Durations of Job Steps - Appdynamics/AppDynamics.DEXTER GitHub Wiki
The timing report is output to Reports folder as StepDurations.csv.
Duration of Extract Phase
The duration of the Extract phase depends on:
- The number of Applications to query
- More Applications => More work to retrieve anything)
- The complexity of Applications being queried
- More Tiers, Business Transactions, Configuration rules and so on => More time to retrieve Metrics and Flowmaps
- The busy-ness of Applications being queried More Snapshots => More time to retrieve Snapshots
- More Events => More time to retrieve Events
- Whether sequential or parallel mode is chosen
- Sequential => Longer
- Ability of Controller to respond to queries in timely interval, especially for Snapshot retrieval
- Slower Controller => Longer
- Your network connection
- Slower => Longer
Phase | Duration |
---|---|
ExtractController, ExtractControllerConfiguration | 1-4 seconds per Controller |
ExtractControllerUsersGroupsRolesAndPermissions | ~10 seconds per Controller |
ExtractDashboards | 10-15 seconds per 50 dashboards |
ExtractApplicationEventsAndHealthRuleViolations | 10-30 seconds for each Application |
Extract***Configuration | 2-3 seconds per Application |
Extract***Entities | 5-10 seconds per Controller |
ExtractAPMMetrics | 5-20 seconds per Entity Type (Tier, Node, Backend, etc.), depending on number of metric sets requested and number of time ranges |
ExtractAPMFlowmaps | 1-2 minutes for each 200 Entities per Application or ~100 ms per entity (Tier, Node, Backend, etc.) |
ExtractAPMSnapshots | 0.5-5 seconds per Snapshot, depending on Snapshot size |
Duration of Index Phase
During Index phase, large number of small JSON files are read, converted in memory and results are output to CSV files.
The duration of Index phase is driven by number of CPUs and CPU speed and speed of local IO system for reading. Local SSD is highly recommended. More CPUs => More work done.
Setting | Duration |
---|---|
IndexControllerUsersGroupsRolesAndPermissions | 2-5 seconds per Controller |
IndexDashboards | 1-2 seconds per 50 dashboards |
IndexApplicationEventsAndHealthRuleViolations | 3-5 seconds for each Application |
Index***Configuration | 2-3 seconds per Application |
Extract***Entities | 1-2 seconds per Application |
IndexApplicationConfigurationComparison | 2-3 seconds per Application |
IndexAPMMetrics | 5-20 seconds per Application |
IndexAPMFlowmaps | 5-10 seconds for each 200 Entities per Application Or ~100 ms per entity (Tier, Node, Backend, etc. |
IndexSnapshots | ~20 ms per Snapshot |
Duration of Report Phase
Setting | Duration |
---|---|
ReportUsersGroupsRolesAndPermissions | 5-10 seconds |
ReportControllerAndApplicationConfiguration | 2-5 seconds per Application |
ReportApplicationEventsAndHealthRuleViolations | 1-2 second per Application |
Report***Entities | 2-5 seconds per Application |
ReportApplicationAndEntityMetrics | 5-10 seconds per Application, and longer the more time intervals are present |
ReportApplicationAndEntityMetricGraphs | 30-60 seconds per Entity Type, and longer the more time intervals are present |
ReportSnapshots | Up to several minutes, if > 10K snapshots |
ReportSnapshotsMethodCallLines | Up to several minutes, if > 10K snapshots |
ReportIndividualApplicationAndEntityDetails | 5-10 seconds per Entity |
ReportFlameGraphs | 5-10 seconds per Application |