# Atlas CLI Benchmark Metrics > Depth-first rabbithole dossier for Atlas CLI Benchmark Metrics; source-anchored research pack. Parent: [Atlas CLI](https://llms-explorer.com/tree/atlas-cli/) · 4 facets · 42 facts · page: https://llms-explorer.com/tree/atlas-cli-benchmark-metrics/ ## How it works - [low] When specifying granularity, the Atlas CLI requires either a period or both start and end times to be specified. Sources: https://www.mongodb.com/docs/atlas/cli/current/command/atlas-metrics-processes.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] MongoDB Atlas provides rich cluster and shard-level metrics, but using mongodb-exporter alone does not capture metrics at the same level of detail as the Atlas UI, even when connected to all shards. Sources: https://community.grafana.com/t/how-to-integrate-atlas-metrics-with-prometheus-and-visualize-in-grafana/157754 — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The Atlas CLI package conflicts with the 'atlas' and 'nim' packages because both install the `atlas` executable. Sources: https://formulae.brew.sh/formula/mongodb-atlas-cli — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#troubleshooting` ## Measurements and reference values - [low] The Atlas CLI supports only the following ISO 8601-formatted time periods: PT10S, PT1M, PT5M, PT1H, and P1D. Sources: https://www.mongodb.com/docs/atlas/cli/current/command/atlas-metrics-processes.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` ## Problems, failure modes and limitations - [low] If the error “Error scraping target: server returned HTTP status 429 Too Many Requests” appears, it indicates Atlas has imposed a 100 requests per minute limit, requiring adjustment of the scrape_interval. Sources: https://community.grafana.com/t/how-to-integrate-atlas-metrics-with-prometheus-and-visualize-in-grafana/157754 — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#troubleshooting` ## Facts and statements - [low] The number of items per results page defaults to 100 with a maximum of 500; if results exceed 500, the --page option must be specified to change the results page. Sources: https://www.mongodb.com/docs/atlas/cli/current/command/atlas-metrics-processes.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The page number parameter defaults to 1 and specifies a page of results. Sources: https://www.mongodb.com/docs/atlas/cli/current/command/atlas-metrics-processes.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Using the Atlas CLI command requires authentication with a user account, service account, or API key holding the Project Read Only role. Sources: https://www.mongodb.com/docs/atlas/cli/current/command/atlas-metrics-processes.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas gathers metrics data at a 1-minute granularity by default, but gathers Replication Lag and Replication Headroom metrics at an 85-second granularity. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] With premium monitoring enabled, Atlas gathers metrics data at a 10-second granularity. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas retains 10-second granularity metrics data for 8 hours. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas retains 1-minute and 5-minute granularity metrics data for 48 hours. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas retains 1-hour granularity metrics data for 63 days. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas retains 1-day granularity metrics data forever. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Creating a new cluster with the same deployment type, name, and project as a previously deleted cluster carries over the historic metrics from the deleted cluster. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The Atlas CLI is stable at version 1.59.2. (verified as of 2026-10-15) Sources: https://formulae.brew.sh/formula/mongodb-atlas-cli — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The Atlas CLI uses mongosh 2.13.0 for connecting, configuring, querying, and working with MongoDB databases. (verified as of 2026-10-15) Sources: https://formulae.brew.sh/formula/mongodb-atlas-cli — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The Atlas CLI is built with go 1.27.2. (verified as of 2026-10-15) Sources: https://formulae.brew.sh/formula/mongodb-atlas-cli — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The Homebrew analytics for the Atlas CLI package show 1,412 installs in the last 30 days, 4,613 in the last 90 days, and 46,250 in the last 365 days, with 0 build errors. (verified as of 2026-10-15) Sources: https://formulae.brew.sh/formula/mongodb-atlas-cli — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Query shape metrics prefixed `mongodb.atlas.querystats` carry `samplingstrategy` and `samplingrate` tags describing how MongoDB Atlas collected underlying operations. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.latencyStats.read.p50`, `p95`, and `p99` rate metrics report the number of operations on the 50th, 95th, and 99th percentiles in the latency histogram for read requests, respectively. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.latencyStats.write.p50`, `p95`, and `p99` rate metrics report the number of operations on the 50th, 95th, and 99th percentiles in the latency histogram for write requests, respectively. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas retains all database-specific statistics, while MongoDB log data is retained at a maximum rate of 2000 lines per 2 minutes. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas pauses monitoring for Free clusters that have had no connection activity for 7 days. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Scaling a cluster to the M40 tier or higher qualifies it for premium monitoring. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas compacts each group of 60 minutes into a single hour unit, and after 63 days, compacts each group of 24 hours into a single day unit. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The value of a burst reporting metric at a given granularity is the maximum reported value from all reports generated at the next finer granularity level during that interval. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] Atlas computes the Connections metric in the 1-hour report as the average of all values across the 60 1-minute reports generated in that hour. Sources: https://www.mongodb.com/docs/atlas/monitor-cluster-metrics.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The MongoDB Atlas integration is only available on M10+ clusters. (verified as of 2026-10-15) Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] On clusters running MongoDB 9.0 or later with sample-based collection, the `samplingstrategy:sampled` tag indicates query shape metrics reflect a random sample of read and write operations. (verified as of 2026-10-15) Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] On earlier versions, the `samplingstrategy:rate_limited` tag indicates Atlas records up to a maximum of 100 read queries per second per host by default, and write operations are not collected. (verified as of 2026-10-15) Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.dbstats.num_extents` gauge metric reports the number of contiguous blocks of space in a data file allocated for a collection or index. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.dbstats.objects` gauge metric reports the number of objects (documents) in the database across all collections. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.dbstats.storage_size` gauge metric reports the sum of disk space allocated to all collections in the database for document storage, including free space. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.dbstats.views` gauge metric reports the number of views in the database. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.global.lock.current.queue.reads` gauge metric reports the number of operations queued waiting for a read lock. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.global.lock.current.queue.total` gauge metric reports the number of operations queued waiting for any lock. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.global.lock.current.queue.writes` gauge metric reports the number of operations queued waiting for a write lock. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.latencyStats.read.sum` rate metric reports latency for total read requests. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.latencyStats.read.count` rate metric reports the number of operations performed on the collection since startup for read requests. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.latencyStats.write.sum` rate metric reports latency for total write requests. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` - [low] The `mongodb.atlas.latencyStats.write.count` rate metric reports the number of operations performed on the collection since startup for write requests. Sources: https://docs.datadoghq.com/integrations/mongodb-atlas — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/atlas-cli-benchmark-metrics-8bb738b567/rabbithole-synthesis.md#core` ## Related concepts - Atlas — is a part of Atlas CLI Benchmark Metrics - Metrics — is a part of Atlas CLI Benchmark Metrics - CLI — is a part of Atlas CLI Benchmark Metrics