Creating aggregates

About creating aggregates

When a single, unmirrored aggregate is first created, all the disks in the single plex must come from the same disk pool.

How Data ONTAP enforces checksum type rules

As mentioned in Chapter 3, Data ONTAP uses the disk's checksum type for RAID parity checksums. You must be aware of a disk's checksum type because Data ONTAP enforces the following rules when creating aggregates or adding disks to existing aggregates (these rules also apply to creating traditional volumes or adding disks to them):

An aggregate can have only one checksum type, and it applies to the entire aggregate.
When you create an aggregate:
- Data ONTAP determines the checksum type of the aggregate, based on the type of disks available.
- If enough block checksum disks (BCDs) are available, the aggregate uses BCDs.
- Otherwise, the aggregate uses zoned checksum disks (ZCDs).
- To use BCDs when you create a new aggregate, you must have at least the same number of block checksum spare disks available that you specify in the aggr create command.
When you add disks to an existing aggregate:
- You can add a BCD to either a block checksum aggregate or a zoned checksum aggregate.
- You cannot add a ZCD to a block checksum aggregate.

If you have a system with both BCDs and ZCDs, Data ONTAP attempts to use the BCDs first. For example, if you issue the command to create an aggregate, Data ONTAP checks to see whether there are enough BCDs available.

If there are enough BCDs, Data ONTAP creates a block checksum aggregate.
If there are not enough BCDs, and there are no ZCDs available, the command to create an aggregate fails.
If there are not enough BCDs, and there are ZCDs available, Data ONTAP checks to see whether there are enough of them to create the aggregate.
- If there are not enough ZCDs, Data ONTAP checks to see whether there are enough mixed disks to create the aggregate.
- If there are enough mixed disks, Data ONTAP mixes block and zoned checksum disks to create a zoned checksum aggregate.
- If there are not enough mixed disks, the command to create an aggregate fails.

Once an aggregate is created on a storage system, you cannot change the format of a disk. However, on NetApp V-Series systems, you can convert a disk from one checksum type to the other with the disk assign -c block | zoned command. For more information, see the V-Series Software Setup, Installation, and Management Guide.

Data ONTAP automatically creates Snapshot copies of aggregates to support commands related to the SnapMirror software, which provides volume-level mirroring. For example, Data ONTAP uses Snapshot copies when data in two plexes of a mirrored aggregate need to be resynchronized.

You can accept or modify the default Snapshot copy schedule. You can also create one or more Snapshot copies at any time. For information about aggregate Snapshot copies, see the System Administration Guide. For information about Snapshot copies, plexes, and SyncMirror, see the Data Protection Online Backup and Recovery Guide.

Creating an aggregate

When you create an aggregate, you must provide the following information:

A name for the aggregate

The names must follow these naming conventions:

Begin with either a letter or an underscore (_)
Contain only letters, digits, and underscores
Contain no more than 255 characters

Disks to include in the aggregate

You specify disks by using the -d option and their IDs or by the number of disks of a specified size.

All of the disks in an aggregate must follow these rules:

Disks must be of the same type (FC-AL, ATA, or SCSI).
Disks must have the same RPM.

If disks with different speeds are present on a storage system (for example, both 10,000 RPM and 15,000 RPM disks), Data ONTAP avoids mixing them within one aggregate. By default, Data ONTAP selects disks

With the same speed when creating an aggregate in response to the following commands:
- aggr create
- vol create
That match the speed of existing disks in the aggregate that requires expansion or mirroring in response to the following commands:
- aggr add
- aggr mirror
- vol add
- vol mirror

If you use the -d option to specify a list of disks for commands that add disks, the operation will fail if the speeds of the disks differ from each other or differ from the speed of disks already included in the aggregate. The commands for which the -d option will fail in this case are aggr create, aggr add, aggr mirror, vol create, vol add, and vol mirror. For example, if you enter aggr create vol4 -d 9b.25 9b.26 9b.27 and two of the disks are of different speeds, the operation fails.

When using the aggr create or vol create commands, you can use the -R rpm option to specify the type of disk to used based on speed. You only need to use this option on storage systems that have different disks with different speeds. Typical values for rpm are: 5400, 7200, 10000, and 15000. The -R option cannot be used with the -d option.

If you have any question concerning the speed of a disk that you are planning to specify, use the sysconfig -r command to ascertain the speed of the disks that you want to specify.

Attention
It is possible to override the RPM check with option -f, but doing this might have a negative impact on the performance of the resulting aggregate.

Data ONTAP periodically checks if adequate spares are available for the storage system. In those checks, only disks with matching or higher speeds are considered as adequate spares. However, if a disk fails and a spare with matching speed is not available, Data ONTAP may use a spare with a different (higher or lower) speed for RAID reconstruction.

Note
If an aggregate happens to include disks with different speeds and adequate spares are present, you can use the disk replace command to replace mismatched disks. Data ONTAP will use Rapid RAID Recovery to copy such disks to more appropriate replacements.

Note
If you are setting up aggregates on an FAS270c storage system with two internal system heads or a system licensed for SnapMover, you might have to assign the disks to one of the storage systems before creating aggregates on those systems. For more information, see Software-based disk ownership.

For information about creating aggregates, see the na_aggr man page.

To create an aggregate, complete the following steps.

Step	Action
1	View a list of the spare disks on your storage system. These disks are available for you to assign to the aggregate that you want to create. Enter the following command: aggr status -s Result The output of `aggr status -s` lists all the spare disks that you can select for inclusion in the aggregate and their capacities.
2	Enter the following command: aggr create aggr_name [-f] [-m] [-n] [-t { raid4 \| raid_dp} ] [-r raidsize] [-T disk-type][-R rpm] disk- list aggr_name is the name for the new aggregate. `-f` overrides the default behavior that does not permit disks in a plex to span disk pools. This option also allows you to mix disks with different RPM speeds. `-m` specifies the optional creation of a SyncMirror-replicated volume if you want to supplement RAID protection with SyncMirror protection. A SyncMirror license is required for this feature. `-t {raid4 \| raid_dp}` specifies the level of RAID protection you want to provide for this aggregate. If no RAID level is specified, the default value (`raid_dp)` is applied. `-r` raidsize is the maximum number of disks that you want RAID groups created in this aggregate to consist of. If the last RAID group created contains fewer disks than the value specified, any new disks that are added to this aggregate are added to this RAID group until that RAID group reaches the number of disks specified. When that point is reached, a new RAID group will be created for any additional disks added to the aggregate.
	`-T` disk-type specifies one of the following types of disk to be used: ATA, EATA, FCAL, LUN, and SCSI. This option is only needed when creating aggregates on systems that have mixed disks. Mixing disks of different types in one aggregate is not allowed. You cannot use the `-T` option in combination with the `-d` option. `-R` rpm specifies the type of disk to used based on its speed. Use this option only on storage systems having different disks with different speeds. Typical values for rpm are: 5400, 7200, 10000, and 15000. The `-R` option cannot be used with the `-d` option.
	disk-list is one of the following: ndisks[@disk-size] ndisks is the number of disks to use. It must be at least 2 (3 if RAID-DP is configured). disk-size is the disk size to use, in gigabytes. You must have at least ndisks available disks of the size you specify. `-d` disk_name1 disk_name2... disk_nameN disk_name1, disk_name2, and disk_nameN are disk IDs of one or more available disks; use a space to separate multiple disks.
3	Enter the following command to verify that the aggregate exists as you specified: *aggr status aggr_name* -r** aggr_name is the name of the aggregate whose existence you want to verify. Result: The system displays the RAID groups and disks of the specified aggregate on your storage system.

Aggregate creation example

The following command creates an aggregate called newaggr, with no more than eight disks in a RAID group consisting of the disks with disk IDs 8.1, 8.2, 8.3, and 8.4.

aggr create newaggr -r 8 -d 8.1 8.2 8.3 8.4.