# FTDC Storage Overhead & Retention Budgeting > Depth-first rabbithole dossier for FTDC Storage Overhead & Retention Budgeting; source-anchored research pack. Parent: [FTDC and Log Tooling](https://llms-explorer.com/tree/ftdc-and-log-tooling/) · 7 facets · 28 facts · page: https://llms-explorer.com/tree/ftdc-storage-overhead-retention-budgeting/ ## Definitions - [low] FTDC is the first artifact MongoDB support asks for on a performance ticket, and engineers work from the full second-by-second record. Sources: https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` ## Structure and components - [high] FTDC data never contains samples of queries, query predicates or results, data sampled from end-user collections or indexes, or credentials/certificates; it captures metadata and does not inspect network packets. It does contain host information such as hostnames, OS information and startup options. (A Percona author found no database or collection names in one 228 KB sample.) (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md, — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [low] FTDC data contains host machine information such as hostnames, operating system information, and the options or settings used to start `mongod` or `mongos`. Sources: https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` ## How it works - [low] Reports aggregate metrics for FTDC collection runs since the MongoDB process started, including the number of runs that met or exceeded the configured sampling period. Sources: https://www.mongodb.com/docs/manual/reference/command/serverStatus.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [high] FTDC thread failures are fatal and stop the parent `mongod` or `mongos` process. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#troubleshooting` - [high] FTDC data is lost when a host is terminated or rebuilt. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#troubleshooting` ## Measurements and reference values - [low] On an idle cluster, FTDC usage is about 0.93 MB per 32 minutes per node, resulting in a retention of close to ten days; on a busy production cluster, expect two to five days of retention. (verified as of 2026-10-15) Sources: https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` ## Problems, failure modes and limitations - [high] MongoDB stores FTDC data only on the local host, and you cannot use FTDC data forwarded by OTel or other log shipping pipelines for diagnostic purposes. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [low] In rare circumstances, a write can fail due to cache pressure, causing MongoDB to issue a `TemporarilyUnavailable` error and increment the `temporarilyUnavailableErrors` counter in the slow query log and FTDC. Sources: https://www.mongodb.com/docs/manual/reference/parameters.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#troubleshooting` ## Comparisons and alternatives - [high] FTDC reports utilization statistics from the perspective of the container, calculating memory utilization against the container's RAM limit rather than the host operating system's limit. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [low] A new 10MB FTDC file may be rolled every 8 minutes instead of the typical 4-8 hours in some observed cases. (verified as of 2026-10-15) Sources: https://jira.mongodb.org/browse/SERVER-85636 — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#troubleshooting` ## Facts and statements - [low] On self-managed servers, you can control FTDC retention by raising `diagnosticDataCollectionDirectorySizeMB` or copying the directory to durable storage on a cron. Sources: https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#patterns` - [high] FTDC is enabled by default, and these defaults are designed to provide useful data to MongoDB engineers with minimal impact on performance or storage size. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [high] FTDC does not include data related to storage capacity. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [low] The metrics page for FTDC data typically has a few dozen series and seven days of retention. (verified as of 2026-10-15) Sources: https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [high] FTDC captures data every 1 second, writing approximately 5,700 metrics per second into the `diagnostic.data` folder. (verified as of 2026-10-15) Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md, https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [high] FTDC data files are compressed and not human-readable; the data is delta-encoded and compressed so aggressively that days of history fit in a few hundred megabytes. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md, https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [high] FTDC collects statistics produced by specific commands on file rotation or startup and periodically collects statistics produced by other commands. Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [low] Modify FTDC default values only if requested by MongoDB engineers for specific diagnostic purposes. Sources: https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [low] The FTDC bundle includes a metadata document with `getCmdLineOpts` that indicates how Atlas starts `mongod`. Sources: https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [low] Decoding FTDC data is CPU-bound and takes minutes, so commands default to the last 4 hours and skip older chunks before decompressing them. Sources: https://github.com/saqibameen86/mdbkit — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#core` - [high] To disable FTDC, start `mongod` or `mongos` with `diagnosticDataCollectionEnabled: false` in the `setParameter` settings (not recommended for production). Sources: https://www.mongodb.com/docs/manual/administration/full-time-diagnostic-data-capture.md, https://www.mongodb.com/docs/v7.0/administration/full-time-diagnostic-data-capture.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#tools` - [low] ComputeStats takes an FTDC diagnostic file as an io.Reader and computes statistics for all metrics on each chunk. Sources: https://pkg.go.dev/github.com/jenunes/ftdc-utils — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#tools` - [low] ComputeStatsInterval takes an FTDC diagnostic file as an io.Reader and computes statistics for all metrics within a given time frame, clipping chunks to fit. Sources: https://pkg.go.dev/github.com/jenunes/ftdc-utils — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#tools` - [low] Accessing FTDC data via API requires a programmatic API key, not a database user, and if an access list is required, the IP must be added to avoid ORG_REQUIRES_ACCESS_LIST errors. Sources: https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#tools` - [low] A script can send all FTDC metric data to a Dockerized InfluxDB instance. Sources: https://www.percona.com/blog/enhancing-keyhole-pulling-more-metrics-from-mongodbs-ftdc/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#tools` - [low] The described FTDC analysis was run against an Atlas M10 on MongoDB 8.0.29. (verified as of 2026-10-15) Sources: https://www.percona.com/blog/the-diagnostic-data-mongodb-atlas-doesnt-hand-you/ — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#methodology` - [low] An increasing `delayed` value may indicate that the server is under heavy load, which can delay FTDC data collection. Sources: https://www.mongodb.com/docs/manual/reference/command/serverStatus.md — source: `~/.global-ai-hub/research/.pipeline/mongodb-2026-10-14/frb/ftdc-storage-overhead-retention-budgeting-529cb1c2e1/rabbithole-synthesis.md#troubleshooting` ## Related concepts - FTDC — is a part of FTDC Storage Overhead & Retention Budgeting - Storage — is a part of FTDC Storage Overhead & Retention Budgeting - Retention — is a part of FTDC Storage Overhead & Retention Budgeting