Indonesia has strict requirements about PII remaining in region.

We have recently upgraded from a shared tier to a dedicated cluster. As document, this process involved a downtime. However, what was not document is that the cluster becomes reachable for clients again before the data has been fully restored. That way, clients can invoke operations on the cluster, that may even interfere with the mongorestore process. I would propose to make the cluster available again to public clients only after all data has been restored to avoid the chance for such race conditions.

As a Project Owner, when I start rolling out dashboards to new users, I would find it very useful to see the dashboards as they can see them with their current permissions. I would like to be able to "View As" a particular user, so I can see that I have granted them the correct set of dashboards and test their permissions work for them. Obviously a feature for Owners only.

MongoDB Atlas should give each user a referral link, and a benefit of this could be that a free account could get increased storage space depending on how many/if the referred person starts paying for Atlas services.

While setting up alert allow additional text to be added. This would help to write SOP for action needed on such alert

Our team uses and loves Stitch/Realm triggers. We've got them wired up so that we can update them from a bash script. They work well, but they are hard to debug -- it'd be nice if there was a CLI option to test them against a local mongodb instance.

In both mobile and desktop version of Chrome and Safari, there is a feature for automatically generating a strong password.

In the registration form for MongoDB Atlas, the password field in the registration form cannot take values entered this way. Please add this feature since password managers are in fact quite normal in this era.

As a diabetic who relies on Nightscout but is not very experienced regarding building apps and utilising databases, etc., I just wanted to say thank you very much for the clear step-by-step mlab migration instructions and video that you provided. I was able to follow these and it all seems to have worked! Thank you.

Get rid of the dropdowns and give me a text box. Or at least an option to toggle off type safety.

I see the strength of Mongo as being untyped, but this editor is forcing in type. In addition, it is incredibly clunky to use when adding new fields by hand.

Add an alias to cluster name, as we are not allowed to change the cluster name. If we have an option to add an alias to cluster name. We can use it, in-case the user created the cluster with the wrong name. And in-screen u can show as clustername(alias). For better understanding.

I suggest that after user registered they should ve able to log in regardless of his account confirmation state, but if he is in a pending state, he could not do any read/write operation unless his own user data and his own user extra data,

And the user can also detect if his confirmation state is change

When changes, such as when collections are added or removed, occur to an existing database with BI Connector enabled, the BI Connector needs to be cycled off and then on. This process is not well explained for Atlas. I feel that a refresh button that ripples Bi Connector in Atlas when changes occur would be beneficial, rather than having to dive in to edit configurations --> slide the "enable bi connector" from off, wait, than back to on.

As per the mongorestore

1. to preserve UUID must use drop collection
2. if the collection is dropped then it will affect user/login access.

This cascading effect makes it's impractical to use. We wanted to use mongorestore UUID as an idempotent operation, in case halfway things go wrong and have to start again then it's useful to preserve UUID.

Also, in the mongorestore there should be an option to ignore/skip index creation, we were trying to bring data from the legacy app without indexes. It seems that's not possible at this moment.

Confluent cloud -Mongodb Sink connector requires Atlas whitelist IP to 0.0.0.0/0 which is a security issue on our side. Can mongodb partner with confluent and provide a solution that can be accessed through VPC private link? We are very interested in mnogodb Sink Connector for our solution so we don't have to maintain aws ec2 instances for the sink connectors.

https://www.confluent.io/blog/confluent-cloud-offers-kafka-mongodb-connector-fully-managed-clusters/

Add some way to restore old aggregations pipelines from history, that way, even if I accidently closed the UI. I will have backup.

Please could you extend the modify cluster API so that when an M20+ cluster is downsized to the next smaller size, the API command executes a sanity check to see if the smaller size has sufficient resources?

Return a 40x if the smaller size has insufficient resources.

Many user have Production systems that don't need huge resources over night so smaller machine size is sufficient

The cloud atlas frontend is very buggy, and not usable at all, even with a fast connection takes minutes to load a page on click.

https://cloud.mongodb.com/

I think we are loading all the components and required data from server side.
Instead we can simply load the core components, and show a user friendly spinner, that will stop spinning once content is loaded via an XHR request. This would be so much more user friendly, just showing them a spinner (maybe estimated time as well).

Currently the page stays blank white, which is terrible UX.

To avoid use of terminology that for some will have racial connotations remove the term "whitelist" from the Atlas UI.