Atlas
- A brief description of what you are looking to do
- How you think this will help
- Why this matters to you
169 results found
-
Integrar la alerta de Replication lag de Atlas en el API de Prometheus
Se solicita incluir métricas en el API de prometheus para poder implementar la alarma de Replication lag
6 votes -
Ability to "Mass Kill" slow running queries
Currently, Atlas has a "Kill Op" option which is useful to kill single long-running queries.
When upgrading to MongoDB 7.0, we were faced with a situation where the Slot-Based Query Engine (SBE) was causing 1000s of queries to execute slowly, we wanted to kill them all, but it was more than a human could do by clicking "Kill Op" 1-by-1. Hence a "Mass Kill" feature which kills queries longer than X seconds (X is configurable) would have helped us greatly in an outage scenario. We ultimately rebooted our cluster to kill queries, then manually implemented a script which did this…
7 votes -
Atlas metrics granularity after 48 hours
For metrics older than 48 hours, the data is presented in 1-hour intervals. This level of granularity is often too coarse for a thorough examination of past events and trends. Such a broad view can obscure smaller yet significant details critical for understanding and resolving performance issues that occurred in the past.
Suggested Improvement:
having a smaller granularity value for historical metrics beyond the 48-hour timeframe. Providing data in smaller intervals would greatly enhance our ability to conduct in-depth analyses and diagnose past performance issues accurately. This would be particularly beneficial for conducting detailed investigations of historical data and identifying…
4 votes -
Prometheus database and collection metrics
We checked the Prometheus metrics provided by MongoDB Atlas and didn't find metrics for the following:
Database sizeCollection storage size
Record per collection
Indexes per collection
Index size
We would like to have this kind of metrics to add to dashboards.
20 votes -
Add document count to Datadog metrics
We'd like to monitor the number of documents in a collection via DataDog.
For On-Premise MongoDB the stats are already reported via mongodb.collection.count, mongodb.collection.size and mongodb.collection.avgobjsize.
If the same metrics could be made available for Atlas (E.g. mongodb.atlas.stats.collection.count) that would really help in monitoring.
E.g. spikes in different parts of the application could be tied to the number of documents on a glance. Without having that metric available, it is hard to pinpoint if a recent change had a negative impact on performance.
If the metrics can't be made available on the collection but only on the database level, this…
8 votes -
View cluster/storage scalable (yes/no) in "All cluster" dashboard
Right now in the "all cluster" view, one cannot see, if the clusters are enabled for auto scaling. my idea is to add this also to avoid last minute performance alerts for non scaling enabled clusters
1 vote -
Collection query audit logs
It would be nice to be able to see what queries were run on which collection over the last x days.
1 vote -
MongoDB Atlas historical stats by collection
It would be helpful to have historical metrics for storage by collection. Storage space used, indexspace used, number of indexes.
Thanks1 vote -
Metric reporting private endpoint state
On Mongo Atlas platform we are able to see the status of both Atlas Private Endpoint and Azure Private Endpoint. It would be helpful to have these statuses available as a metric on the prometheus integration.
4 votes -
Change Streams Monitoring and Alerting
Change streams can cause performance issues if not used properly. In some cases, administrators of multi-tenant dbs have no control (and shouldn't) over how various clients create change streams.
I think it is important that we accommodate these use-cases and provide useful metrics in the OM/Atlas metrics pages, and alerts on those metrics. Some potential metrics:
1. Number of change streams open
2. Average change stream lifetime
3. Query targeting ratios for change streams
4. Avg time between consecutive polls of the change stream (and other statistics)
--thought here is that change streams that are polled infrequently will result in…5 votes -
Improve Metric Correlation
Would be nice to better correlate metrics, server events, and individual operations.
For example would be helpful to have in the profiler indication of automatic scale up/down of the cluster in order to easily correlate actions that trigger them in the last 24h.
1 vote -
Metric Grouping
A huge improvement and help when it comes to metrics would be the ability to query by grouping (e.g. for database access users). This way if you were to use a specific database user per a specific service connection, we could see how much load to the database that specific service is causing.
Any form of implementation would be helpful, one example could be adding labels to the prometheus metrics per user, replica/shard etc.1 vote -
Throttle instances within a shared tier
Right now, if a cluster is in a shared tier, other "noisy neighbor" tenants can destroy performance by overusing the resources, forcing a restart, causing unstable behevior etc.
Throttle noisy neighbors so that others don't suffer!
1 vote -
Dump audit logs to log analytics workspace
Why can't we dump audit logs to log analytics workspace when we are using azure private endpoints to connect atlas
1 vote -
Investigation Buckets (dropdown menu) to enable predefined set of MongoDB & Hardware Metrics to investigate certain Data Layer SLAs
We have many metrics under "MongoDB Metrics" & "Hardware Metrics". User will have a good idea which metrics to enable while troubleshooting certain requirements.
E.g. User is investigating current lags in the data Layer. User has to select all the metrics that would help them get the data they are looking.
We can help the customer bit more by provide buckets like.
Options ( select one )
1. Investigate Lags
2. Investigate IOPS
3. Investigate Replication
4. Investigate Search Index
... and etc..This will pre-select certain metrics from list of Mongodb & Hardware metrics which will give them all…
1 vote -
Monitoring Metrics on dhandle
We'd like to monitor the WiredTiger dhandle over the time, directly from Cloud Atlas Monitoring view. It would allow to directly see the impact when updating cluster settings.
We'd like also being able to configure alert triggers on it, the goal for us is being alerted when an excessive amount of files (collections & indexes) is loaded into the MongoDB Memory, to avoid reaching an Out Of Memory error.
2 votes -
prometheus integration to use PrivateLink
There is a possibility to integrate Prometheus into an Atlas project.
However, for enabling this integration, one needs to add Prometheus's IP address in the IP Access List.
This procedure has 2 flaws in it:
1. Prometheus runs as pods on some use-cases, meaning that its IP is ephemeral.
2. For projects that work solely with PrivateLink enabled and no open IP in the IP Access List, one cannot use the Prometheus integration (already talked with support about that).The improvement here is to add the Prometheus integration to work as well in "PrivateLink-only" mode.
74 votes -
Export only selected Atlas metrics
This is a request from a customer named: UPAX partnered with Grupo Salinas.
Inside the cluster metrics, where the Status is displayed, we are able to find the Export option (PDF, PNG) and they would like to export for only specific metrics (only the toggle charts) not the whole of them for reporting purposes with other teams and management.1 vote -
Duplication detection
One of our custmoers would like to see duplication detection as their project is spread out over various projects.
Value: It is important as different regions create sometimes duplicates and duplicates can be avoided in advance.
1 vote -
Metrics: Fix Y axis of "max CPU %" charts to 100%
In the various CPU % usage charts in the shard metrics, it looks like the upper limit of the Y axis isn't fixed, and it changes based on what the maximum value in the range of time being displayed is. For both normalized and regular CPU% values, there's a natural maximum to the value, I think: 100% for the normalized CPU%, and ncores*100% for the non-normalized CPU%.
With a changeable Y limit, it's hard to tell at a glance whether the value is actually really high or not. Oh, the blue line is hovering up near the top of the…
1 vote
- Don't see your idea?