ONTAP Discussions
ONTAP Discussions
8.3 docs on "Physical Storage" only refers to Flash Pools when talking about using Storage Pools on SSDs. Is there some reason the Storage Pool allocations can't/shouldn't be used to create all-SSD aggregates as well? Nos supported? Not BP?
Solved! See The Solution
HI Francis,
Storage pools partition the SSDs differently. In a pure SSD aggregate they are either unpartitioned, or root/data partitioned if the platform was initialized with ADP. Storage Pool's partition them into 4 slices so you can apply cache to more aggregates with fewer SSDs.
https://library.netapp.com/ecmdocs/ECMP1636022/html/GUID-8284CB30-5BC5-4D10-B0AA-AA8F8DAA752E.html
HI Francis,
Storage pools partition the SSDs differently. In a pure SSD aggregate they are either unpartitioned, or root/data partitioned if the platform was initialized with ADP. Storage Pool's partition them into 4 slices so you can apply cache to more aggregates with fewer SSDs.
https://library.netapp.com/ecmdocs/ECMP1636022/html/GUID-8284CB30-5BC5-4D10-B0AA-AA8F8DAA752E.html
Sean,
My core question is, how disk-like are the four SSD partitions that get created when you create Storage Pools? Can I make a pure SSD aggregate out of them, or is their use limited to adding cache space to Flash Pools?
What I wanted to explore was to see if we can apply the parity amortization advantage of Storage Pools beyond Flash Pools.
I think what you're looking for is ADPv2. There each SSD gets 2 data slices and a root slice. The data slices become 2 data aggregates. Each controller gets a data aggregate and half the root slices.
The storage pools are only intended for cache. but now that you've piqued my interest some vsims will be sacrificed exploring that question.
Curiosity satisfied. Once SSDs are placed into a storage pool they are prevented from being used in an aggregate.
Here I've made a storage pool called "cache1", which appears as the container name in the disk list:
crshtst1::storage pool*> disk show Usable Disk Container Container Disk Size Shelf Bay Type Type Name Owner ---------------- ---------- ----- --- ------- ----------- --------- -------- VMw-1.1 7.10GB - 0 VMDISK aggregate aggr0 crshtst1-01 VMw-1.2 7.10GB - 1 VMDISK aggregate aggr0 crshtst1-01 VMw-1.3 7.10GB - 2 VMDISK aggregate aggr0 crshtst1-01 VMw-1.4 7.10GB - 3 VMDISK spare Pool0 crshtst1-01 VMw-1.5 7.10GB - 4 VMDISK spare Pool0 crshtst1-01 VMw-1.6 7.10GB - 5 VMDISK spare Pool0 crshtst1-01 VMw-1.7 7.10GB - 6 VMDISK spare Pool0 crshtst1-01 VMw-1.8 7.10GB - 8 VMDISK spare Pool0 crshtst1-01 VMw-1.9 7.10GB - 9 VMDISK spare Pool0 crshtst1-01 VMw-1.10 7.10GB - 10 VMDISK spare Pool0 crshtst1-01 VMw-1.11 7.10GB - 11 VMDISK spare Pool0 crshtst1-01 VMw-1.12 7.10GB - 12 VMDISK spare Pool0 crshtst1-01 VMw-1.13 7.10GB - 13 VMDISK spare Pool0 crshtst1-01 VMw-1.14 7.10GB - 14 VMDISK spare Pool0 crshtst1-01 VMw-1.15 7.10GB - 15 VMDISK spare Pool0 crshtst1-01 VMw-1.16 3.55GB - 0 SSD shared cache1 crshtst1-01 VMw-1.17 3.55GB - 1 SSD shared cache1 crshtst1-01 VMw-1.18 3.55GB - 2 SSD shared cache1 crshtst1-01 VMw-1.19 3.55GB - 3 SSD shared cache1 crshtst1-01 VMw-1.20 3.55GB - 4 SSD shared cache1 crshtst1-01 VMw-1.21 3.55GB - 5 SSD spare Pool0 crshtst1-01 21 entries were displayed.
Now the usual aggr creation method fails, because I don't have enough spare SSD left:
crshtst1::storage pool*> aggr create -aggregate ssd1 -diskcount 5 -disktype ssD [Job 21] Job is queued: Create ssd1. Error: command failed: [Job 21] Job failed: Failed to create aggregate "ssd1" on "crshtst1-01". Reason: 5 disks needed from one pool, but not enough matching disks are available in either pool.
I can see the partitions though:
crshtst1::storage disk partition*> show Usable Container Container Partition Size Type Name Owner ------------------------- ------- ------------- ----------------- ----------------- VMw-1.16.P1 895.5MB spare Pool0 crshtst1-01 VMw-1.16.P2 895.5MB spare Pool0 crshtst1-01 VMw-1.16.P3 895.5MB spare Pool0 crshtst1-01 VMw-1.16.P4 895.5MB spare Pool0 crshtst1-01 VMw-1.17.P1 895.5MB spare Pool0 crshtst1-01 VMw-1.17.P2 895.5MB spare Pool0 crshtst1-01 VMw-1.17.P3 895.5MB spare Pool0 crshtst1-01 VMw-1.17.P4 895.5MB spare Pool0 crshtst1-01 VMw-1.18.P1 895.5MB spare Pool0 crshtst1-01 VMw-1.18.P2 895.5MB spare Pool0 crshtst1-01 VMw-1.18.P3 895.5MB spare Pool0 crshtst1-01 VMw-1.18.P4 895.5MB spare Pool0 crshtst1-01 VMw-1.19.P1 895.5MB spare Pool0 crshtst1-01 VMw-1.19.P2 895.5MB spare Pool0 crshtst1-01 VMw-1.19.P3 895.5MB spare Pool0 crshtst1-01 VMw-1.19.P4 895.5MB spare Pool0 crshtst1-01 VMw-1.20.P1 895.5MB spare Pool0 crshtst1-01 VMw-1.20.P2 895.5MB spare Pool0 crshtst1-01 VMw-1.20.P3 895.5MB spare Pool0 crshtst1-01 VMw-1.20.P4 895.5MB spare Pool0 crshtst1-01 20 entries were displayed.
But I still can't use them to create an aggregate
crshtst1::storage disk partition*> aggr create -aggregate test1 -partitionlist VMw-1.16.P1,VMw-1.17.P1,VMw-1.18.P1,VMw-1.19.P1,VMw-1.20.P1 [Job 22] Job is queued: Create test1. Error: command failed: [Job 22] Job failed: Failed to create aggregate "test1" on "crshtst1-01". Reason: Disk "VMw-1.16.P1" is part of a storage pool. Please use the 'aggregate add' command with the '-storage-pool' option to add capacity from a storage pool to an aggregate.