Basic questions 2 X FAS2552 on a DS2246 (24 disks) | NetApp | Page 1

chilly knoll Jun 14, 2023, 12:17 PM

#

Hello again folks,

Several questions. I am running OnTap 9.8P4 on both nodes.

I have 24 disks total (1.09TB / 10K SAS) in one shelf DS2246

Correct me if I'm wrong when I elaborate so I can learn properly.
I am looking for the best scenario in my case before ongoing with vservers, volumes etc.
Question #1

On picture 1 we have root aggregate on genesis-01 (10 disks) and genesis-02 (10 disks)

I was wondering can I create a new aggregate on 24 disks (system won't let me) I can only do it on 12 disks. Local Tiers on Ontap GUI creates 2 aggregates of 11 disks for a total of 22 disks. I imagine the 2 others will be spare. But at the moment they shared.

**Question #2 **

Still on picture 1 I was trying a lot to change the 4 disks (1.0.20-1.0.23) from shared to spare but I am having a hard time finding a solution after some time searching and trying.

Could you please help me out ?

Thank you 🙂 By the way cool channel Nick I am currently watching the ADP video.

chilly knoll Jun 14, 2023, 1:03 PM

#

Question #3 for FAS2552 can I install the latest ONTAP versions ? 9.12 (example)

#

vivid dirge Jun 14, 2023, 2:24 PM

#

For question 3, you can check the maximum supported ONTAP version by looking up your FAS model on our Hardware Universe site @ hwu.netapp.com

chilly knoll Jun 14, 2023, 2:52 PM

#

vivid dirge For question 3, you can check the maximum supported ONTAP version by looking up ...

Yep that did the trick. 9.8 P19 ! Thank you Ross 🙂

mortal holly Jun 14, 2023, 5:11 PM

#

for Question 1, yes, but you have to manually reassign the disks to a single node to create one big aggregate

#

for question 2, the disks will stay shared because they are partitioned. the partitions are spare. try set advanced; storage disk partition show

chilly knoll Jun 14, 2023, 5:40 PM

#

mortal holly for Question 1, yes, but you have to manually reassign the disks to a single nod...

Not to sure how to pull this off really. Tell me with your experience. Which scenario is the best. 24 disks aggr or 2 aggr of 12 disks.

chilly knoll Jun 14, 2023, 5:41 PM

#

mortal holly for question 2, the disks will stay shared because they are partitioned. the par...

I don't have partition after " storage disk "

mortal holly Jun 14, 2023, 8:02 PM

#

chilly knoll I don't have partition after " storage disk "

try set advanced , press enter, and then try storage disk <tab>

red fox Jun 14, 2023, 8:55 PM

#

try: storage disk create-partition

mortal holly Jun 14, 2023, 9:10 PM

#

red fox try: storage disk create-partition

but he already has the partitions 😉
he needs to reassign them. reassignment of the data partition works with storage disk removeowner -disk ... -data true, followed by storage disk assign -disk ... -data true -owner <node>. After all partitions are reassigned to the same owner, you should be able to create an aggregate using all disks (23 out of the 24, because you still need a spare... make sure to set max-raid-size to at least 23 or you'll end up with a lot less free space than what you expected)

chilly knoll Jun 15, 2023, 10:12 AM

#

@mortal holly By doing will I still have HA if one node dies ?

mortal holly Jun 15, 2023, 1:48 PM

#

chilly knoll <@189749672921792512> By doing will I still have HA if one node dies ?

of course

chilly knoll Jun 15, 2023, 2:16 PM

#

I'll give it a try. Thank you sir.

chilly knoll Jun 15, 2023, 4:09 PM

#

Here's an update.

#

If I try Y after : storage disk removeowner -disk 1.0.0 -data true - apparently the disk is auto-assign so won't affect my changes. Perhaps I'm mistaken

red fox Jun 15, 2023, 5:11 PM

#

There is an easier way to do this

#

You have 24x drives and each node has 12x data partitions each

#

So, on one node (pick your favorite one) create an aggregate using all 12x of the data partitions on that node.

#

Then, relocate the aggregate to the partner using: storage aggregate relocation start -aggregate [aggr-name] -node [node name owning the 12x partition aggregate] -destination [node name to own the aggregate]

#

this will, in turn, automatically assign all of those partitions to the new, destination node

#

from there, add 11x more partitions into RG0, so that you retain 1x spare for the node holding the aggregate

mortal holly Jun 15, 2023, 5:24 PM

#

that's actually a pretty smart solution 🙂

#

honestly, the disk assignment (especially in MetroCluster IP) is what makes me want to hurt myself (or the storage system 😉 ) every time I have to do it ...

chilly knoll Jun 15, 2023, 6:46 PM

#

Yes Scott I agree it's a good solution I'll try that and will let you know. Thanks.

chilly knoll Jun 15, 2023, 7:29 PM

#

It worked. Now trying to find the commands for the other part

mortal holly Jun 15, 2023, 8:50 PM

#

first you need to delete one of the aggregates again. then set maxraidsize on the other one to 23 (if you didn't do that already) and add all but one disks to it

chilly knoll Jun 15, 2023, 11:58 PM

#

Awesome it worked. Now then the spare disk is it ok if genesis-02 owns it or that disk needs to be without owner ?

red fox Jun 16, 2023, 8:47 PM

#

You should reassign it over to node 1. Since it's a spare, it should be easy to do that.

#

or, you can relocate the aggregate over to node 2 for an even simpler course of action

#

either way, the result is the same

chilly knoll Jul 4, 2023, 11:37 AM

#

Just letting you know it worked. Thanks Scott !

#

Case closed thank you !

#Basic questions 2 X FAS2552 on a DS2246 (24 disks)