Table of ContentsView in Frames

Creating an aggregate

You create an aggregate to provide storage to one or more FlexVol volumes.

Before you begin

About this task

Aggregate names must conform to the following requirements:

Steps

  1. Display a list of available spares: aggr status -s
  2. Create the aggregate: aggr create aggr_name [-f] [-m] [-n] [-t {raid0 | raid4 | raid_dp}] [-r raidsize] [-T disk-type] -R rpm] [-L] [-p] disk-list
    aggr_name is the name for the new aggregate.

    -f overrides the default behavior that does not permit drives in a plex to belong to different pools. This option also enables you to mix drives with different RPM speeds even if the appropriate raid.rpm option is not off.

    -m specifies the optional creation of a SyncMirror-replicated volume if you want to supplement RAID protection with SyncMirror protection.

    -n displays the results of the command but does not execute it. This is useful for displaying the drives that would be automatically selected prior to executing the command.

    -t {raid0 | raid4 | raid_dp} specifies the level of RAID protection you want to provide for this aggregate. If no RAID level is specified for an aggregate composed of drives, the default value (raid_dp) is applied. raid0 is used only for array LUNs.

    -r raidsize is the maximum size of the RAID groups for this aggregate. If no size is specified, the default is used.

    -T disk-type specifies the Data ONTAP drive type. This option is needed when creating aggregates on systems that have mixed drive types or both drives and array LUNs.

    -R rpm specifies the type of drive to use based on its speed. Valid values for rpm include 5400, 7200, 10000, and 15000.

    -L creates a SnapLock aggregate. For more information about the SnapLock feature, see the na_aggr(1) man page or the Data ONTAP Archive and Compliance Management Guide for 7-Mode.

    -p specifies the pool from which the drives are selected.

    disk-list is one of the following values:

    • ndisks[@disk-size]

      ndisks is the number of drives to use.

      disk-size is the drive size to use, in gigabytes.

    • -d disk_name1 disk_name2... disk_nameN

      disk_name1, disk_name2, and disk_nameN are drive IDs of available drives; use a space to separate drive IDs.

  3. Verify the RAID groups and drives of your new aggregate: aggr status -r aggr_name

Example

You want to create a storage configuration for a single node cluster and have 4 shelves of SAS disks totaling 96 drives, with 3 drives used for the root aggregate. You decide to allocate the remaining drives in the following manner:
  • 2 spares
  • 91 drives used for data in one aggregate:
    • 4 RAID groups of 18 drives each
    • 1 RAID group of 19 drives

To accomplish this goal, you use the following steps:

  1. Create the first RAID group of 19 disks: aggr create n01sas01 -r 19 -T SAS 19
  2. Change the aggregate's maximum RAID group size to 18: aggr options n01sas01 raidsize 18
  3. Add the remaining 54 disks as 4 RAID groups of 18 disks each: aggr add n01sas01 -T SAS 54

Later, if you added a 24-drive shelf to the node and wanted to add the storage to this aggregate, you could again modify the RAID group size to 24 so that the entire shelf could be accommodated in a single RAID group, provided that a RAID group size of 24 is acceptable for your installation.