NSX Edge Gateway - On Demand Failover

[rajeevsrikant]

From NSX Release 6.2.3 , there is On Demand fail over.

http://pubs.vmware.com/Release_Notes/en/nsx/6.2.3/releasenotes_nsx_vsphere_623.html

I have the below queries.

1) I am running version 6.2.2 in my environment.

     So wanted to confirm there is no way to do On Demand Failover of NSX Edge HA either with CLI or through REST API.

     The only way to confirm is by shutting down the active Edge Gateway.

     Let me know if my above understanding is right.

2) In NSX version 6.2.3 & above its mentioned that On Demand Failover is possible.

     The below article says how to achieve it.

     So wanted to confirm this can be achieved  only via REST API or can it be done via CLI interface of NSX Edge Gateway.

Force HA Failover on NSX Edges

[mhampto]

Hello,

The API calls are not available in 6.2.2.

In 6.2.3 it is a REST API call.

To force a manual failover:

<haAdminState> was added to control which Edge is active and can be used for a user initiated Edge failover:

PUT /api/4.0/edges/{edge-id}/appliances/

<appliances>

   <appliance>

    <highAvailabilityIndex>0</highAvailabilityIndex>

    <haAdminState>down</haAdminState>

...

  </appliance>

  <appliance>

    <highAvailabilityIndex>1</highAvailabilityIndex>

    <haAdminState>up</haAdminState>

...

  </appliance>

</appliances>

This will initiate a failover on Edge with Index 0 from Active to Standby, and Edge with Index 1 will become Active.

You can also make the call on a specific Edge:

PUT /api/4.0/edges/{edge-id}/appliances/{ha-index}

<appliance>

  <highAvailabilityIndex>{ha-index}</highAvailabilityIndex>

  <haAdminState>down</haAdminState>

...

</appliance>

Does this help?

Matt

[rajeevsrikant]

on version 6.2.2 even via CLI I will not be able to achieve it.

Is my understanding is  right.

[mhampto · Accepted solution]

Correct, the functionality is not in 6.2.2.