS mount or to a Windows host through a CIFS share.
You assign the following attributes to every volume, whether it is a traditional or a FlexVol volume, except where noted:
Data ONTAP automatically creates and deletes Snapshot copies of data in volumes to support commands related to Snapshot technology.
For more information about the default Snapshot copy schedule, Snapshot copies, plexes, and SyncMirror, see the Data Protection Online Backup and Recovery Guide
With all new storage systems, Data ONTAP is installed at the factory with a root volume already configured. The root volume is named vol0 by default.
The root volume contains the storage system's configuration files, including the /etc/rc file, which includes startup commands and log files. You use the root volume to set up and maintain the configuration files.
Only one root volume is allowed on an storage system. The root volume contains log files, so for traditional volumes, make sure your root volume spans four to six disks to handle the increased traffic.
A volume is the most inclusive of the logical containers. It can store files and directories, qtrees, and LUNs. You can use qtrees to organize files and directories, as well as LUNs. You can use LUNs to serve as virtual disks in SAN environments to store files and directories. For information about qtrees, see How qtrees are used. For information about LUNs, see How LUNs are used.
The following diagram shows how you can use volumes, qtrees, and LUNs to store files and directories.
For more information about volumes, see Chapter 6, Volume Management.
Each volume depends on its containing aggregate for all its physical storage. The way a volume is associated with its containing aggregate depends on whether the volume is a traditional volume or a FlexVol volume.
A traditional volume is contained by a single, dedicated, aggregate. A traditional volume is tightly coupled with its containing aggregate. The only way to increase the size of a traditional volume is to add entire disks to its containing aggregate. It is impossible to decrease the size of a traditional volume.
The smallest possible traditional volume must occupy all of two disks (for RAID4) or three disks (for RAID-DP). Thus, the minimum size of a traditional volume depends on the size and number of disks used to create the traditional volume.
No other volume can use the storage associated with a traditional volume's containing aggregate.
When you create a traditional volume, Data ONTAP creates its underlying containing aggregate based on the parameters you choose with the vol create command or with the FilerView® Volume Wizard. Once created, you can manage the traditional volume's containing aggregate with the aggr command. You can also use FilerView to perform some management tasks.
The aggregate portion of each traditional volume is assigned its own pool of disks that are used to create its RAID groups, which are then organized into one or two plexes. Because traditional volumes are defined by their own set of disks and RAID groups, they exist outside of and independently of any other aggregates that might be defined on the storage system.
The following diagram illustrates how a traditional volume, trad_volA, is tightly coupled to its containing aggregate. When trad_volA was created, its size was determined by the amount of disk space requested, the number of disks and their capacity to be used, or a list of disks to be used.
A FlexVol volume is loosely coupled with its containing aggregate. Because the volume is managed separately from the aggregate, FlexVol volumes give you a lot more options for managing the size of the volume. FlexVol volumes provide the following advantages:
These volumes stripe their data across all the disks and RAID groups in their containing aggregate.
A FlexVol volume can share its containing aggregate with other FlexVol volumes. Thus, a single aggregate is the shared source of all the storage used by the FlexVol volumes it contains.
In the following diagram, aggrB contains four FlexVol volumes of varying sizes. Note that one of the FlexVol volumes is a FlexClone.
You can create traditional volumes and FlexVol volumes on the same storage system, up to the maximum number of volumes allowed. For information about maximum limits, see Maximum numbers of volumes.
A snapshot is a space-efficient, point-in-time image of the data in a volume or an aggregate. Snapshots are used for such purposes as backup and error recovery.
Data ONTAP automatically creates and deletes snapshots of data in volumes to support commands related to Snapshot technology. Data ONTAP also automatically creates snapshots of aggregates to support commands related to the SnapMirror® software, which provides volume-level mirroring. For example, Data ONTAP uses snapshots when data in two plexes of a mirrored aggregate need to be resynchronized.
You can accept the default snapshot schedule, or modify it. You can also create one or more snapshots at any time. For information about aggregate snapshots, see the System Administration Guide. For more information about snapshots, plexes, and SyncMirror, see the Data Protection Online Backup and Recovery Guide.
Copyright © 1994-2005, Network Appliance, Inc. All rights
reserved.
Part No. 210-01085_B0
File: 1intro2.htm
Last updated: 01-Dec-2005 16:51:21