#Add Second Node To FAS2552 Bornt As Single Node on 9.8P17 ADP

1 messages · Page 1 of 1 (latest)

brave hound
#

I got a FAS2552 system bornt as a single node on 9.8P17 with ADP.
Can I add a second node to It?

I saw that the partitions that would be assigned to the second node root aggregate are available as spares to the first node.

#

I have assigned the 9 spare partitions to the new node already.

storage disk assign -sysid XXXXXXXXX -root true -force -disk 2.0.1
storage disk assign -sysid XXXXXXXXX -root true -force -disk X.X.X
...
#

I have also defined the partner-sysid on both nodes.

#

I tried to run (4) Clean configuration and initialize all disks.
but it did not create the root aggregate.

#

After the first run I can see...
(4) Clean configuration and initialize owned disks (10 disks are owned by this filer).

brave hound
# 
bootarg.bootmenu.selection is |4a|
Sanown: Considering disks only from internal shelf
Aug 23 17:47:56 [localhost:raid.autoPart.disabled:ALERT]: Disk auto-partitioning is disabled on this system: the system needs a minimum of 8 usable internal hard disks.
BOOTMGR: already_assigned=9, min_to_boot=3, num_assigned=0
Root aggregate creation failed.
Aborting Initialization.
Rebooting... (press ctrl-c during boot to reattempt initialization)
Unable to create root aggregate: Not enough disks

Aug 23 17:47:56 [localhost:callhome.raid.adp.disabled:ALERT]: Disk auto-partitioning is disabled on this system: ADP DISABLED.
Aug 23 17:47:57 [localhost:cf.fm.noMBDisksOrIc:error]: Could not find the local mailbox disks. Could not determine the firmware state of the partner through the HA interconnect.
Aug 23 17:48:03 [localhost:raid.disk.fast.zero.done:notice]: Disk 0a.00.14P2 Shelf 0 Bay 14 [NETAPP   X423_TAL13900A10 NA01] S/N [XXXXXXXXXXXXXXXXX] UID [60000396:381070D8:500A0981:00000002:00000000:00000000:00000000:00000000:00000000:00000000] : disk zeroing complete (0x64e6465304861248).
...
System rebooting...
Uptime: 3m15s
sharp moat
#

did you set the controller and chassis ha-config from single to ha on both nodes? (in Maintenance Mode)

#

otherwise they cannot communicate through the HA interconnect

brave hound
#

Node 2 (New Node).

*> ha-config show
   Chassis HA configuration: ha
Controller HA configuration: ha
sharp moat
#

did you do it on both nodes?

brave hound
#

Node 1 (Current Node).

*> ha-config show
   Chassis HA configuration: non-ha
Controller HA configuration: non-ha
#

Fixed...

*> ha-config modify controller ha
Controller configuration will change from non-ha to ha.
Warning: Incorrect configuration may lead to data loss. Do you want to continue? [no] y
Controller FRU PROM successfully modified to ha

*> ha-config modify chassis ha
Chassis configuration will change from non-ha to ha.
Warning: Incorrect configuration may lead to data loss. Do you want to continue? [no] y
Chassis FRU PROM successfully modified to ha

*> ha-config show
   Chassis HA configuration: ha
Controller HA configuration: ha
brave hound
#

Tried to clean and initialize again... Same problem.

PANIC: raid: Unable to find root aggregate. Reason: Unknown. (DS=87, DL=10, DA=31, BDTOC=0, BDLBL=0, BLMAG=68 BLCRC=0, BLVER=0, BLSZ=0, BLTOC=0, BLOBJ=0) in SK process rc on release 9.8P17 (C) on Wed Aug 23 20:05:20 GMT 2023
sharp moat
#

hm. strange. are the 9 partitions enough? how many root partitions does the other node have? maybe it's too small for a root aggr.

#

I mean there's always the option of creating the aggregate on the other node, then reassigning it to the new node and setting the cfo option on the aggr. that should be enough to make it create an AUTOROOT volume and boot... you should be able to join the cluster then

#

(if the new node was repurposed, be sure to do a wipeconfig first if you haven't already, to clear the NVRAM from any remaining config)

brave hound
#

Thanks for dedicating some time here, @sharp moat.
I was able to get it done after some experimentation and tricky procedures.

sharp moat