HA setup

[deuch]

Hello,

We have an issue regarding HA and ONTAP series.

I have 2 NAS, on primary and one backup (with 1h asynchronous replication) on 2 DataCenters.

So in fact, i need 2 config files like those

For DC1 { "version": 1, "storageDriverName": "ontap-nas", "managementLIF": "DNS-adm", "dataLIF": "DNS-nfs1", "svm": "SVM-DC1", "username": "docker", "password": "xxxxxxxxxx", "aggregate": "aggr1_DC1" }

Pour DC2 { "version": 1, "storageDriverName": "ontap-nas", "managementLIF": "DNS-adm", "dataLIF": "DNS-nfs1", "svm": "SVM-DC2", "username": "docker", "password": "xxxxxxxxxx", "aggregate": "aggr1_DC2" }

For the management LIF and dataLIF, we're using a DNS name with a swith in case of failure of the NAS or the DC (few seconds to switch).

But we've 2 differents parameter for svm and aggregate ... I think we need to have the possibility to set a primary and secondary parameter for svm and aggregate. In case of NAS failure, the plugin will re-mount the volume on the healthy NAS/DC.

Is it something possible to do in the plugin ?

[adkerr]

Hi @deuch,

I do not believe what you're requesting is currently possible, but is something we could consider for a future release.

[deuch]

Ok thanks. We plan to use ONTAP NAS for statefull containers, but if the plugin can not handle a HA/DR situation, it's a not go for us.

Actually, we have to stop all the containers with netapp volumes, stop the plugin, modify the plugin configuration, restart the plugin (and maybe the engine) and restart all containers ... Not so safe for a production environment ! :-D

[adkerr]

Tracked internally by [US2476]

[adkerr]

Moved to netapp/trident#80