We would like to install Cloud Manager Agent on arm64 based linux machines.

Submitting on behalf of customer - Request to add Cloud Manager support for Azure servers (provisioning, installing an agent, making a custom hostname and doing DNS) and Azure blob for backups. Customer has shared they have had a great experience using Atlas. However due to business requirements, some application stay must on prem. Customer is partnered with Azure and unable to use Cloud Manager for on prem deployments as they cannot use AWS for servers and backups

My automation manager posts an info message about unexpected fields in process memory metrics map every minute. It's polluting my logs. I would like a way to fix the underlying problem or, if it cannot be fixed, then don't tell me about it in the logs.

Here are some examples:
mongodb-data-prod-vm2.cloud.ca 2020-08-05 - 01:10:47 PM info runtimestatsPrinter Unexpected field in process memory metrics map: thp_enabled. Skipping

mongodb-data-prod-vm2.cloud.ca 2020-08-05 - 01:10:47 PM info runtimestatsPrinter Unexpected field in process memory metrics map: speculationstorebypass. Skipping

What is the problem that needs to be solved? MongoDB Agent service (mongodb-mms-automation-agent.service) does not start (by default (disabled; vendor preset: disabled)) automatically after OS reboot.

Why is it a problem? (the pain) OS restarts can be frequent, and current default behavior (disabled; vendor preset: disabled) of MongoDB Agent not to start after OS restart causing additional downtime for customers.

With the Cloud Manger API it is possible to create new indexes. But there is no possibility to delete indexes by the Cloud Manger API.

There doesn't seem to be an equivalent for Ops Manager All Clusters page in Cloud Manager. It is a good way to quickly glance over all cluster states, stats and versions.
Can it be added?

When creating new clusters in Cloud Manager, there isn't an option to create them with a specific FCV. We just had a need to build a 4.2 cluster with FCV 4.0, and had to do it manually while shutting down automation.

When we create a new collection with a shard key using mongosh the same is not reflected in cloud manager. In cloud manager / ops manager the shard key for the new collection is not seen in the shard tab.

The FCV could be different than the MongoDB version. Also, sometime the FCV on different shards/CSRS in a sharded cluster might be different (e.g. FCV is upgraded on the shards, but for some reason FCV is not upgraded on the CSRS), and this could cause issues.

It would be nice if FCV is displayed next to the MongoDB version information. So that we can spot out the discrepancy quickly.

What is the problem that needs to be solved? Ability to set alerts at a global level for all projects, similar to functionality currently in Ops Manager

Why is it a problem? (the pain) There is a basic set of alerting for all clusters that we use; instead of setting up each project with the same alerts, have a global alert level that covers all projects. Ops Manager currently has this functionality

I am not able to find a way to connect cloud manager to on prem docker containers running MongoDB. Maybe i've missed something and this feature already exists?