how to enable auto-scaling for swarm-worker-vmss on the basis of Memory usage

[shivamgupta211]

Expected behavior

• Wanted to enable the auto-scaling on swarm-worker-vmss, for our memory hungry services i.e. when average memory usage of swarm-worker-vmss is greater than 70%, then worker instance should be increased
• wanted to see memory % usage for swarm manager and worker as well on the metrices section.

Actual behavior

• Not able to find any rule or way to add auto-scaling for swarm-workers on the basis of their Memory usages.
• Getting the memory metrices stats/logs in the MBs, but wanted to get in terms of %age.

Information

• Full output of the diagnostics from "docker-diagnose" ran from one of the instance

swarm-manager000000:/$ docker-diagnose
curl: (7) Failed to connect to 10.0.0.4 port 44554: Connection refused
OK hostname=swarm-worker000000 session=1533969549-DxagcZ10QEbaxhQddRUr6eShWjlkusMr
Done requesting diagnostics.
Your diagnostics session ID is 1533969549-DxagcZ10QEbaxhQddRUr6eShWjlkusMr
Please provide this session ID to the maintainer debugging your issue.