We are assessing what work is required to make sure our vRA/vRO 7.6 workflows will port to vRA/vRO8
We have hit a snagging point on one workflow which imports an ssh key to the vRO appliance to allow it to connect to another vm to run a command.
We are using the internal vRO on the vRA Appliances and the workflow is the inbuilt workflow (Copy file from guest to vCO)
This ssh key needs to be copied to all nodes in the cluster -
In vRA7.X
We could access the API directly through each individual vRA Appliance node - we didnt need to go to the API through the Load Balancer - when we ran the workflow then it would import the ssh key into the desired vRA appliance
So basically we would
1) Use rest api to trigger workflow in vRA node 1 which would import a file into the OS (ssh key) which could be used by the embedded vRO
2) We would then do the same for node 2
3) If present then also for node 3
After this the ssh key would be present in all 3 vRA nodes - and in turn all embedded vRO nodes would be able to run the command on the remote machine as they would have that remote machines ssh key.
In vRA8
We only seem to be able to access the api through the VIP which means we cant control which node the ssh.key is being imported into. It is down to the LB and which node it balances to
My questions are
1) Is there anyway to contact the api directly on an individual appliance rather than having to go through load balancer in 8 - like there was in 7.
2) If the VIP has to be used is there anyway (other than disabling nodes) to ensure that a workflow being ran is being ran on the desired node in the vRA/vRO cluster
Any other help or guidance would be gratefully received.