Assigned all disks to one controller under one Aggregate

[#1]

Hello Community,

I am not sure if the configuration I am after is the proper Netapp way of setting up aggregates.

I have a Netapp with version 7.0.4 firmware.

It has a total of 10 disks. I want one giant Aggregate and create several LUNs for my Citrix Xenserver pool to use.

I did a factory reset and it split up the disks to each controller (5 on each).

I then unowned them from one of the controllers and assigned the 5 to the other controller.

Now I have one giant Aggregate, but I am getting cluster messages on my console:

NORWLKCTSAN01*> Thu Jan  1 11:28:11 EST [NORWLKCTSAN01: cf.fsm.firmwareStatus:info]: Cluster monitor: partner kernel loaded

Thu Jan  1 11:28:17 EST [NORWLKCTSAN01: cf.fsm.takeoverByPartnerEnabled:notice]: Cluster monitor: takeover of NORWLKCTSAN01 by netapp01c1 enabled

Thu Jan  1 11:28:20 EST [NORWLKCTSAN01: cf.fsm.takeoverByPartnerDisabled:notice]: Cluster monitor: takeover of NORWLKCTSAN01 by netapp01c1 disabled (unsynchronized log)

Thu Jan  1 11:28:33 EST [NORWLKCTSAN01: asup.throttle.drop:info]: Too many autosupport messages in too short a time, throttling autosupport: PARTNER DOWN, TAKEOVER IMPOSSIBLE

Thu Jan  1 11:28:39 EST [NORWLKCTSAN01: cf.fsm.firmwareStatus:info]: Cluster monitor: partner kernel loaded

Thu Jan  1 11:28:44 EST [NORWLKCTSAN01: cf.fsm.takeoverByPartnerEnabled:notice]: Cluster monitor: takeover of NORWLKCTSAN01 by netapp01c1 enabled

Thu Jan  1 11:28:47 EST [NORWLKCTSAN01: cf.fsm.takeoverByPartnerDisabled:notice]: Cluster monitor: takeover of NORWLKCTSAN01 by netapp01c1 disabled (unsynchronized log)

Thu Jan  1 11:29:00 EST [NORWLKCTSAN01: asup.throttle.drop:info]: Too many autosupport messages in too short a time, throttling autosupport: PARTNER DOWN, TAKEOVER IMPOSSIBLE

Thu Jan  1 11:29:06 EST [NORWLKCTSAN01: cf.fsm.firmwareStatus:info]: Cluster monitor: partner kernel loaded

Thu Jan  1 11:29:11 EST [NORWLKCTSAN01: cf.fsm.takeoverByPartnerEnabled:notice]: Cluster monitor: takeover of NORWLKCTSAN01 by netapp01c1 enabled

Thu Jan  1 11:29:15 EST [NORWLKCTSAN01: cf.fsm.takeoverByPartnerDisabled:notice]: Cluster monitor: takeover of NORWLKCTSAN01 by netapp01c1 disabled (unsynchronized log)

Thu Jan  1 11:29:27 EST [NORWLKCTSAN01: asup.throttle.drop:info]: Too many autosupport messages in too short a time, throttling autosupport: PARTNER DOWN, TAKEOVER IMPOSSIBLE

Thu Jan  1 11:29:33 EST [NORWLKCTSAN01: cf.fsm.firmwareStatus:info]: Cluster monitor: partner kernel loaded

I then disabled Cluster Takeover via the GUI and I am now getting the below messages on my console:

NORWLKCTSAN01*> Thu Jan  1 11:57:00 EST [NORWLKCTSAN01: cf.fsm.firmwareStatus:info]: Cluster monitor: partner kernel loaded

Thu Jan  1 11:59:11 EST last message repeated 4 times

Thu Jan  1 11:59:44 EST [NORWLKCTSAN01: cf.fsm.firmwareStatus:info]: Cluster monitor: partner kernel loaded

Thu Jan  1 12:00:00 EST [NORWLKCTSAN01: kern.uptime.filer:info]:  12:00pm up 17:06 0 NFS ops, 0 CIFS ops, 669 HTTP ops, 0 DAFS ops, 0 FCP ops, 0 iSCSI ops

Thu Jan  1 12:00:01 EST [NORWLKCTSAN01: statd:ALERT]: Cluster is licensed but takeover of partner is disabled

Thu Jan  1 12:00:01 EST [NORWLKCTSAN01: monitor.chassisPowerSupply.degraded:notice]: Chassis power supply 2 is degraded: PSU 2 Fan 2 is spinning below the critical level. Replace it immediately to avoid overheating.

Thu Jan  1 12:00:01 EST [NORWLKCTSAN01: cf.ic.hourlyRVnag:error]: Cluster Interconnect sessions with partner have been DOWN for 0 minute(s)

So my goal is to have one large aggregate, many LUNS under that aggregate and the ability if one controller stops working, then the other controller will pick up the aggregate and things will transition without downtime.

Thank You in advance!

[#2 · Accepted solution]

Ian,

to cut things short: You cant.

BOTH of the filer heads do need at least 2 disks to be properly up and running. So if you assign all of the 10 disks to one head you cannot fail over to the 2nd one and you will get plenty of errors as the controller is constantly rebooting.

Regards,

Thomas

[#3]

Hi Thomas,

So I should maybe add 2 small disks to my Netapp. Assign them to that controller?

basically create a new small aggregate for that other controller using those 2 small disks?

This will then allow both of my controllers to properly work in the event of a failed controller?

If my main controller went down then the secondary controller (which would contain 2 small disks) would be able to still handle everything?

Thanks again.

[#4]

Yes to all.