What is an aggregate?
Applies to
- Clustered Data ONTAP
- Data ONTAP 7-Mode
Answer
What is an aggregate?
An aggregate is a collection of disks (or partitions) arranged into one or more RAID groups. It is the most basic storage object within ONTAP and is required to allow for the provisioning of space for connected hosts.
How is an aggregate created?
- In 7-Mode, use the aggr create command with the desired parameters.
- In Clustered Data ONTAP, use the storage aggregate create command from the cluster shell with the desired parameters
Note that in order to create an aggregate on a given node, all of the required disks must be owned by the respective node.
Can an aggregate contain disks from different shelves?
Yes, an aggregate can be created using disks across different shelves.
How do you view aggregate details?
There are several commands available to to see aggregate details. The specific commands are determined by the type of information required. The most common commands are as shown.
- To determine the physical size of an aggregate, use df -A <aggregate name> to see the space usage for the aggregate in question.
- To see the physical disks that are in a given aggregate, the aggr status -r <aggregate name> command can be run with an output similar to :
7-Mode:
Example:
controller> aggr status -r aggr0
Aggregate aggr0 (online, raid_dp) (block checksums)
Plex /aggr0/plex0 (online, normal, active)
RAID group /aggr0/plex0/rg0 (normal)
RAID Disk Device HA SHELF BAY CHAN Pool Type RPM Used (MB/blks) Phys (MB/blks)
--------- ------ ------------- ---- ---- ---- ----- -------------- --------------
dparity 0a.16 0a 1 0 FC:A - FCAL 15000 136000/278528000 137104/280790184
parity 0a.17 0a 1 1 FC:A - FCAL 15000 136000/278528000 137104/280790184
data 0c.00.1 0c 0 1 SA:B - SAS 15000 136000/278528000 280104/573653840
Clustered Data ONTAP:
Example:
cluster1::> aggr status -r aggr1_node1
Owner_Node: cluster1-01
Aggregate: aggr1_node1 (online, raid_dp) (block checksums)
Plex: /aggr1_node1/plex0 (online, normal, active, pool0)
RAID Group /aggr1_node1/plex0/rg0 (normal, block checksums)
Usable Physical
Position Disk Pool Type RPM Size Size Status
-------- --------------------------- ---- ----- ------ -------- -------- ----------
dparity 1.1.0 0 BSAS 7200 2.42TB 2.43TB (normal)
parity 1.1.1 0 BSAS 7200 2.42TB 2.43TB (normal)
data 1.1.2 0 BSAS 7200 2.42TB 2.43TB (normal)
data 1.1.3 0 BSAS 7200 2.42TB 2.43TB (normal)
data 1.1.5 0 BSAS 7200 2.42TB 2.43TB (normal)
data 1.1.6 0 BSAS 7200 2.42TB 2.43TB (normal)
6 entries were displayed.
To see specific aggregate details, use the following commands:
7-Mode:
controller> aggr status -v <aggr name>
Clustered Data ONTAP:
cluster1::> node run -node <node name> -command "aggr status -v <aggr name>"
or
cluster1::> aggr show <aggr name> -instance
What to do with a failed aggregate?
An aggregate may have failed when one or more RAID groups have gone offline. Before the aggregate can be brought back online, a root cause will need to be determined and fixed.
Please call NetApp Technical Support for assistance.
What to do with a foreign aggregate?
In clustered Data ONTAP, while adding a shelf from a different cluster (or a 7-Mode system) to a node, any existing aggregates, or parts of aggregates; on that shelf will be flagged as foreign and the drives belonging to that aggregate as orphaned disks. This issue can also be encountered after replacing a failed disk.
For example, an aggregate can show up as follows under the aggr status command:
aggr0(1) failed
This is caused due to the disks not being zeroed (and containing a RAID label) and therefore retaining the aggregate information from an earlier/different system. These foreign aggregates are visible only when using the node shell.
I created an aggregate but it does not show up in the list for creating volumes in OnCommand System Manager.
Check to see if the new aggregate is present in the list of aggregates available for this SVM:
cluster1::> vserver show -vserver <name> -fields aggr-list
What causes a degraded aggregate?
When a disk fails, Data ONTAP can continue to serve data, but it must reconstruct the data from the failed disk using RAID parity. If there are no available spare disks to start the reconstruction the aggregate will be degraded.
A RAID group goes into degraded mode in the following scenarios:
- A single disk fails in a RAID4 group.
- One or two disks fail in a RAID-DP RAID group.
Note: If an aggregate contains at least one RAID group with double disk failure that is not being reconstructed (RAID-DP aggregate only) it is double degraded.
- A disk is taken offline by Data ONTAP (i.e. Disk Firmware Updates).
After the offline disk is brought back online, the RAID group returns to normal mode.
How do I identify a degraded aggregate?
The following examples show a raid group with 1 failed disk and no available spares for reconstruction.
Clustered Data ONTAP:
Cluster::*> aggr show aggr1 -fields state,raidstatus
aggregate raidstatus state
--------- ------------------------------ ------
aggr1 raid_dp, reconstruct, degraded online
7-Mode:
Node> aggr status aggr1
Aggr State Status Options
aggr1 online raid_dp, aggr nosnap=on
reconstruct
degraded
64-bit
Plex /aggr1/plex0: online, normal, active
RAID group /aggr1/plex0/rg0: reconstruction 5% completed, block checksums
Node> sysconfig -r (Note: aggr status -r provides the same output)
Aggregate aggr1 (online, raid_dp, reconstruct, degraded) (block checksums)
Plex /aggr1/plex0 (online, normal, active, pool0)
RAID group /aggr1/plex0/rg0 (reconstruction 17% completed, block checksums)
The following example shows a degraded aggregate with a degraded and double degraded raid group.
(Note: to view the double degraded raid group you must run the commands from the nodeshell. These are the same commands used in 7-mode)
Cluster::*> aggr show aggr2 -fields state,raidstatus
aggregate raidstatus state
--------- ----------------- ------
aggr2 raid_dp, degraded online
Cluster::*> node run -node node_name aggr status -r aggr2
Aggregate aggr2 (online, raid_dp, degraded) (block checksums)
Plex /aggr2/plex0 (online, normal, active, pool0)
RAID group /aggr2/plex0/rg0 (degraded, block checksums)
RAID Disk Device HA SHELF BAY CHAN Pool Type RPM Used (MB/blks) Phys (MB/blks)
--------- ------ ------------- ---- ---- ---- ----- -------------- --------------
dparity FAILED N/A 2538546/ -
parity 3c.10.11 3c 10 11 SA:B 0 BSAS 7200 2538546/5198943744 2543634/5209362816
data 3c.10.12 3c 10 12 SA:B 0 BSAS 7200 2538546/5198943744 2543634/5209362816
data 3c.10.13 3c 10 13 SA:B 0 BSAS 7200 2538546/5198943744 2543634/5209362816
data 3c.10.14 3c 10 14 SA:B 0 BSAS 7200 2538546/5198943744 2543634/5209362816
RAID group /aggr2/plex0/rg1 (double degraded, block checksums)
RAID Disk Device HA SHELF BAY CHAN Pool Type RPM Used (MB/blks) Phys (MB/blks)
--------- ------ ------------- ---- ---- ---- ----- -------------- --------------
dparity FAILED N/A 2538546/ -
parity 3c.10.16 3c 10 16 SA:B 0 BSAS 7200 2538546/5198943744 2543634/5209362816
data FAILED N/A 2538546/ -
data 3c.10.18 3c 10 18 SA:B 0 BSAS 7200 2538546/5198943744 2543634/5209362816
data 3c.10.19 3c 10 19 SA:B 0 BSAS 7200 2538546/5198943744 2543634/5209362816
What to do if have degraded aggr?
Replace the failed disks. After the disks are reconstructed to a spare, the RAID group returns to normal mode.