0
votes

i have created the Azure AKS cluster with autoscale feature enabled, by following the link.

Deployed Django, Celery and Rabbitmq based application and created Keda in it to scale pods based on Rabbitmq queue length. Keda is able to scale pods but Nodes are not getting scaled in node pool.

Can some one help me with it?

1
your question lacks a lot of information. how many nodes are there? what's your min/max node count? at what point do you expect the nodes to scale out? how much load is on the nodes? what are the resource limits? - Markus Dresch

1 Answers

1
votes

The following is answer I got from azure support team on this -

"Unfortunately, Autoscaling feature on Virtual machine availabilty sets is natively support by Azure kubernetes for now. We have VMSS autoscaler feature and that too in preview phase."

They were focusing on manual scaling for now.

They also mentioned one github repo to refer but azure won't provide any support for it.

Its mentioned as follows -

I have done a quick research ,please find the github link where we have procedure on autoscaling of VM availbility sets, Kindly go through standard deployment section in the link. This is not directly supported by us and if you have any issues or concerns you can approach github for the same.

click here