Dynatrace stores and retains different types of monitored data from your environments. The monitoring data is stored on the Dynatrace cluster. The following table shows the general retention periods for service data (distributed traces), Real User Monitoring (user actions and user sessions), synthetic monitors, Log Management and Analytics, and metric time series data.
After a 15-day trial account expires, Dynatrace continues to store the monitoring data from the account for at least 30 days to ensure that no data is lost.
Dynatrace stores all monitoring data from purchased accounts for at least 35 days after account expiration. For active Dynatrace accounts, the following retention rates are set by default:
Data type
Retention rate
10 days
35 days
35 days
35 days
35 days
Synthetic
35 days
Configurable, with maximum 10 years of retention time
35 days
15 months
5 years
OneAgent diagnostics (support archives and analysis results)
30 days
14 months
Dynatrace stores the complete details of every transaction for 10 days . This enables you to analyze individual transactions and get all the details available with your instrumentation.
For trial users, an additional storage-size limit applies, which might lead to shorter retention times.
Code-level insights are available with OneAgent instrumentation for 10 days.
After 10 days, session data is optimized for aggregated views. Non-aggregated and aggregated code-level data produce comparable results for longer timeframes, while differences may be expected for shorter timeframes.
Short-term storage of the data related to service metrics used in multidimensional analysis and request charting. This data is available for 35 days with the following interval granularity levels:
A short-timeframe analysis accesses code-level data that is available for 10 days.
After 10 days, session data is optimized for aggregated views. Non-aggregated and aggregated code-level data produce comparable results for longer timeframes, while differences may be expected for shorter timeframes.
Aggregated user action metrics, which are used in tables like Top user actions and Top JavaScript errors, are available for 35 days. After 10 days, user actions data is optimized for aggregated views, and some individual user actions become unavailable for individual analysis. However, the sample set is large enough for statistically correct aggregations.
For key user actions, raw user action data is also kept for 35 days. The retention for timeseries data of key user actions is the same as for timeseries metrics.
Includes Session Replay data. All user session data is stored for 35 days. Note that waterfall analysis and JavaScript error data is stored with distributed trace code-level insights and errors.
Includes all crash data and stack traces of mobile and custom applications. The data is stored for 35 days.
Minimum size of required Session Replay storage volume is entirely load-dependent. A maximum size isn't required.
A dedicated disk is used for Session Replay data.
Log Management and Analytics enables you to ingest, process, retain and analyze log data stored in the Grail data lakehouse in SaaS environments.
With Grail storage, you don't have to worry about managing data storage performance, availability, or free space. Select the desired retention period for your logs in the bucket configuration. For log buckets, the available retention period ranges from 1 day to 10 years, with an additional week.
Log Monitoring Classic enables you to store all logs centrally within external storage. This makes log data available independent of log files themselves.
Log files are stored in Amazon Elastic File System in the zone where your Dynatrace environment resides. You don't have to worry about storage performance, availability, or free space. Disk storage costs are included in your Log Monitoring Classic subscription.
Memory dumps are immediately deleted from the disk once they're uploaded to ActiveGate. When an upload isn't possible, memory dumps up to 20 GB are stored on the disk for up to 2 hours.
Metrics powered by Grail provides a default 1-minute interval granularity for 15 months. Metrics with this granularity and retention can be accessed via Platform applications, such as Dashboards and Notebooks. Learn more at Metrics Limits.
The following interval granularity levels are available for dashboarding and API access:
To provide accurate calculations for timeseries metrics, Dynatrace uses the P2 algorithm to calculate the quantiles dynamically. This algorithm is known to yield good results and it works well with values in the long tails of value distributions. However, the aggregation algorithm is neither associative ((a + b ) + c == a + ( b + c )) nor commutative (a + b + c == c + b + a). For some metrics, for example, response times, this might lead to different quantile values each time the algorithm runs or when the data is aggregated in different ways, for example, one metric is split by URL and another by browser.
OneAgent diagnostics and ActiveGate diagnostics are optional features that enable you to collect and analyze support archives for anomalies.
Support archives are created by Dynatrace OneAgent or Dynatrace ActiveGate and stored in Cassandra, where they are automatically deleted after 30 days. When you allow Dynatrace to analyze an issue, an additional copy of the support archive is stored in the configured AWS S3 bucket. Results of the issue analysis and the support archive are also automatically deleted from the AWS S3 bucket after 30 days. Dynatrace OneAgent and Dynatrace ActiveGate do not keep copies of created support archives.
You can delete OneAgent or ActiveGate diagnostics issues at any time. If you delete an issue, the related support archive and analysis report are deleted from Cassandra and the AWS S3 bucket immediately. The analysis result in Dynatrace Health Control is deleted after 30 days.