DevOps / AppDynamics Interview questions
How does AppDynamics decide when to capture a transaction snapshot?
Rather than recording every single execution of every Business Transaction — which would be far too much data — AppDynamics captures snapshots selectively, triggered by specific conditions on either the originating or a continuing tier.
| Trigger | What it Means |
| Slow | Response time exceeds the baseline by a set threshold |
| Very Slow | Response time far exceeds baseline, a more severe threshold |
| Stall | The request has been running longer than a stall threshold and hasn't finished |
| Error | The transaction threw an exception or returned an error status |
| Diagnostic / Adaptive Session | A pattern of repeated problems triggers deeper, sustained sampling |
Continuing (downstream) tiers only capture their portion of a snapshot when the transaction is already flagged as slow, stalled, or erroring somewhere in the chain, and volume is capped — for example, continuing snapshots are limited to roughly 200 per node per minute — so instrumentation overhead stays bounded even under heavy load.
Which of these is NOT a typical transaction snapshot trigger?
Why does AppDynamics cap the number of continuing snapshots per node per minute?
More Related questions...
