Atlas
- A brief description of what you are looking to do
- How you think this will help
- Why this matters to you
1258 results found
-
Bulk explain() command
I want to be able to run the "explain" command for multiple database operations so that I can quickly estimate the RPUs and WPUs for my workload (instead of having to do it one by one for each operation).
2 votes -
Allow multiregion sharded clusters to have Primary nodes in multiple regions
For multi-region sharded clusters, the primaries for each shard remain in a single (highest priority) region in Atlas. If these primaries can be spread across multiple regions, that will a) provide consistent average response time for applications that connect from multiple regions as active-active b) Limit the blast radius if the highest priority region is unavailable c) Better distribution of workload from an application side.
This architecture is currently possible with EA deployments but not with Atlas.
6 votes -
Private endpoint termination protection
Similar to the Cluster Termination Protection, it would be nice to have Private Endpoint termination protection to prevent accidental deletes of private endpoints, which could very likely result in application connectivity loss and downtime.
5 votes -
Map IDP groups to Atlas teams
At the moment Atlas does not support mapping IDP groups to existing Atlas teams. We would like that the integration would support that. For example:
okta group "devops" --- mapped to ---> Atlas team "devops"
Each time the customer adds a user to this idp group, the user will be given the proper permissions in Atlas.
10 votes -
tgw support for Atlas
Can you please provide TGW support for Atlas? We are using confluent.io kafka with TGW, and found ourselves caught up in situation where kafka has no direct access to atlas, both our EKS and kafka are directly connected to Kafka, but there's no way for kafka to access atlas, because EKS and atlas are connected over vpc peering. We do not take interest in PrivateLink. Our Kafka needs direct access to Atlas because we are running the mongo source connector. Only viable solution now is to run the mongo source connector on our EKS, which while works defeats the purpose…
2 votes -
Retrieve latest/max oplog timestamp via the Atlas Administration API
At the moment the latest/max oplog timestamp can be retrieved via the Atlas UI in the backups section for point in time restores. This feedback post is to make this information available as part of the atlas administration api.
For example, in the UI it states:
You can only restore up to this oplog timestamp: 1692580549 and this increment: 1.
Would be good to have this value available to be retrieved via the API for restores requiring oplog timestamp values in this format.
3 votes -
Need a Instance with unlimted storage .
We are seeking an instance with 2 vCPUs, 4 GB of memory, and a storage capacity of 500 GB.While we understand that the available M20 instance type supports the desired vCPU and memory specifications, we noticed that the maximum storage capacity for M20 instances is limited to 256 GB. Unfortunately, this falls short of our project's storage needs.
Considering this, we kindly request assistance in identifying an alternative instance type or solution that can meet our requirements for 2 vCPUs, 4 GB of memory, and a larger storage capacity of 500 GB. We would appreciate your guidance in selecting an…1 vote -
Support europe-west4 in GCP
As the title says really, we want to make extensive use of serverless but can't do so until you support west4. You support west4 on dedicated instances, but support say there are no plans for west4 on serverless, which is disappointing.
1 vote -
Serverless queries could return their RPU & WPU along with the results
Currently it's impossible to control the serverless spending. By being able to log and alert on queries resulting in huge RPU/WPU usage during development it would be safer to release changes without billing surprises
1 vote -
Support migration of existing projects, clusters, ... to be CloudFormation managed
When MongoDB::Atlas::* resources are created and the specifed project, cluster, ... already exists, then these resources should not raise an error on resource creation but use the existing project, cluster, ... and apply the given configuration. In other words, the Create event that is initially sent by CloudFormation should be handled similar to an Update event but return the physical id of already existing resources. Apparently, this behaviour should be enabled by an opt-in flag.
3 votes -
Modify MongoDB Autoscaling Logic for CPU Usage Interval
Currently, the autoscaling feature in MongoDB is configured to initiate scaling based on the last 1-hour CPU usage. The trigger for autoscaling is set at 75% CPU usage within this hourly interval. However, there is a requirement to adjust this logic to a more granular level.
Request:
The request is to modify the autoscaling logic in MongoDB to consider the last 30 minutes of CPU usage instead of the current 1-hour interval. Specifically, scaling should be triggered if the CPU usage exceeds 75% within the last 30 minutes.
1 vote -
Enable database users separate read/write permissions based on cluster
We have multiple clusters in the same project, and we need to limit write access to certain CLUSTERS while leaving read only access to any other cluster. The problem is that we cannot do this by DATABASE name as there are many of them, as well as dynamically addition of new ones in clusters, so it will become obsolete and very hard to maintain.
Trying to use the restriction does not help as it limits all access, including read only, to those clusters.
What we'd like to see is that we can limit write access by cluster (not database/collection) and…3 votes -
Add a data source for the Atlas Control Plane inbound/outbound IPs
It would be nice to be able to fetch all outbound and inbound IPs for the Atlas control plane so that we can add them to the list of allowed IPs on Azure keyvaults for example
3 votes -
Migrate specific collection / database from a replica set to sharded cluster(not empty) without downtime.
It would be great if MongoDB has a tool to migrate specific collections/databases from a replica set to sharded cluster and vice versa. Migrating data with this tool should not cause any downtime in both source and destination clusters.
2 votes -
Sync Atlas alert configurations between projects.
Since we can export all Alert configurations from a project, add the ability to sync alerts on all the projects to be consistent with notifications/alerts.
Alerts can be created individually but not in bulk. Please enhance the Mongo Atlas API for the following functionalities:
- The ability to sync all alerts from one project to another.
Thank you
1 vote -
Specifying who receives notifications for auto-scaling events
Having ability to manage wich roles receive auto-scaling notifications when a cluster scales up or down.
Now this is limited to Project Owner but this feature will be helpfull to manage notifications to inform other Team Members wich doesn't have to get all permissions related to Project Owner.
1 vote -
1 vote
-
Google Private Service Connect
Greeting from Fivetran!
This is somewhat related to https://feedback.mongodb.com/forums/924145-atlas/suggestions/45272014-allow-customers-to-specify-the-number-of-service-a . Having 50 service attachments is not scalable for us which requires 50 IP addresses for each PSC. We have a large customer base and having each of them create PSC would require a lot of IP addresses and would quickly exhaust our subnets.
From the support case it seems that the decision to use 50 PSC attachments comes from the fact that GCP load balancer does not allow more than one pool of servers per service attachment and that the ports are passed through as is as opposed to AWS…
4 votes -
On-Demand Pricing
I have been an enthusiastic MongoDB user for a couple of years now. My business is very small and the utilization of the database is very irregular. For example, I only have a higher workload on one day of the week and on most other days I only have a load in the morning and often nothing at all at night and at weekends. That's why the smallest available plan that can also be used in production (M30) is actually too expensive for me. A pay-per-use pricing concept would be ideal and I'm not sure whether this would be possible…
1 vote -
kill process for rolling index build from atlas
If index builds are rolling, then the Atlas console does not have the functionality to kill them if they take a longer time on a node. Is it possible to kill it like regular index creation from atlas UI instead raising a ticket to mongo
2 votes
- Don't see your idea?