2018-07-19 10:00 AM
Hey guys, have a fun one here. We have a single SVM called SVM01-WH, this has DDNS enabled for CIFS LIF's and users connect to \\SVM01-WH - this same SVM is running iSCSI and NFS protocols.
The SVM connections are added into SC with the preffered IP address of the SVM MGMT - this works fine, I can see all disks on the hosts, I can manage them, add resources, all that stuff works. When I come to run a verify or backup job, the preffered IP is ignored and the windows SQL host tries to do its own DNS resolution, which of course resolves to the CIFS LIF.
In the good old days we would simply use command “sdcli preferredip set -f SVM01-WH -IP <IP Address>" on snap drive and everything would be fine.
1) Is this a known bug? I've emailed some contacts at NetApp but thought I would ask the community too!
2) The preferred IP should be passed down to the configuraiton of the windows plugin - does anyone know which file / regkey this would be in?
I can "fix" this by adding a host file on the SQL box, but of course it then can no longer connect to the SMB shares as the MGMT LIF isn't going to let me have them!
Yes yes, more than one SVM would help but it's too late for that and this should work by design.
Any help is appreciated peeps!
Solved! See The Solution
1 REPLY 1
2018-07-19 10:15 AM
May of just fixed this by allowing a management connection over the CIFS data LIF.
net int modify -vserver SVM01-WH -lif SVM01_WH_CIFS01 -firewall-policy mgmt
net int modify -vserver SVM01-WH -lif SVM01_WH_CIFS02 -firewall-policy mgmt