Replace manually a broken disk with a spare one

UnitIT — Fri, 30 Aug 2024 13:37:30 GMT

Hello,

Since this Monday, I've had two discs fall into failed state. One was automatically replaced by one of the spare disks, but the other was not. For information, my aggregate is made up of 7200RPM disks.

I actually had 3 spare disks:
- 1 x 7200RPM (the one that replaced the failed disk)
- 2 x 15000RPM (which don't automatically launch the replacement)

I guess the replacement doesn't start automatically because the disks have different specifications (I understand that and I'm quite okay with it).

However, I'd like to force the disk replacement even if the spare disk is faster.

Most of the disks have been running for as long as the disks that failed, and I'd like to minimize the chances of ending up with 3 failed disks (and then the drama...).

My failed aggregate:

Position Disk Pool Type RPM Size Size Status
-------- --------------------------- ---- ----- ------ -------- -------- ----------
shared 2.0.8 0 FSAS 7200 3.57TB 3.64TB (normal)
shared 2.0.10 0 FSAS 7200 3.57TB 3.64TB (normal)
shared 2.0.12 0 FSAS 7200 3.57TB 3.64TB (normal)
shared 2.0.14 0 FSAS 7200 3.57TB 3.64TB (normal)
shared 2.0.16 0 FSAS 7200 3.57TB 3.64TB (normal)
shared 2.1.23 0 FSAS 7200 3.57TB 3.64TB (normal)
shared 2.0.4 0 FSAS 7200 3.57TB 3.64TB (normal)
shared FAILED - - - 3.57TB - (failed)
shared 2.0.22 0 FSAS 7200 3.57TB 3.64TB (normal)

My spare disks:

Disk HA Shelf Bay Chan Pool Type Class RPM Size Size Owner
--------------- ------------ ---- ------ ------ ----------- ------ -------- -------- --------
2.0.6 0a 0 6 B Pool0 SAS performance 15000 3.63TB 3.64TB SAME_OWNER_AS_THE_AGGREGATE
2.0.17 0a 0 17 B Pool0 SAS performance 15000 3.63TB 3.64TB SAME_OWNER_AS_THE_AGGREGATE

Thank you in advance !

Re: Replace manually a broken disk with a spare one

cedric_renauld — Sun, 01 Sep 2024 16:11:47 GMT

Hello,

It's normal way, different type of disk could not be in the same AGGR, for you : SAS and FSAS (Sata)

But it's a best practice to avoid mixed disk type and avoid performance issue, if your are aware abot this issues you can try this command to authorize the mix :

7-Mode
>raid.disktype.enable

Cluster ONTAP
>storage raid-options modify - raid.disktype.enable

The source is this article :

https://community.netapp.com/t5/Network-and-Storage-Protocols/mixed-disk-types-in-one-aggregate/td-p/9777

topic Replace manually a broken disk with a spare one in ONTAP Discussions

Replace manually a broken disk with a spare one

Re: Replace manually a broken disk with a spare one