ONTAP Hardware
ONTAP Hardware
Hello everyone,
We are planning to extend the storage capacity of a FAS 2240-2 (Ontap 8.2.4 in 7-Mode) with 4 DS14-MK2-AT shelves.
The FAS 2240-2 has currently 3 DS2246 shelves attached and running well which I cabled as per the recommanded guide.
I am concerned about the cabling between these shelves as I don't see any documentation.
Does anyone know how to cable FAS 2240-2 with DS2246 and DS14-MK2-AT shelves ?
https://library.netapp.com/ecm/ecm_download_file/ECMP1139842
HA pair with FC shelves and FC tape backup (page 26) - DS14MK2-AT connect to Mezzanine FC ports
HA pair with external SAS shelves (Page 10) - DS2246 connect to the onboard SAS ports
The e0a,e0b,e0c,e0d onboard ports connect to host's for iSCSI, NFS,CIFS
no other ports available.
hope it helps
That's a good doc to point to... but Example 12 (pg 29) is even more apt (no tape devices involved - just HA cabling DS14 shelves)
Hi there!
Maximum drive count for a FAS2240-2 is 144 drives.
internal shelf + 3 external = 4 * 24= 96 drives
4 x DS14-MK2-AT = 56 drives
56 + 96 = 152 drives. Just leave 6 8 disks out.
Good point, Alex... but the math dictates pulling 8 disks out...
Thank you guyes for your help and replies and I apologize fot the delay as I was out of the office moving the shelves to another datacenter.
If my understanding is good, to get the DS14MK2-AT shelves attached to the FAS2240-2 I need a Mezzanine card in each one of the controllers. Once these cards are up I can then connect the DS14MK2-AT shelves following the guide. Does this operation require any specific configuration ? is it hot pluggable ? is it autoconfiguration ?
Also, I will not need to mix connection between DS14MK2-AT and DS2246 at anytime ? I was confused because DS2246 uses ACP cables and DS14MK2-AT does not.
Thanks again.
If you don't already have the FC mezzanine module, I'm not sure where you'll be able to find them - NetApp no longer sells this platform or module.
The marketing part number is X1150A-R6.
You have to pull the PCM canister to install this module.
The ports are by default ready for attachment to DS14 storage shelves (assuming you have the cables).
There is no mixing between the FC-AL and SAS shelves. They are separate storage domains. ACP is only applicable to the SAS shelves and the e0P ports on the nodes.
Andris, your reply is very useful and relevant.
I went back to the DC and took a picture of the FAS 2240-2 rear's view which you can see on the attachments.
In the documentation https://library.netapp.com/ecm/ecm_download_file/ECMP1139842 (page 6), the ports e1a and e1b on both controllers are refered as 10-GbE mezzanine cards.
Could you please confirm whether these ports are the required mezzanine cards? and also which command I issue use to check they are mezzanine cards ? ( I used sysconfig -v but didn't get what I want).
Thank you so much.
Now we absouloutely need to purchase 2 mezzanine cards (PN: X1150A-R6), but how to install them, unless I am mistaken there is no free placement. Any help please ?
Hello,
Sorry for the delay as we just received the two purshased X1150A-R6 module.
I have done some component replacement (DIMM) on FAS8020 and would like you to point me out to the correct guide for the FAS2240-2.
Should I refer to this https://library.netapp.com/ecm/ecm_download_file/ECMM1280402 ?
Also, knowing that we have no FC configuration yet, I would really appreciate your help on how to configure FC correctly from scratch.
C1NETAPP-VIRTU> fcadmin config
Warning: This command is deprecated. Use the ucadmin command to configure adapters.
Local
Adapter Type State Status
---------------------------------------------------
C1NETAPP-VIRTU> cf status
Controller Failover enabled, C2NETAPP-VIRTU is up.
Negotiated failover enabled (network_interface).
VIA Interconnect is up (link up).
Thanks again.
To connect the DS14mk2 FC Shelves to the mezzanine card refer to this link on how to stack the disk shelves with the FAS224x system
https://library.netapp.com/ecm/ecm_download_file/ECMP1139842
Example 10: HA pair with FC shelves (Page 25)
then change the mezzaning fc ports from target mode to initiator mode if not configured as initiator ports, and then reboot the each controller to change mode,
refer to the link below
https://library.netapp.com/ecmdocs/ECMP1368845/html/GUID-8D8CB1A6-ECC4-4FE7-8F37-635242CD65F0.html
the system will be able to see the disk shelves and the disk drives.
Disk shelves ID should be unique
(eg. DS14mk2 shelf 1, to have ID 01 f to have ID 02... etc)
If stacking is correct, run a sysconfig -a or a fcadmin device_map to check if all the disks / shelfs appear
assign the disks to each node as you like to create the aggregates.
hope it helps
Firstly, I followed the "Adding or replacing the mezzanine card" to change our 10G Ethernet mezzanine cards with the new FC X1150A-R6 cards.
https://library.netapp.com/ecm/ecm_download_file/ECMP1141723
I went into maintenance mode and test the newly installed hardware which was OK.
I did not have to change the mezzaning fc ports from target mode to initiator mode as they were already configured so
C1NETAPP-VIRTU> ucadmin show
Current Current Pending Pending Admin
Adapter Mode Type Mode Type Status
------- ------- --------- ------- --------- -------
1a fc initiator - - Online
1b fc initiator - - Online
Secondly, I set a unique ID for each of the 4 DS14mk2 shelves (4,5,6 and 7 in my case) and rebooted them.
Then I used the below guide to connect the DS14mk2 FC Shelves to the mezzanine card (stack the disk shelves with the FAS224x system)
Example 12: HA pair with FC disk shelves (Page 29)
https://library.netapp.com/ecm/ecm_download_file/ECMP1139842
It's worth noting that these FC shelves were already used in production with another system. I was 100% sure that the stacking was correct but sysconfig -a or fcadmin device_map would not return me all the newly disks. I was able to detect this little issue by using a storage show which showed all disks and their respective owners, in the present case it showed that the old system still owning these FC disks. After that, I issued disk assign [DiskID] -s unowned -f on these disks, and the auto assign option took care to assign the newly 48 disks to my both nodes.
Aggregates and volumes creation went flawless.
I would like to thank you for all your availabiliy and assistance that you provided us with. Much apreciated.
Regards,
Radouane