# Cluster Pause > Depth-first rabbithole dossier for Cluster Pause; source-anchored research pack. Parent: [MongoDB Atlas Cost Optimization](https://llms-explorer.com/tree/mongodb-atlas-cost-optimization/) · 5 facets · 13 facts · page: https://llms-explorer.com/tree/cluster-pause/ ## Definitions - [high] If a MongoDB Atlas M10 or larger cluster is not resumed within 30 days of being paused, Atlas automatically resumes the cluster. Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md, https://www.mongodb.com/developer/products/atlas/5-ways-reduce-costs-atlas/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` ## How it works - [high] To pause or resume a MongoDB Atlas cluster, users must have Project Cluster Manager access or higher, and can initiate the action via UI, CLI, or the Atlas Administration API by setting the paused field to true. (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/architecture/current/cost-saving-config.md, https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md, https://cloudavocado.com/blog/schedule-mongodb-atlas-m20-clusters-and-cut-70-of-compute-spend-when-you-use-50-hours-week/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#tools` - [high] When a MongoDB Atlas cluster is paused, Atlas charges only for storage, ceasing charges for compute and data transfer, and paused clusters are billed at a lower rate than active clusters. Sources: https://www.mongodb.com/docs/atlas/billing/cluster-configuration-costs.md, https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md, https://www.mongodb.com/developer/products/atlas/5-ways-reduce-costs-atlas/, https://cloudavocado.com/blog/schedule-mongodb-atlas-m20-clusters-and-cut-70-of-compute-spend-when-you-use-50-hours-week/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` - [low] MongoDB Atlas recommends using M10 or M30 clusters for development and test environments, and suggests setting alerts in third-party systems to trigger if a dev or test cluster has no activity for over one week. Sources: https://www.mongodb.com/docs/atlas/architecture/current/cost-saving-config.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#patterns` - [low] Scaling scheduling across three M20 dev clusters can save approximately $300 per month on compute costs before rightsizing, though scheduling prevents burning subscription credits on idle hours without retroactively reducing commitments. Sources: https://cloudavocado.com/blog/schedule-mongodb-atlas-m20-clusters-and-cut-70-of-compute-spend-when-you-use-50-hours-week/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#patterns` ## Measurements and reference values - [high] MongoDB Atlas allows pausing M10 or larger clusters that do not use NVMe storage for up to 30 days, provided the cluster is healthy, disk usage is below 95%, and it has run for at least 60 minutes after resuming. (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md, https://www.mongodb.com/developer/products/atlas/5-ways-reduce-costs-atlas/, https://cloudavocado.com/blog/schedule-mongodb-atlas-m20-clusters-and-cut-70-of-compute-spend-when-you-use-50-hours-week/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` - [low] An M20 MongoDB Atlas cluster costs $0.20 per hour, approximately $146.72 per month for 24x7 operation, and scheduling clusters to run only during business hours (approx. 50 hours/week) can reduce compute costs by about 70%. (verified as of 2026-10-15) Sources: https://cloudavocado.com/blog/schedule-mongodb-atlas-m20-clusters-and-cut-70-of-compute-spend-when-you-use-50-hours-week/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#patterns` ## Problems, failure modes and limitations - [high] MongoDB Atlas automatically pauses inactive Free (M0) clusters after 30 days of inactivity, disallowing connections until resumed, while Flex and Serverless clusters cannot be paused and are not automatically paused. Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md, https://cloudavocado.com/blog/schedule-mongodb-atlas-m20-clusters-and-cut-70-of-compute-spend-when-you-use-50-hours-week/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` - [low] Users can retry pausing an unhealthy Atlas cluster once it returns to a healthy state. For a paused Free cluster, they can resume it at any time unless it was running on an old version that Atlas cannot restore to the current version. Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#troubleshooting` ## Facts and statements - [low] For inactive Free (M0) clusters, Atlas sends an email seven days before pausing the cluster and another email after the pause. Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` - [low] Pausing a MongoDB Atlas cluster deletes data from Search Nodes, which are recreated and indexes rebuilt upon resumption, and may pause required maintenance tasks that start immediately upon unpausing. Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` - [low] If a Backup Compliance Policy is enabled, resuming a paused MongoDB Atlas cluster automatically enables Cloud Backup. Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` - [low] Terminating a MongoDB Atlas cluster deletes any backup snapshots for that cluster, and if Termination Protection is enabled, it must be disabled before termination. Sources: https://www.mongodb.com/docs/atlas/pause-terminate-cluster.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/cluster-pause-2f418bc636/rabbithole-synthesis.md#core` ## Related concepts - Cluster — is a part of Cluster Pause - Pause — is a part of Cluster Pause