PurePath Analytics
Leveraging the PurePath transaction detail, dynaTrace provides a series of powerful
views and analytics built-in. These views and analytics take the large quantity
of PurePath data entering the dynaTrace server and make them digestible, actionable and
insightful for each stakeholder group – development, testers, architects, performance
specialists, operations and business owners.
For business owners, we automatically analyze every PurePath to
check them against SLAs, separate synthetics from real users, and split them by
key business transactions. This allows instant dashboard and reporting of the key
application parameters business users are looking for to make timely, informed decisions.
For operations, PurePath analytics are applied to disaggregate
application details from system details for familiar dashboard monitoring views
and faster, more accurate triage. Analytics review every PurePath against preset
thresholds and problem patterns to alert operators of hotspots and breaches. And
additional analytics are applied to aggregate PurePath details into a logical application
topology view that can be filtered and viewed by each business transaction. This
view is automatically created, fully interactive and provides end-to-end transaction
flow insight (see Advance Monitoring/Transaction Flow Map).
For architects, PurePath analytics aggregate PurePath detail to
provide an interactive UML diagram displaying all tiers, services, and interaction,
perfect for runtime validation of your distributed architecture. Analytics are also
applied across numerous parameters to highlight performance hotspots, bottlenecks
and much more (see Diagnostics). And for quick determination of what has changed
from version to version, or what is different between this transaction and that
transaction, dynaTrace has built in a Compare capability that quickly analyses the
associated PurePaths and determines what is better (green), what is worse (red)
and what is new (grey).
For testers, analytics are applied to connect PurePaths to loadtest
trouble-spots. Now pass on to development PurePath Sessions, complete with integrated
logs, alongside your load test report. For every slow performing page/component
in the report, development can now drill into PurePath detail – no more test reruns,
no more ping-pong (see Test Center Edition).
For development and performance teams, PurePath analytics automatically
calculate hotspots – longest running transactions, high number of database calls
per request, memory leaks, duplicate threads, chattiness among tiers and services
and much more. Hotspots are automatically detected to the method-level and show the full root-cause information (call stack).
These many diagnostic algorithms have been developed through years
of field engagements with direct customer feedback. With very little learning, and
no knowledge required of the detailed architecture of an application, problem isolation
and diagnostics have never been easier.