Add-disk simulate failed issue

[AdminNetapp091320]

HI Experts, does anyone know how to solve this?

Cluster1::> storage aggregate add-disks -aggregate mcluster_aggr_siteA -raidgroup rg0 -diskcount 1 -simulate true

Error: command failed: Addition of disks would fail for aggregate "metrocluster_aggr_siteA_1" on node "Cluster1". Reason: The number of disks specified must be a multiple of the number
of plexes in the target aggregate 'mcluster_aggr_siteA '.

I have 10 unassigned disk and 2 spare disk

[pedro_rocha]

Hello,

Have you tried disk count '2'?

Can you post the disk layout for the aggr?

Regards,

Pedro

[TMACMD]

Everything is usually mirrored in metro cluster

 therefore if I add a disk, I need to add 2, one for the local plex and one for the remote plex

 just like the last post said. 

if you want to add 4 disks you would use 8

[AdminNetapp091320]

Thanks, should I need to move those unassigned disks as spares before re initiating the simulate?

[Fabian1993]

If you have unassigned Disks yes, ensure that you assign same Number of Disk to plex0 (Local) and plex1 (Remote) to expand ths existing Aggregate...

[AdminNetapp091320]

Hi Fabian,

How do we know if  a certain disk is for local or should be for remote plex?  Or regardless, i just need  to put the same number of Disk to Local and Remote. right?

[Fabian1993]

Hi @AdminNetapp091320,

you have to assign the Disks, this is based on the owner and in a MetroCluster you have to set the right plex.

It means, if you want to create on your Site A Cluster a Aggr with 12 Disks you need 12 Disks local in plex0 and 12 Disks on Site B, which are assigned to the owner on SITE A in plex1..

Could you please tell me your MetroCluster Enviroment (2,4-Node Cluster)?

[AdminNetapp091320]

Hi Fabian,

Thanks! I was able to run the simulate successfully using diskcount 2. I'm thinking to run storage aggregate add-disks with diskcount 4 initially (ones another simulation completed). Bit confuse about those 12 spares on how the proper way to point them to a correct group or configuration. How do I know what unassigned disks are for site 1 and what are for site 2 since those unassigned disks are both visible on both sites.

[Fabian1993]

Hi @AdminNetapp091320,

you must check the Ownership, use the command below. Each Disk have a owner. Based on the Ownership, you have Disk available for Site A or for Site B in your Enviroment, also you need the plex0&1 for the Disk.

12 Spare Disk Phy. located on Site A and 12 Spare Disk Phy. located on Site B and if you want to expand the Aggregate on Site A. You have to assign the 12 Spare Disks which are Phy. located on Site A to plex0 - Node1 on Site A, the 12 Disk which are located phy. on Site B, you have to assign to SiteA Node1 to plex1.

storage disk show -ownership

[pedro_rocha]

Hello,

You need to account for disk localization in the disk shelves. You probably have 2 sites or 2 separates racks (A and B for exemple).

Every aggregate has 2 plexes, one local (pool 0) and one remote (pool 1).

Site A

Controller A1/A2 (or just A1)

Shelves with disks for:

  - disks for the A1/A2 local plexes (A pool 0);

  - disks for the B1/B2 remote plexes (B pool 1);

Site B

Controller B1/B2 (or just B1)

Shelves with disks for:

  - disks for the B1/B2 local plexes (B pool 0);

  - disks for the A1/A2 remote plexes (A pool 1);

Every deployment is unique and the distribution of disks in terms of which pool/plex has to be designed according to platform/disk shelves available  at each site.

Give us more information on you equipment.

Regards,

Pedro

[AdminNetapp091320]

Hi Pedro,

Have a look at these. I have a single cluster per site with 12 unassigned disk and would like to move these to my existing single aggregate per site.

cluster1::> storage disk show -container-type unassigned -fields type,physical-size,host-adapter,shelf
disk	host-adapter	physical-size	shelf	type
1.0.1	5d	3.94TB	0	SSD
1.0.2	5d	3.94TB	0	SSD
1.0.3	5d	3.94TB	0	SSD
1.0.4	5d	3.94TB	0	SSD
1.0.5	5a	3.94TB	0	SSD
1.0.6	5d	3.94TB	0	SSD
2.0.15	5c	3.94TB	10	SSD
2.0.16	5c	3.94TB	10	SSD
2.0.17	5c	3.94TB	10	SSD
2.0.18	5c	3.94TB	10	SSD
2.0.19	5c	3.94TB	10	SSD
2.0.20	5c	3.94TB	10	SSD

12 entries were displayed.

cluster1::>storage aggregate show-spare-disks -original-owner cluster1-n1
Original Owner: cluster1-n1
Pool0
Spare Pool
Usable Physical
Disk             Type   Class          RPM Checksum           Size     Size Status
---------------- ------ ----------- ------ -------------- -------- -------- --------
1.0.20            SSD    solid-state      - block           745.0GB  745.2GB zeroed
1.1.11           SSD    solid-state      - block            3.49TB   3.49TB zeroed
Original Owner: cluster1-n1
Pool1
Spare Pool
Usable Physical
Disk             Type   Class          RPM Checksum           Size     Size Status
---------------- ------ ----------- ------ -------------- -------- -------- --------
2.10.5           SSD    solid-state      - block           745.0GB  745.2GB zeroed
2.11.23          SSD    solid-state      - block            3.49TB   3.49TB zeroed
4 entries were displayed.

[pedro_rocha]

How many disk shelves do you have on each site?

Only one data aggr per site? What is the RG size on each aggr?

[AdminNetapp091320]

Hi Pedro,

I have 4 dshelf per site and approximately 30+TB for capacity

aggregate                 node         availsize percent-used size    state  usedsize volcount physical-used physical-used-percent
------------------------- ------------ --------- ------------ ------- ------ -------- -------- ------------- ---------------------
mcluster_aggr_siteA_1 cluster1-n1 11.42TB   73%          41.81TB online 30.39TB  23       30.23TB       69%
node         shelf shelf-id disk-count module-fw-revision psu-fw-version
------------ ----- -------- ---------- ------------------ --------------
cluster1-n1 1.0   0        22         0191,0191          020F,020F
cluster1-n1 1.1   1        24         0191,0191          020F,020F
cluster1-n1 2.10  10       22         0191,0191          020F,020F
cluster1-n1 2.11  11       24         0191,0191          020F,020F
4 entries were displayed.
aggregate                 node         availsize percent-used size    state  usedsize volcount physical-used physical-used-percent
------------------------- ------------ --------- ------------ ------- ------ -------- -------- ------------- ---------------------
mcluster_aggr_siteB_1 cluster2-n1 10.51TB   75%          41.81TB online 31.30TB  23       31.11TB       71%
node         shelf shelf-id disk-count module-fw-revision psu-fw-version
------------ ----- -------- ---------- ------------------ --------------
cluster2-n1 1.0   0        22         0191,0191          020F,020F
cluster2-n1 1.1   1        24         0191,0191          020F,020F
cluster2-n1 2.10  10       22         0191,0191          020F,020F
cluster2-n1 2.11  11       24         0191,0191          020F,020F
4 entries were displayed.

[Fabian1993]

Do you want to expand the Data Aggregate on Cluster1? Than you have to check if they are 12 Disk on Pool 0 and on SITE 2 you need 12 Disk that are assigned at Pool 1 for Cluster1.

If you want to expand a existing Aggregate, check the current Raidgroupsize..

[AdminNetapp091320]

Hi Fabian,

Yes, I wanted to add few more disk to my existing metrocluster aggregates. And if run storage disk show -container-type unassigned command. I can only see total of 12 disk available . Are these still sufficient ?

Also, There were 32 disks assigned in site 1 aggregate mcluster_aggr_siteA_1
16 are under plex0 | pool0
16 are under plex7 | pool1
Maximum Raid size is 23
RaidGroups:
mcluster_aggr_siteA_1/plex0/rg0 (block)
mcluster_aggr_siteA_1/plex7/rg0 (block)
# of spare disk: 2

There were also 32 disk assigned in site 2 aggregate mcluster_aggr_siteB_1
16 are under plex1 | pool1
16 are under plex0 | pool0
Maximum Raid size is 23
RaidGroups:
mcluster_aggr_siteB_1/plex0/rg0 (block)
mcluster_aggr_siteB_1/plex1/rg0 (block)
# of spare disk: 2

Total # of un assigned disk: 12 ( visible on both site)

Given these 12 unassigned disk, can i use the command "

storage disk assign

 storage disk assign -disk <diskname> -owner <owner name>

 to manually assigned some of these unassigned disk to become spare drives? Please advise

[Fabian1993 · Accepted solution]

system node run -node * -command disk show -n

First you have to find out where the 12 Disk are phy. located (Which Shelf or Shelfs).

You musst ensure that you have the same Number of Disk in Pool0 and Pool1..

And than you could assign the 6 Disk to pool0 and 6 Disk to Pool1 (For only Site A) or for both Sides A (3 Disk pool0 / 3 Disk pool1) and the same for Site B...