Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page
ADP disk layout Active-Active HA Pair - Node Root aggregate
2018-02-07
05:11 AM
11,051 Views
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi everyone,
I've got a question regarding a brand new implementation and the way ADP has partitionned my disks. The system is a FAS2650 with 12*960GB SSD, it came with ONTAP 9.1P7 and has been upgraded to 9.2P2.
As described below (coming from the attached document, Page 43) I exepected the following layout:
But instead of that, I've noticed into the configAdvisor report that my partitions are configured like this:
Node A, disks: 0-2-4-6-8-10
Node B, disks: 1-3-5-7-9-11
(I'm only talking about root paritions, no data aggregate has been configured yet)
My question is, should I keep this configuration? I mean, in the attached document this configuration is valid for a AFF A200 system but not for FAS... I don't really know what is the impact if there's one...
Many thanks beforehand for you help,
Miguel
Solved! See The Solution
1 ACCEPTED SOLUTION
brasseurm has accepted the solution
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi there,
I don't believe the root data partition location is deterministic - ie, depending when the drives come back after wipeinfo, they get picked up differently.
There's no issue with where the drives are, but if you wish to shut the system down and do a "disk tango" and just physically move them, it will come back up fine.
8 REPLIES 8
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
My gut feeling is that integrated systems (which AFF A200 is) default to even/odd bay . But yes, I'd love to see defaults explained somewhere, including description of what "bay" disk auto assignment policy actually does.
brasseurm has accepted the solution
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi there,
I don't believe the root data partition location is deterministic - ie, depending when the drives come back after wipeinfo, they get picked up differently.
There's no issue with where the drives are, but if you wish to shut the system down and do a "disk tango" and just physically move them, it will come back up fine.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you guys for your help.
I got an underlying question, customer wants to use Flash Pool with remaining data from these SSDs, the related documentation says:
"You cannot use SSDs that have been partitioned for root-data partitioning in a storage pool"
How can I configure FlashPool with this partitioning? Synergy tool shows a different a different root-data partitioning layout with 11*53.88GiB partition for root aggregate instead of 5*143.65GiB I had out of box...
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi there,
If I understand the question correctly, the system has partitioned the SSDs, so it is probably an all-SSD or AFF config system, and there is no need for a FlashPool under this configuration. You are welcome to send me a DM with the serial number of the system and I will review what it shipped with if you wish.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Per our discussion over message, we have identified that the correct solution is to the unpartition a number of the internal SSDs, in order to create a storage pool.
As this operation is potentially destructive to data, we prefer admins to do the unpartition with guidance of our support organisation over webex, referring to KB 1015130.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello and thank you for your answer.
The solution to repartition the disk sounds good but, what about the root partitions? Should I move it to the disk shelf HDD? Since we cannot do ADP on external disk shelves, this would lead us to "lose" multiple 8TB disks... Please find below the initial Synergy output, which I really have difficult to understand. The first one is a FAS2650 + 12*960GB SSD configuration, no shelf and matching with our disks layout:
But as soon as I add the shelf to the configuration:
In this configuration, where are located the 11*53.88GB parititions?
-> If it's on flash, how are the Flash Pools created ?
Many thanks for your help...
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you for the screenshot to explain what you are seeing. I have verified that I see the same output, and I also don't understand it.
I have contacted the Synergy team for explanation. I will let you know what I find out.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
As ONTAP 9.2 supports partitioning external drives on these systems, synergy uses these rules:
- FAS 2650 + internal HDD + external HDD --- internal drives are partitioned, irrespective of the drive capacities.
- FAS 2650 + internal HDD + external SSD --- internal HDD drives are partitioned
- FAS 2650 + internal SSD + external HDD --- External drives are partitioned
- FAS 2650 + internal SSD + external SSD – higher capacity drive is partitioned
So Synergy was right.
In the case of the original poster here, we have corresponded externally, and identified that the system was shipped with 9.1, then upgraded to 9.2, which does not repartition the drives to the 9.2 standard. We have identified that the most appropriate way forward is to verify there is no data present, and then wipe and repartition using option 9a then 9b from the boot menu.
