AFF

Aggregate expansion

AlexeyF
281 Views

Hi

We have a followig configuration: 

AFF250 with 10 disks of 13,97TB.

ADPv2 is in place.

2 data aggregates of 9 partitions each 6,94 TiB in size.

 

We've just added 8 disks which appear as spares, not partitioned.

Tryig to expand the aggregate via GUI (ONTAP 9.10) I get the following estimation message:

     Adding drives to local tier "aggr02" will increase its capacity by approximately 25 TiB

I wonder how is it going to expand the aggregate?

I would expect 8 partitions of 6,94 TiB added to the existing raid group what would result in 8x6,94TiB

Can I do it manually via command line?

Have not found any documentation other than Add capacity to a local tier (add disks to an aggregate) (netapp.com) but it is not detailed enough.

 

Thanks in advance for your advices.

1 ACCEPTED SOLUTION

SpindleNinja
276 Views

It'll partition the drives and add the partition to the aggr.  the remaining partition should get assigned to the partner node. 

You can do it via CLI, I prefer it actually.   The -simulate true flag is very helpful.  

You'll end up with 17 disks in the raid group and a shared spare. 

View solution in original post

6 REPLIES 6

SpindleNinja
277 Views

It'll partition the drives and add the partition to the aggr.  the remaining partition should get assigned to the partner node. 

You can do it via CLI, I prefer it actually.   The -simulate true flag is very helpful.  

You'll end up with 17 disks in the raid group and a shared spare. 

AlexeyF
266 Views

why then only 25TB instead of  55TB (8x6,94TB)?

SpindleNinja
263 Views

by "13,97TB". I'm assuming you mean 15.3TB NVMe Drives? 

Good question.. checking maths - 

 

10 total disks looks to be - 43.7 TiB useable per data aggr.   RG size is 9x6.9TiB partitions 

18 total disks looks to be - 91.11 TiB useable per data aggr. RG size is 17x6.9TiB partitions 

 

But I would check it via the CLI - could be a GUI bug? 

AlexeyF
249 Views

CLUSTER02::> storage aggregate add-disks -aggregate aggr02 -diskcount 8 -simulate true

Error: command failed: Failed to add disks to aggregate "aggr02". Reason: Not enough spares on node Node2.

 

4 new disks are owned by Node1 and other 4 new disks are owned by Node2

1 spare already partitioned is in 1.0.4

 

The command would work only with max 5 disks:

storage aggregate add-disks -aggregate aggr02 -diskcount 5 -simulate true

Disks would be added to aggregate "aggr02" on node "Node2" in the following manner:

First Plex

RAID Group rg0, 5 disks (block checksum, raid_dp)
Usable Physical
Position Disk Type Size Size
---------- ------------------------- ---------- -------- --------
shared 1.0.4 SSD-NVM 6.94TB 6.94TB
shared 1.0.5 SSD-NVM 6.94TB 6.94TB
shared 1.0.6 SSD-NVM 6.94TB 6.94TB
shared 1.0.7 SSD-NVM 6.94TB 6.94TB
shared 1.0.8 SSD-NVM 6.94TB 6.94TB

Aggregate capacity available for volume use would be increased by 31.23TB.

The following disks would be partitioned: 1.0.5, 1.0.6, 1.0.7, 1.0.8.
Warning: Not enough spares of disk type SSD-NVM on node  Node2.

 

So maybe should first add 4 disks to every aggregate, it will partition them and therefore permit to add another 4 disks ?

andris
150 Views

You shouldn't have to... what is your spare partition/disk situation, now?
You still need to maintain spares after the aggregate expansions.

nodeshell CLI "sysconfig -r" for each node might provide more clarity.

Or open a technical case for formal assistance...

AlexeyF
83 Views

So, adding 4 disks first (those 4 owened by the aggregates node) to one aggregate and the other 4 to another aggregates has partitioned them and I could add another 4 disk partitions to each aggregate (owened by partner node) afterwards.

 

It has added 25TBx2 though instead of 6.94TBx4x2.

Public